Case Reports

Sarcoidosis is a systemic inflammatory condition with pulmonary and extrapulmonary manifestations. The etiology of sarcoidosis remains unknown. Iannuzzi and colleagues hypothesize that an unknown antigen sets off a cycle of chronic granulomatous inflammation in a genetically susceptible host.¹

Diagnosis

A diagnosis of sarcoidosis is typically based on a patient having an appropriate clinical presentation and a biopsy, often of lungs or skin, showing noncaseating granulomas.

Symptoms

Of the protean manifestations of sarcoidosis, respiratory symptoms are the most common and typically include subacute or chronic cough and progressive dyspnea on exertion.² Chest imaging may show only hilar or mediastinal lymphadenopathy, diffuse micronodular lung disease, or signs of chronic inflammation and fibrosis.² Upper airway involvement and progressive lung disease may lead to increased risk of sleep-disordered breathing, particularly obstructive sleep apnea (OSA).³

Sarcoidosis also can develop in the skin, neurologic system, heart, and other systems. It typically presents as areas of patchy, infiltrative inflammation. In the heart, this can lead to heart failure, often with reduced ejection fraction (EF) and ventricular arrhythmias.¹ Pulmonary hypertension (PH) may result from multiple possible mechanisms, including left-heart disease, parenchymal lung disease, sleep-disordered breathing, and possibly direct inflammation and compression of the pulmonary vasculature.^2-4

Sarcoidosis in Obese Patients

Emerging evidence shows that sarcoidosis occurs at higher rates in obese patients, suggesting that obesity may be a risk factor for the disease.^5-7 Rates of morbid obesity are increasing in the US. From 2000 to 2010, the prevalence of morbid obesity, defined as body mass index (BMI) > 40, increased by 70%, with even larger relative increases in the number of patients with BMI > 50.⁸ Among veterans who receive health care at the US Department of Veterans Affairs (VA) medical centers, 28% are obese.⁹ As a result, VA physicians will encounter more patients with morbid obesity and another significant comorbid condition.

Managing symptomatic sarcoidosis in patients with morbid obesity poses a dilemma. Typical treatment for symptomatic pulmonary sarcoidosis is prednisone 20 mg to 40 mg daily.^10,11 Higher doses are suggested for involvement of other organs, such as the heart.^2,12 Associated weight gain from corticosteroid treatment with possible sleep-disordered breathing increases an already high risk of metabolic complications in morbidly obese patients.¹³ No clear consensus exists on how corticosteroid doses should be adjusted. We present 2 cases that highlight the complexity of corticosteroid management in the obese sarcoidosis patient.

Case 1: Pulmonary Sarcoidosis

A 43-year-old morbidly obese man presented to his primary care provider with subacute onset of dyspnea. He had a history of OSA that was diagnosed empirically at another institution without polysomnogram and treated with autotitrating continuous positive airway pressure (CPAP).

The patient was admitted for expedited evaluation. His BMI was 63.2 with declining exercise tolerance and hypoxemia on ambulation. His oxyhemoglobin saturation rate was 85% after walking a short distance. Ongoing CPAP therapy for sleep-disordered breathing made laboratory evaluation for obesity hypoventilation syndrome (OHS) challenging. The patient’s serum bicarbonate test result was normal. Serum markers as well as induced sputum stains and cultures were negative for evidence of mycobacterial or fungal infections. A chest radiograph showed bilateral hilar adenopathy and miliary nodularity. Pulmonary function testing revealed severe obstruction and restriction as well as a moderate diffusion impairment. Bronchoscopy with biopsy revealed noncaseating granulomas consistent with sarcoidosis. An electrocardiogram (ECG) was normal. Transthoracic echocardiogram showed evidence of diastolic dysfunction and a mildly dilated right ventricle with normal function, suggestive of possible PH. We were unable to assess his pulmonary artery pressure.

Upon release, the patient began a course of 50 mg (0.24 mg/kg actual body weight) oral prednisone daily and home oxygen.

Six weeks after initiation of steroids, the patient reported that his dyspnea had improved. However, after 6 months of steroid treatment, his weight increased from 462 pounds to 503 pounds. He was evaluated for possible neurosarcoidosis with hypothalamic or pituitary involvement as a possible cause for the weight gain. Brain magnetic resonance imaging and hormonal testing were normal. We considered starting him on a steroid-sparing agent. However, after early efficacy, prednisone was gradually tapered and, after 1 year of treatment, discontinued. At that time, symptoms had substantially improved: His pulmonary function tests had normalized, and he was weaned off oxygen; repeat chest imaging showed only residual enlargement of the hilar lymph nodes. After cessation of steroids, the patient was able to lose 20 pounds.

Case 2: Cardiac Sarcoidosis

A 57-year-old morbidly obese man presented to the emergency department with subacute increasing dyspnea on exertion. He had a known history of sarcoidosis diagnosed by skin biopsy 28 years earlier but had been without treatment for decades. His history also included prediabetes, heart failure with preserved ejection fraction (HFpEF), OSA with an apnea hypopnea index (AHI) of 114.7 per hour, PH diagnosed by prior echocardiogram, and paroxysmal atrial fibrillation (AF). He required 2 L/m home oxygen and bilevel positive airway pressure (PAP) of 22/17 cm H₂O while sleeping.

On physical examination, the patient’s BMI was 54.6. He was tachycardic and hypoxemic on his usual oxygen flow rate. His serum bicarbonate, arterial blood pH, and PaCO₂ blood levels were normal. We heard bibasilar crackles over the lungs. Chest radiograph revealed an enlarged cardiac silhouette and bilateral infiltrates concerning for cardiogenic pulmonary edema. An echocardiogram showed a restrictive filling pattern with preserved EF and moderate dilation and dysfunction of the right ventricle, consistent with PH. A positron emission tomography (PET)/computed tomography scan, the preferred study for cardiac sarcoidosis, suggested active infiltrative septal cardiac disease and active hilar and mediastinal adenopathy. This was concerning for both cardiac and pulmonary sarcoidosis. Ongoing treatment of sleep-disordered breathing made laboratory assessment for OHS challenging. Given his intact EF, the absence of ventricular arrhythmias, and improvement with diuretics and bilevel PAP, specific treatment of sarcoidosis was not initiated. He was discharged home with a plan to re-evaluate sarcoidosis symptoms and initiate treatment as an outpatient.

The patient was readmitted 2 weeks later with worsened dyspnea, hypoxemia, and volume overload. A right heart catheterization confirmed PH with a mean pulmonary artery pressure of 44 mm Hg (68/32 mm Hg) and pulmonary vascular resistance of 4.6 Wood units. We also found evidence of left-heart dysfunction with a pulmonary capillary wedge pressure of 16 mm Hg.

Given his recurrent symptoms, evidence of active myocardial inflammation on recent PET, and prior biopsy-proven sarcoidosis, we made the decision to pursue treatment for symptomatic sarcoidosis. He began a course of 40 mg (0.20 mg/kg actual body weight) oral prednisone daily. He now required 6 L/m supplemental oxygen. After IV diuretic therapy during his hospitalization, the patient was discharged on his preadmission oral diuretic dose. Pulmonary vasodilator therapy was not initiated for PH as left heart disease and sleep-disordered breathing needed to be managed first.

One month after steroid initiation, the patient reported that the dyspnea and hypoxemia had markedly improved. His oxygen flow rate was reduced to 2 L/m. He remained normotensive and had no further difficulties with fluid retention or volume overload on a stable dose of oral diuretics. He had elevated blood glucose with a glycated hemoglobin (HbA_1c) of 6.4%. He began treatment with glipizide 5 mg daily.

After 3 months, he returned to the emergency department with hyperosmolar nonketotic hyperglycemia due to steroid-induced diabetes mellitus (DM). His HbA_1c was now 17.1%. The patient was started on a home insulin regimen, and his blood sugar values subsequently improved. He remained symptomatically better and lost 40 pounds with a guided weight management program and a stable diuretic regimen. He underwent arrhythmia evaluation with a Holter monitor that showed AF without ventricular arrhythmias.

Unfortunately, he did not return for cardiac or pulmonary reevaluation, and was lost to follow-up. Nine months after initiation of treatment, the patient died after an out-of-hospital cardiac arrest.

Discussion

These 2 cases highlight therapeutic challenges that may arise in the management of sarcoidosis with symptomatic vital organ involvement and coexistent extreme obesity. Both patients showed symptom improvement with moderate doses of prednisone (40 mg to 50 mg daily), but serious treatment-related complications developed: further weight gain in the first patient, and severe DM in the second. Although DM may have been a direct treatment complication in our second patient, his HFpEF and PH were high-risk comorbidities; he did not present with acute symptomatic worsening after treatment initiation. His symptoms were never reassessed when he was lost to follow-up.

Sarcoidosis/Obesity Relationship

Recent evidence suggests that patients with obesity are at increased risk of developing sarcoidosis.^5-7 Although the mechanism of association is unclear, several possibilities have been proposed.

Neurosarcoidosis. One known but rare cause of obesity is neurosarcoidosis of the hypothalamus or pituitary.¹⁴ This was investigated in one of our patients.

Proinflammatory responses. Another possible mechanism for the association of sarcoidosis and obesity is the proinflammatory properties of increased fat and adipose tissue.¹⁵ Obesity has been linked to an aberrant expansion of inflammatory cells and mediators, including macrophages, proinflammatory cytokines, T cells, and B cells.¹⁵ Leptin, produced primarily by adipocytes, also is higher in obese patients and has been found to be proinflammatory.¹⁶ These seem to underlie the link between obesity and other inflammatory diseases, including type 2 DM, gout, and atherosclerosis.¹⁵

Behavioral link. There also is a possible behavioral link between sarcoidosis and obesity: A patient might develop symptomatic sarcoidosis and later become less active due to dyspnea, which could predispose to weight gain.⁵

Management of Comorbid Sarcoidosis and Obesity

Regardless of the exact mechanism of this association, management of the co-occurrence of sarcoidosis and obesity poses a clinical problem, especially in cases of extreme obesity. Corticosteroids are generally considered the treatment of choice for symptomatic sarcoidosis. The initial treatment of symptomatic pulmonary sarcoidosis is 20 mg to 40 mg prednisone daily.^10,11 Higher daily doses such as 60 mg to 80 mg or 0.5 mg/kg are typically used to treat cardiac sarcoidosis, although no clear consensus exists on the appropriate dose.^12,17 One recent study showed no difference in cardiac outcomes in patients treated with high- and low-dose prednisone.¹⁸

For patients who are obese and require steroids to treat a medical condition, there is conflicting evidence on whether steroid doses should be increased in proportion to total body weight. Milsap and colleagues found clearance of prednisolone correlated strongly with degree of obesity, suggesting steroid dose should be increased in accordance with actual weight.¹⁹ In contrast, Dunn and colleagues found decreased clearance of methylprednisolone in obese patients, suggesting that ideal body weight dosing is appropriate.²⁰

Identifying the appropriate steroid dose is important because corticosteroids place obese patients at higher risk of developing complications. Treatment-related comorbidities include DM, hypertension, fluid retention, osteoporosis, and infection. Further weight gain due to steroid use is a risk for progressive OSA and, even though not generally associated with sarcoidosis alone, OHS. For patients with sarcoidosis, these complications (DM, fluid retention, hypertension, sleep-disordered breathing) may increase the risk of cardiovascular disease and PH.^21-23 Cardiomyopathy, especially with reduced EF and increased PH, can be associated with a poor prognosis in sarcoidosis.^4,24-26 PH also can be challenging to treat patients with sarcoidosis because the response of PH to steroids is unclear.²⁷ Small trials have shown the benefit of pulmonary vasodilators on hemodynamics, but these have generally been used in patients with stable sarcoidosis who do not have left-heart disease.^28-30

Our Prescription Model

We empirically prescribed moderate total doses of prednisone—although low on a mg/kg basis—to balance efficacy and the risk of adverse effects in these 2 morbidly obese patients. We also managed treatment-related complications with guided weight-management programs, CPAP, or noninvasive ventilation for sleep-disordered breathing, and DM treatment.

Our cases demonstrate the need for close monitoring of weight, blood pressure, and blood glucose to detect and treat any complications that may arise during corticosteroid treatment. Aggressive treatment of hyperglycemia with insulin or oral alternatives associated with weight loss such as metformin, sulfonylureas, dipeptidyl peptidase 4 inhibitors, or glucagon-like peptide 1 receptor agonists, may help prevent further DM complications. Sleep-disordered breathing should be assessed and treated. Bariatric surgery may be an option to treat obesity and minimize resultant complications. In our patients, and likely many others, the degree of respiratory and cardiac disease coupled with poor wound healing due to chronic prednisone, may increase the procedural risks.

Conclusion

Our experiences with these patients illustrate that symptomatic and objective improvement in sarcoidosis may be achieved in morbidly obese patients with doses of prednisone that are generally considered moderate, though quite low on a mg/kg basis.

We believe ours is the first report to describe the use of corticosteroids for the treatment of sarcoidosis in patients with morbid obesity. That 2 patients were treated at a single VA medical center within 1-year likely reflects the rising incidence of morbid obesity in the US veteran population and suggests that other federal practitioners might encounter similar patients.

Further study may show that, as an alternative to initial moderate-dose prednisone, patients with symptomatic sarcoidosis and extreme obesity might be started on antimetabolite or antitumor necrosis factor medication or on low-dose prednisone and a second steroid-sparing agent.

Sarcoidosis is a systemic inflammatory condition with pulmonary and extrapulmonary manifestations. The etiology of sarcoidosis remains unknown. Iannuzzi and colleagues hypothesize that an unknown antigen sets off a cycle of chronic granulomatous inflammation in a genetically susceptible host.¹

Diagnosis

A diagnosis of sarcoidosis is typically based on a patient having an appropriate clinical presentation and a biopsy, often of lungs or skin, showing noncaseating granulomas.

Symptoms

Of the protean manifestations of sarcoidosis, respiratory symptoms are the most common and typically include subacute or chronic cough and progressive dyspnea on exertion.² Chest imaging may show only hilar or mediastinal lymphadenopathy, diffuse micronodular lung disease, or signs of chronic inflammation and fibrosis.² Upper airway involvement and progressive lung disease may lead to increased risk of sleep-disordered breathing, particularly obstructive sleep apnea (OSA).³

Sarcoidosis also can develop in the skin, neurologic system, heart, and other systems. It typically presents as areas of patchy, infiltrative inflammation. In the heart, this can lead to heart failure, often with reduced ejection fraction (EF) and ventricular arrhythmias.¹ Pulmonary hypertension (PH) may result from multiple possible mechanisms, including left-heart disease, parenchymal lung disease, sleep-disordered breathing, and possibly direct inflammation and compression of the pulmonary vasculature.^2-4

Sarcoidosis in Obese Patients

Emerging evidence shows that sarcoidosis occurs at higher rates in obese patients, suggesting that obesity may be a risk factor for the disease.^5-7 Rates of morbid obesity are increasing in the US. From 2000 to 2010, the prevalence of morbid obesity, defined as body mass index (BMI) > 40, increased by 70%, with even larger relative increases in the number of patients with BMI > 50.⁸ Among veterans who receive health care at the US Department of Veterans Affairs (VA) medical centers, 28% are obese.⁹ As a result, VA physicians will encounter more patients with morbid obesity and another significant comorbid condition.

Managing symptomatic sarcoidosis in patients with morbid obesity poses a dilemma. Typical treatment for symptomatic pulmonary sarcoidosis is prednisone 20 mg to 40 mg daily.^10,11 Higher doses are suggested for involvement of other organs, such as the heart.^2,12 Associated weight gain from corticosteroid treatment with possible sleep-disordered breathing increases an already high risk of metabolic complications in morbidly obese patients.¹³ No clear consensus exists on how corticosteroid doses should be adjusted. We present 2 cases that highlight the complexity of corticosteroid management in the obese sarcoidosis patient.

Case 1: Pulmonary Sarcoidosis

A 43-year-old morbidly obese man presented to his primary care provider with subacute onset of dyspnea. He had a history of OSA that was diagnosed empirically at another institution without polysomnogram and treated with autotitrating continuous positive airway pressure (CPAP).

The patient was admitted for expedited evaluation. His BMI was 63.2 with declining exercise tolerance and hypoxemia on ambulation. His oxyhemoglobin saturation rate was 85% after walking a short distance. Ongoing CPAP therapy for sleep-disordered breathing made laboratory evaluation for obesity hypoventilation syndrome (OHS) challenging. The patient’s serum bicarbonate test result was normal. Serum markers as well as induced sputum stains and cultures were negative for evidence of mycobacterial or fungal infections. A chest radiograph showed bilateral hilar adenopathy and miliary nodularity. Pulmonary function testing revealed severe obstruction and restriction as well as a moderate diffusion impairment. Bronchoscopy with biopsy revealed noncaseating granulomas consistent with sarcoidosis. An electrocardiogram (ECG) was normal. Transthoracic echocardiogram showed evidence of diastolic dysfunction and a mildly dilated right ventricle with normal function, suggestive of possible PH. We were unable to assess his pulmonary artery pressure.

Upon release, the patient began a course of 50 mg (0.24 mg/kg actual body weight) oral prednisone daily and home oxygen.

Six weeks after initiation of steroids, the patient reported that his dyspnea had improved. However, after 6 months of steroid treatment, his weight increased from 462 pounds to 503 pounds. He was evaluated for possible neurosarcoidosis with hypothalamic or pituitary involvement as a possible cause for the weight gain. Brain magnetic resonance imaging and hormonal testing were normal. We considered starting him on a steroid-sparing agent. However, after early efficacy, prednisone was gradually tapered and, after 1 year of treatment, discontinued. At that time, symptoms had substantially improved: His pulmonary function tests had normalized, and he was weaned off oxygen; repeat chest imaging showed only residual enlargement of the hilar lymph nodes. After cessation of steroids, the patient was able to lose 20 pounds.

Case 2: Cardiac Sarcoidosis

A 57-year-old morbidly obese man presented to the emergency department with subacute increasing dyspnea on exertion. He had a known history of sarcoidosis diagnosed by skin biopsy 28 years earlier but had been without treatment for decades. His history also included prediabetes, heart failure with preserved ejection fraction (HFpEF), OSA with an apnea hypopnea index (AHI) of 114.7 per hour, PH diagnosed by prior echocardiogram, and paroxysmal atrial fibrillation (AF). He required 2 L/m home oxygen and bilevel positive airway pressure (PAP) of 22/17 cm H₂O while sleeping.

On physical examination, the patient’s BMI was 54.6. He was tachycardic and hypoxemic on his usual oxygen flow rate. His serum bicarbonate, arterial blood pH, and PaCO₂ blood levels were normal. We heard bibasilar crackles over the lungs. Chest radiograph revealed an enlarged cardiac silhouette and bilateral infiltrates concerning for cardiogenic pulmonary edema. An echocardiogram showed a restrictive filling pattern with preserved EF and moderate dilation and dysfunction of the right ventricle, consistent with PH. A positron emission tomography (PET)/computed tomography scan, the preferred study for cardiac sarcoidosis, suggested active infiltrative septal cardiac disease and active hilar and mediastinal adenopathy. This was concerning for both cardiac and pulmonary sarcoidosis. Ongoing treatment of sleep-disordered breathing made laboratory assessment for OHS challenging. Given his intact EF, the absence of ventricular arrhythmias, and improvement with diuretics and bilevel PAP, specific treatment of sarcoidosis was not initiated. He was discharged home with a plan to re-evaluate sarcoidosis symptoms and initiate treatment as an outpatient.

The patient was readmitted 2 weeks later with worsened dyspnea, hypoxemia, and volume overload. A right heart catheterization confirmed PH with a mean pulmonary artery pressure of 44 mm Hg (68/32 mm Hg) and pulmonary vascular resistance of 4.6 Wood units. We also found evidence of left-heart dysfunction with a pulmonary capillary wedge pressure of 16 mm Hg.

Given his recurrent symptoms, evidence of active myocardial inflammation on recent PET, and prior biopsy-proven sarcoidosis, we made the decision to pursue treatment for symptomatic sarcoidosis. He began a course of 40 mg (0.20 mg/kg actual body weight) oral prednisone daily. He now required 6 L/m supplemental oxygen. After IV diuretic therapy during his hospitalization, the patient was discharged on his preadmission oral diuretic dose. Pulmonary vasodilator therapy was not initiated for PH as left heart disease and sleep-disordered breathing needed to be managed first.

One month after steroid initiation, the patient reported that the dyspnea and hypoxemia had markedly improved. His oxygen flow rate was reduced to 2 L/m. He remained normotensive and had no further difficulties with fluid retention or volume overload on a stable dose of oral diuretics. He had elevated blood glucose with a glycated hemoglobin (HbA_1c) of 6.4%. He began treatment with glipizide 5 mg daily.

After 3 months, he returned to the emergency department with hyperosmolar nonketotic hyperglycemia due to steroid-induced diabetes mellitus (DM). His HbA_1c was now 17.1%. The patient was started on a home insulin regimen, and his blood sugar values subsequently improved. He remained symptomatically better and lost 40 pounds with a guided weight management program and a stable diuretic regimen. He underwent arrhythmia evaluation with a Holter monitor that showed AF without ventricular arrhythmias.

Unfortunately, he did not return for cardiac or pulmonary reevaluation, and was lost to follow-up. Nine months after initiation of treatment, the patient died after an out-of-hospital cardiac arrest.

Discussion

These 2 cases highlight therapeutic challenges that may arise in the management of sarcoidosis with symptomatic vital organ involvement and coexistent extreme obesity. Both patients showed symptom improvement with moderate doses of prednisone (40 mg to 50 mg daily), but serious treatment-related complications developed: further weight gain in the first patient, and severe DM in the second. Although DM may have been a direct treatment complication in our second patient, his HFpEF and PH were high-risk comorbidities; he did not present with acute symptomatic worsening after treatment initiation. His symptoms were never reassessed when he was lost to follow-up.

Sarcoidosis/Obesity Relationship

Recent evidence suggests that patients with obesity are at increased risk of developing sarcoidosis.^5-7 Although the mechanism of association is unclear, several possibilities have been proposed.

Neurosarcoidosis. One known but rare cause of obesity is neurosarcoidosis of the hypothalamus or pituitary.¹⁴ This was investigated in one of our patients.

Proinflammatory responses. Another possible mechanism for the association of sarcoidosis and obesity is the proinflammatory properties of increased fat and adipose tissue.¹⁵ Obesity has been linked to an aberrant expansion of inflammatory cells and mediators, including macrophages, proinflammatory cytokines, T cells, and B cells.¹⁵ Leptin, produced primarily by adipocytes, also is higher in obese patients and has been found to be proinflammatory.¹⁶ These seem to underlie the link between obesity and other inflammatory diseases, including type 2 DM, gout, and atherosclerosis.¹⁵

Behavioral link. There also is a possible behavioral link between sarcoidosis and obesity: A patient might develop symptomatic sarcoidosis and later become less active due to dyspnea, which could predispose to weight gain.⁵

Management of Comorbid Sarcoidosis and Obesity

Regardless of the exact mechanism of this association, management of the co-occurrence of sarcoidosis and obesity poses a clinical problem, especially in cases of extreme obesity. Corticosteroids are generally considered the treatment of choice for symptomatic sarcoidosis. The initial treatment of symptomatic pulmonary sarcoidosis is 20 mg to 40 mg prednisone daily.^10,11 Higher daily doses such as 60 mg to 80 mg or 0.5 mg/kg are typically used to treat cardiac sarcoidosis, although no clear consensus exists on the appropriate dose.^12,17 One recent study showed no difference in cardiac outcomes in patients treated with high- and low-dose prednisone.¹⁸

For patients who are obese and require steroids to treat a medical condition, there is conflicting evidence on whether steroid doses should be increased in proportion to total body weight. Milsap and colleagues found clearance of prednisolone correlated strongly with degree of obesity, suggesting steroid dose should be increased in accordance with actual weight.¹⁹ In contrast, Dunn and colleagues found decreased clearance of methylprednisolone in obese patients, suggesting that ideal body weight dosing is appropriate.²⁰

Identifying the appropriate steroid dose is important because corticosteroids place obese patients at higher risk of developing complications. Treatment-related comorbidities include DM, hypertension, fluid retention, osteoporosis, and infection. Further weight gain due to steroid use is a risk for progressive OSA and, even though not generally associated with sarcoidosis alone, OHS. For patients with sarcoidosis, these complications (DM, fluid retention, hypertension, sleep-disordered breathing) may increase the risk of cardiovascular disease and PH.^21-23 Cardiomyopathy, especially with reduced EF and increased PH, can be associated with a poor prognosis in sarcoidosis.^4,24-26 PH also can be challenging to treat patients with sarcoidosis because the response of PH to steroids is unclear.²⁷ Small trials have shown the benefit of pulmonary vasodilators on hemodynamics, but these have generally been used in patients with stable sarcoidosis who do not have left-heart disease.^28-30

Our Prescription Model

We empirically prescribed moderate total doses of prednisone—although low on a mg/kg basis—to balance efficacy and the risk of adverse effects in these 2 morbidly obese patients. We also managed treatment-related complications with guided weight-management programs, CPAP, or noninvasive ventilation for sleep-disordered breathing, and DM treatment.

Our cases demonstrate the need for close monitoring of weight, blood pressure, and blood glucose to detect and treat any complications that may arise during corticosteroid treatment. Aggressive treatment of hyperglycemia with insulin or oral alternatives associated with weight loss such as metformin, sulfonylureas, dipeptidyl peptidase 4 inhibitors, or glucagon-like peptide 1 receptor agonists, may help prevent further DM complications. Sleep-disordered breathing should be assessed and treated. Bariatric surgery may be an option to treat obesity and minimize resultant complications. In our patients, and likely many others, the degree of respiratory and cardiac disease coupled with poor wound healing due to chronic prednisone, may increase the procedural risks.

Conclusion

Our experiences with these patients illustrate that symptomatic and objective improvement in sarcoidosis may be achieved in morbidly obese patients with doses of prednisone that are generally considered moderate, though quite low on a mg/kg basis.

We believe ours is the first report to describe the use of corticosteroids for the treatment of sarcoidosis in patients with morbid obesity. That 2 patients were treated at a single VA medical center within 1-year likely reflects the rising incidence of morbid obesity in the US veteran population and suggests that other federal practitioners might encounter similar patients.

Further study may show that, as an alternative to initial moderate-dose prednisone, patients with symptomatic sarcoidosis and extreme obesity might be started on antimetabolite or antitumor necrosis factor medication or on low-dose prednisone and a second steroid-sparing agent.

Sarcoidosis is a systemic inflammatory condition with pulmonary and extrapulmonary manifestations. The etiology of sarcoidosis remains unknown. Iannuzzi and colleagues hypothesize that an unknown antigen sets off a cycle of chronic granulomatous inflammation in a genetically susceptible host.¹

Diagnosis

A diagnosis of sarcoidosis is typically based on a patient having an appropriate clinical presentation and a biopsy, often of lungs or skin, showing noncaseating granulomas.

Symptoms

Of the protean manifestations of sarcoidosis, respiratory symptoms are the most common and typically include subacute or chronic cough and progressive dyspnea on exertion.² Chest imaging may show only hilar or mediastinal lymphadenopathy, diffuse micronodular lung disease, or signs of chronic inflammation and fibrosis.² Upper airway involvement and progressive lung disease may lead to increased risk of sleep-disordered breathing, particularly obstructive sleep apnea (OSA).³

Sarcoidosis also can develop in the skin, neurologic system, heart, and other systems. It typically presents as areas of patchy, infiltrative inflammation. In the heart, this can lead to heart failure, often with reduced ejection fraction (EF) and ventricular arrhythmias.¹ Pulmonary hypertension (PH) may result from multiple possible mechanisms, including left-heart disease, parenchymal lung disease, sleep-disordered breathing, and possibly direct inflammation and compression of the pulmonary vasculature.^2-4

Sarcoidosis in Obese Patients

Emerging evidence shows that sarcoidosis occurs at higher rates in obese patients, suggesting that obesity may be a risk factor for the disease.^5-7 Rates of morbid obesity are increasing in the US. From 2000 to 2010, the prevalence of morbid obesity, defined as body mass index (BMI) > 40, increased by 70%, with even larger relative increases in the number of patients with BMI > 50.⁸ Among veterans who receive health care at the US Department of Veterans Affairs (VA) medical centers, 28% are obese.⁹ As a result, VA physicians will encounter more patients with morbid obesity and another significant comorbid condition.

Managing symptomatic sarcoidosis in patients with morbid obesity poses a dilemma. Typical treatment for symptomatic pulmonary sarcoidosis is prednisone 20 mg to 40 mg daily.^10,11 Higher doses are suggested for involvement of other organs, such as the heart.^2,12 Associated weight gain from corticosteroid treatment with possible sleep-disordered breathing increases an already high risk of metabolic complications in morbidly obese patients.¹³ No clear consensus exists on how corticosteroid doses should be adjusted. We present 2 cases that highlight the complexity of corticosteroid management in the obese sarcoidosis patient.

Case 1: Pulmonary Sarcoidosis

A 43-year-old morbidly obese man presented to his primary care provider with subacute onset of dyspnea. He had a history of OSA that was diagnosed empirically at another institution without polysomnogram and treated with autotitrating continuous positive airway pressure (CPAP).

The patient was admitted for expedited evaluation. His BMI was 63.2 with declining exercise tolerance and hypoxemia on ambulation. His oxyhemoglobin saturation rate was 85% after walking a short distance. Ongoing CPAP therapy for sleep-disordered breathing made laboratory evaluation for obesity hypoventilation syndrome (OHS) challenging. The patient’s serum bicarbonate test result was normal. Serum markers as well as induced sputum stains and cultures were negative for evidence of mycobacterial or fungal infections. A chest radiograph showed bilateral hilar adenopathy and miliary nodularity. Pulmonary function testing revealed severe obstruction and restriction as well as a moderate diffusion impairment. Bronchoscopy with biopsy revealed noncaseating granulomas consistent with sarcoidosis. An electrocardiogram (ECG) was normal. Transthoracic echocardiogram showed evidence of diastolic dysfunction and a mildly dilated right ventricle with normal function, suggestive of possible PH. We were unable to assess his pulmonary artery pressure.

Upon release, the patient began a course of 50 mg (0.24 mg/kg actual body weight) oral prednisone daily and home oxygen.

Six weeks after initiation of steroids, the patient reported that his dyspnea had improved. However, after 6 months of steroid treatment, his weight increased from 462 pounds to 503 pounds. He was evaluated for possible neurosarcoidosis with hypothalamic or pituitary involvement as a possible cause for the weight gain. Brain magnetic resonance imaging and hormonal testing were normal. We considered starting him on a steroid-sparing agent. However, after early efficacy, prednisone was gradually tapered and, after 1 year of treatment, discontinued. At that time, symptoms had substantially improved: His pulmonary function tests had normalized, and he was weaned off oxygen; repeat chest imaging showed only residual enlargement of the hilar lymph nodes. After cessation of steroids, the patient was able to lose 20 pounds.

Case 2: Cardiac Sarcoidosis

A 57-year-old morbidly obese man presented to the emergency department with subacute increasing dyspnea on exertion. He had a known history of sarcoidosis diagnosed by skin biopsy 28 years earlier but had been without treatment for decades. His history also included prediabetes, heart failure with preserved ejection fraction (HFpEF), OSA with an apnea hypopnea index (AHI) of 114.7 per hour, PH diagnosed by prior echocardiogram, and paroxysmal atrial fibrillation (AF). He required 2 L/m home oxygen and bilevel positive airway pressure (PAP) of 22/17 cm H₂O while sleeping.

On physical examination, the patient’s BMI was 54.6. He was tachycardic and hypoxemic on his usual oxygen flow rate. His serum bicarbonate, arterial blood pH, and PaCO₂ blood levels were normal. We heard bibasilar crackles over the lungs. Chest radiograph revealed an enlarged cardiac silhouette and bilateral infiltrates concerning for cardiogenic pulmonary edema. An echocardiogram showed a restrictive filling pattern with preserved EF and moderate dilation and dysfunction of the right ventricle, consistent with PH. A positron emission tomography (PET)/computed tomography scan, the preferred study for cardiac sarcoidosis, suggested active infiltrative septal cardiac disease and active hilar and mediastinal adenopathy. This was concerning for both cardiac and pulmonary sarcoidosis. Ongoing treatment of sleep-disordered breathing made laboratory assessment for OHS challenging. Given his intact EF, the absence of ventricular arrhythmias, and improvement with diuretics and bilevel PAP, specific treatment of sarcoidosis was not initiated. He was discharged home with a plan to re-evaluate sarcoidosis symptoms and initiate treatment as an outpatient.

The patient was readmitted 2 weeks later with worsened dyspnea, hypoxemia, and volume overload. A right heart catheterization confirmed PH with a mean pulmonary artery pressure of 44 mm Hg (68/32 mm Hg) and pulmonary vascular resistance of 4.6 Wood units. We also found evidence of left-heart dysfunction with a pulmonary capillary wedge pressure of 16 mm Hg.

Given his recurrent symptoms, evidence of active myocardial inflammation on recent PET, and prior biopsy-proven sarcoidosis, we made the decision to pursue treatment for symptomatic sarcoidosis. He began a course of 40 mg (0.20 mg/kg actual body weight) oral prednisone daily. He now required 6 L/m supplemental oxygen. After IV diuretic therapy during his hospitalization, the patient was discharged on his preadmission oral diuretic dose. Pulmonary vasodilator therapy was not initiated for PH as left heart disease and sleep-disordered breathing needed to be managed first.

One month after steroid initiation, the patient reported that the dyspnea and hypoxemia had markedly improved. His oxygen flow rate was reduced to 2 L/m. He remained normotensive and had no further difficulties with fluid retention or volume overload on a stable dose of oral diuretics. He had elevated blood glucose with a glycated hemoglobin (HbA_1c) of 6.4%. He began treatment with glipizide 5 mg daily.

After 3 months, he returned to the emergency department with hyperosmolar nonketotic hyperglycemia due to steroid-induced diabetes mellitus (DM). His HbA_1c was now 17.1%. The patient was started on a home insulin regimen, and his blood sugar values subsequently improved. He remained symptomatically better and lost 40 pounds with a guided weight management program and a stable diuretic regimen. He underwent arrhythmia evaluation with a Holter monitor that showed AF without ventricular arrhythmias.

Unfortunately, he did not return for cardiac or pulmonary reevaluation, and was lost to follow-up. Nine months after initiation of treatment, the patient died after an out-of-hospital cardiac arrest.

Discussion

These 2 cases highlight therapeutic challenges that may arise in the management of sarcoidosis with symptomatic vital organ involvement and coexistent extreme obesity. Both patients showed symptom improvement with moderate doses of prednisone (40 mg to 50 mg daily), but serious treatment-related complications developed: further weight gain in the first patient, and severe DM in the second. Although DM may have been a direct treatment complication in our second patient, his HFpEF and PH were high-risk comorbidities; he did not present with acute symptomatic worsening after treatment initiation. His symptoms were never reassessed when he was lost to follow-up.

Sarcoidosis/Obesity Relationship

Recent evidence suggests that patients with obesity are at increased risk of developing sarcoidosis.^5-7 Although the mechanism of association is unclear, several possibilities have been proposed.

Neurosarcoidosis. One known but rare cause of obesity is neurosarcoidosis of the hypothalamus or pituitary.¹⁴ This was investigated in one of our patients.

Proinflammatory responses. Another possible mechanism for the association of sarcoidosis and obesity is the proinflammatory properties of increased fat and adipose tissue.¹⁵ Obesity has been linked to an aberrant expansion of inflammatory cells and mediators, including macrophages, proinflammatory cytokines, T cells, and B cells.¹⁵ Leptin, produced primarily by adipocytes, also is higher in obese patients and has been found to be proinflammatory.¹⁶ These seem to underlie the link between obesity and other inflammatory diseases, including type 2 DM, gout, and atherosclerosis.¹⁵

Behavioral link. There also is a possible behavioral link between sarcoidosis and obesity: A patient might develop symptomatic sarcoidosis and later become less active due to dyspnea, which could predispose to weight gain.⁵

Management of Comorbid Sarcoidosis and Obesity

Regardless of the exact mechanism of this association, management of the co-occurrence of sarcoidosis and obesity poses a clinical problem, especially in cases of extreme obesity. Corticosteroids are generally considered the treatment of choice for symptomatic sarcoidosis. The initial treatment of symptomatic pulmonary sarcoidosis is 20 mg to 40 mg prednisone daily.^10,11 Higher daily doses such as 60 mg to 80 mg or 0.5 mg/kg are typically used to treat cardiac sarcoidosis, although no clear consensus exists on the appropriate dose.^12,17 One recent study showed no difference in cardiac outcomes in patients treated with high- and low-dose prednisone.¹⁸

For patients who are obese and require steroids to treat a medical condition, there is conflicting evidence on whether steroid doses should be increased in proportion to total body weight. Milsap and colleagues found clearance of prednisolone correlated strongly with degree of obesity, suggesting steroid dose should be increased in accordance with actual weight.¹⁹ In contrast, Dunn and colleagues found decreased clearance of methylprednisolone in obese patients, suggesting that ideal body weight dosing is appropriate.²⁰

Identifying the appropriate steroid dose is important because corticosteroids place obese patients at higher risk of developing complications. Treatment-related comorbidities include DM, hypertension, fluid retention, osteoporosis, and infection. Further weight gain due to steroid use is a risk for progressive OSA and, even though not generally associated with sarcoidosis alone, OHS. For patients with sarcoidosis, these complications (DM, fluid retention, hypertension, sleep-disordered breathing) may increase the risk of cardiovascular disease and PH.^21-23 Cardiomyopathy, especially with reduced EF and increased PH, can be associated with a poor prognosis in sarcoidosis.^4,24-26 PH also can be challenging to treat patients with sarcoidosis because the response of PH to steroids is unclear.²⁷ Small trials have shown the benefit of pulmonary vasodilators on hemodynamics, but these have generally been used in patients with stable sarcoidosis who do not have left-heart disease.^28-30

Our Prescription Model

We empirically prescribed moderate total doses of prednisone—although low on a mg/kg basis—to balance efficacy and the risk of adverse effects in these 2 morbidly obese patients. We also managed treatment-related complications with guided weight-management programs, CPAP, or noninvasive ventilation for sleep-disordered breathing, and DM treatment.

Our cases demonstrate the need for close monitoring of weight, blood pressure, and blood glucose to detect and treat any complications that may arise during corticosteroid treatment. Aggressive treatment of hyperglycemia with insulin or oral alternatives associated with weight loss such as metformin, sulfonylureas, dipeptidyl peptidase 4 inhibitors, or glucagon-like peptide 1 receptor agonists, may help prevent further DM complications. Sleep-disordered breathing should be assessed and treated. Bariatric surgery may be an option to treat obesity and minimize resultant complications. In our patients, and likely many others, the degree of respiratory and cardiac disease coupled with poor wound healing due to chronic prednisone, may increase the procedural risks.

Conclusion

Our experiences with these patients illustrate that symptomatic and objective improvement in sarcoidosis may be achieved in morbidly obese patients with doses of prednisone that are generally considered moderate, though quite low on a mg/kg basis.

We believe ours is the first report to describe the use of corticosteroids for the treatment of sarcoidosis in patients with morbid obesity. That 2 patients were treated at a single VA medical center within 1-year likely reflects the rising incidence of morbid obesity in the US veteran population and suggests that other federal practitioners might encounter similar patients.

Further study may show that, as an alternative to initial moderate-dose prednisone, patients with symptomatic sarcoidosis and extreme obesity might be started on antimetabolite or antitumor necrosis factor medication or on low-dose prednisone and a second steroid-sparing agent.

Case Report

A 26-year-old woman with a medical history of newly diagnosed diabetes mellitus (DM), obesity, and asthma was evaluated as a hospital consultation with a vegetative plaque on the left lateral ankle of 13 months’ duration. The lesion first appeared as a red scaly rash that became purulent. The lesion had been treated with multiple rounds of topical antibiotics, oral antibiotics, topical antifungals, and corticosteroids without resolution. The patient denied pain or any decrease in ankle mobility. Review of systems was otherwise negative.

On physical examination, 3 large, pink, scaly, crusted plaques with surrounding erythema were observed (Figure 1A). On palpation, purulent drainage with a foul odor was noted in the area underlying the lesion. Initial punch biopsy demonstrated epidermal hyperplasia with neutrophil-rich sinus tracts consistent with pyoderma vegetans (PV)(Figure 2A). Tissue culture was positive for Staphylococcus aureus and Streptococcus anginosus. Cultures for both fungi and acid-fast bacilli were negative for growth.

The patient was treated with mupirocin ointment 2% and 3 months of cephalexin 250 mg twice daily, which cleared the purulent crust; however, serous drainage, ulceration, and erythema persisted. The patient needed an extended course of antibiotics, which had not been previously administered to clear the purulence. During this treatment regimen, the patient’s DM remained uncontrolled.

A second deeper punch biopsy revealed a layered granulomatous infiltrate with sclerosis throughout the dermis most consistent with necrobiosis lipoidica (NL)(Figure 2B). Direct immunofluorescence biopsy was negative. Once the PV was clear, betamethasone dipropionate ointment 0.05% was initiated to address the residual lesions (Figure 1B).

Physical examination combined with histopathologic findings and staphylococcal- and streptococcal-positive tissue cultures supported a diagnosis of NL with superimposed PV.

Comment

Necrobiosis lipoidica is a chronic granulomatous disease characterized by collagen degeneration, granulomatous formation, and endothelial wall thickening.¹ The condition is most commonly seen in association with insulin-dependent DM, though it also has been described in other inflammatory conditions. A case of NL in monozygotic twins has been reported, suggesting a genetic component in nondiabetic patients with NL.² Necrobiosis lipoidica affects females more often than males.

The pathogenesis of NL is not well understood but likely involves secondary microangiopathy because of glycoprotein deposition in vessel walls, leading to vascular thickening. Histopathology reveals palisading and necrobiotic granulomas comprising large confluent areas of necrobiosis throughout the dermis, giving a layered appearance.³

Clinically, NL presents with asymptomatic, well-circumscribed, violaceous papules and nodules that coalesce into plaques on the lower extremities, face, or trunk. The plaques have a central red-brown hue that progressively becomes more yellow and atrophic. The lesions can become eroded and ulcerated if left untreated.¹

Clinical diagnosis of NL can be challenging due to the similar clinical findings of other granulomatous lesions, such as granuloma annulare and cutaneous sarcoidosis. As reported by Pellicano and colleagues,⁴ dermoscopy has proved to be an excellent tool for differentiating these granulomatous skin lesions. Necrobiosis lipoidica demonstrates elongated serpentine telangiectases overlying a white structureless background, whereas granuloma annulare reveals orange-red structureless peripheral borders.⁵

Treatment of NL is difficult; patients often are refractory. Tight control of blood glucose alone has not been proven to cure NL. The mainstay of treatment is topical and intralesional corticosteroids at the active borders of the lesions. Tumor necrosis factor α inhibitors have shown some success, though recurrence has been reported.⁶ Other treatments, such as topical tretinoin and topical tacrolimus, may be of some benefit for atrophic NL lesions. Studies also have shown that skin grafting can be of surgical benefit in ulcerative NL with a low rate of recurrence.⁶ Control and management of DM plus lifestyle modifications may play a role in decreasing the severity of NL.⁷ Topical psoralen plus UVA light therapy and other experimental treatments, such as antiplatelet medications,⁸ also have been utilized.

The case of NL presented here was complicated by a superimposed suppurative infection consistent with PV, a rare chronic bacterial infection of the skin that presents with vegetative plaques. Pyoderma vegetans is most commonly observed in patients with underlying immunosuppression, likely secondary to DM in this case. Pyoderma vegetans is most often caused by S aureus and β-hemolytic streptococci. The clinical presentation of PV reveals verrucous vegetative plaques with pustules and abscesses. The borders of the lesions may be elevated and have a granulomatous appearance, thus complicating clinical diagnosis. There often is foul-smelling, purulent discharge within the plaques.⁹

Histopathology reveals pseudoepitheliomatous hyperplasia with abscesses and sinus tracts. An acute or chronic granulomatous inflammatory infiltrate may be observed. Basophilic fungus like granules are not seen within specimens of PV, which helps differentiate the disease from botryomycosis.¹⁰

There is no standardized treatment of PV; topical and systemic antibiotics are mainstays.¹⁰ One reported case of PV responded well to acitretin.⁹ Our patient responded well to 3 months of oral antibiotic therapy, followed by topical corticosteroids.

Case Report

A 26-year-old woman with a medical history of newly diagnosed diabetes mellitus (DM), obesity, and asthma was evaluated as a hospital consultation with a vegetative plaque on the left lateral ankle of 13 months’ duration. The lesion first appeared as a red scaly rash that became purulent. The lesion had been treated with multiple rounds of topical antibiotics, oral antibiotics, topical antifungals, and corticosteroids without resolution. The patient denied pain or any decrease in ankle mobility. Review of systems was otherwise negative.

On physical examination, 3 large, pink, scaly, crusted plaques with surrounding erythema were observed (Figure 1A). On palpation, purulent drainage with a foul odor was noted in the area underlying the lesion. Initial punch biopsy demonstrated epidermal hyperplasia with neutrophil-rich sinus tracts consistent with pyoderma vegetans (PV)(Figure 2A). Tissue culture was positive for Staphylococcus aureus and Streptococcus anginosus. Cultures for both fungi and acid-fast bacilli were negative for growth.

The patient was treated with mupirocin ointment 2% and 3 months of cephalexin 250 mg twice daily, which cleared the purulent crust; however, serous drainage, ulceration, and erythema persisted. The patient needed an extended course of antibiotics, which had not been previously administered to clear the purulence. During this treatment regimen, the patient’s DM remained uncontrolled.

A second deeper punch biopsy revealed a layered granulomatous infiltrate with sclerosis throughout the dermis most consistent with necrobiosis lipoidica (NL)(Figure 2B). Direct immunofluorescence biopsy was negative. Once the PV was clear, betamethasone dipropionate ointment 0.05% was initiated to address the residual lesions (Figure 1B).

Physical examination combined with histopathologic findings and staphylococcal- and streptococcal-positive tissue cultures supported a diagnosis of NL with superimposed PV.

Comment

Necrobiosis lipoidica is a chronic granulomatous disease characterized by collagen degeneration, granulomatous formation, and endothelial wall thickening.¹ The condition is most commonly seen in association with insulin-dependent DM, though it also has been described in other inflammatory conditions. A case of NL in monozygotic twins has been reported, suggesting a genetic component in nondiabetic patients with NL.² Necrobiosis lipoidica affects females more often than males.

The pathogenesis of NL is not well understood but likely involves secondary microangiopathy because of glycoprotein deposition in vessel walls, leading to vascular thickening. Histopathology reveals palisading and necrobiotic granulomas comprising large confluent areas of necrobiosis throughout the dermis, giving a layered appearance.³

Clinically, NL presents with asymptomatic, well-circumscribed, violaceous papules and nodules that coalesce into plaques on the lower extremities, face, or trunk. The plaques have a central red-brown hue that progressively becomes more yellow and atrophic. The lesions can become eroded and ulcerated if left untreated.¹

Clinical diagnosis of NL can be challenging due to the similar clinical findings of other granulomatous lesions, such as granuloma annulare and cutaneous sarcoidosis. As reported by Pellicano and colleagues,⁴ dermoscopy has proved to be an excellent tool for differentiating these granulomatous skin lesions. Necrobiosis lipoidica demonstrates elongated serpentine telangiectases overlying a white structureless background, whereas granuloma annulare reveals orange-red structureless peripheral borders.⁵

Treatment of NL is difficult; patients often are refractory. Tight control of blood glucose alone has not been proven to cure NL. The mainstay of treatment is topical and intralesional corticosteroids at the active borders of the lesions. Tumor necrosis factor α inhibitors have shown some success, though recurrence has been reported.⁶ Other treatments, such as topical tretinoin and topical tacrolimus, may be of some benefit for atrophic NL lesions. Studies also have shown that skin grafting can be of surgical benefit in ulcerative NL with a low rate of recurrence.⁶ Control and management of DM plus lifestyle modifications may play a role in decreasing the severity of NL.⁷ Topical psoralen plus UVA light therapy and other experimental treatments, such as antiplatelet medications,⁸ also have been utilized.

The case of NL presented here was complicated by a superimposed suppurative infection consistent with PV, a rare chronic bacterial infection of the skin that presents with vegetative plaques. Pyoderma vegetans is most commonly observed in patients with underlying immunosuppression, likely secondary to DM in this case. Pyoderma vegetans is most often caused by S aureus and β-hemolytic streptococci. The clinical presentation of PV reveals verrucous vegetative plaques with pustules and abscesses. The borders of the lesions may be elevated and have a granulomatous appearance, thus complicating clinical diagnosis. There often is foul-smelling, purulent discharge within the plaques.⁹

Histopathology reveals pseudoepitheliomatous hyperplasia with abscesses and sinus tracts. An acute or chronic granulomatous inflammatory infiltrate may be observed. Basophilic fungus like granules are not seen within specimens of PV, which helps differentiate the disease from botryomycosis.¹⁰

There is no standardized treatment of PV; topical and systemic antibiotics are mainstays.¹⁰ One reported case of PV responded well to acitretin.⁹ Our patient responded well to 3 months of oral antibiotic therapy, followed by topical corticosteroids.

Case Report

A 26-year-old woman with a medical history of newly diagnosed diabetes mellitus (DM), obesity, and asthma was evaluated as a hospital consultation with a vegetative plaque on the left lateral ankle of 13 months’ duration. The lesion first appeared as a red scaly rash that became purulent. The lesion had been treated with multiple rounds of topical antibiotics, oral antibiotics, topical antifungals, and corticosteroids without resolution. The patient denied pain or any decrease in ankle mobility. Review of systems was otherwise negative.

On physical examination, 3 large, pink, scaly, crusted plaques with surrounding erythema were observed (Figure 1A). On palpation, purulent drainage with a foul odor was noted in the area underlying the lesion. Initial punch biopsy demonstrated epidermal hyperplasia with neutrophil-rich sinus tracts consistent with pyoderma vegetans (PV)(Figure 2A). Tissue culture was positive for Staphylococcus aureus and Streptococcus anginosus. Cultures for both fungi and acid-fast bacilli were negative for growth.

The patient was treated with mupirocin ointment 2% and 3 months of cephalexin 250 mg twice daily, which cleared the purulent crust; however, serous drainage, ulceration, and erythema persisted. The patient needed an extended course of antibiotics, which had not been previously administered to clear the purulence. During this treatment regimen, the patient’s DM remained uncontrolled.

A second deeper punch biopsy revealed a layered granulomatous infiltrate with sclerosis throughout the dermis most consistent with necrobiosis lipoidica (NL)(Figure 2B). Direct immunofluorescence biopsy was negative. Once the PV was clear, betamethasone dipropionate ointment 0.05% was initiated to address the residual lesions (Figure 1B).

Physical examination combined with histopathologic findings and staphylococcal- and streptococcal-positive tissue cultures supported a diagnosis of NL with superimposed PV.

Comment

Necrobiosis lipoidica is a chronic granulomatous disease characterized by collagen degeneration, granulomatous formation, and endothelial wall thickening.¹ The condition is most commonly seen in association with insulin-dependent DM, though it also has been described in other inflammatory conditions. A case of NL in monozygotic twins has been reported, suggesting a genetic component in nondiabetic patients with NL.² Necrobiosis lipoidica affects females more often than males.

The pathogenesis of NL is not well understood but likely involves secondary microangiopathy because of glycoprotein deposition in vessel walls, leading to vascular thickening. Histopathology reveals palisading and necrobiotic granulomas comprising large confluent areas of necrobiosis throughout the dermis, giving a layered appearance.³

Clinically, NL presents with asymptomatic, well-circumscribed, violaceous papules and nodules that coalesce into plaques on the lower extremities, face, or trunk. The plaques have a central red-brown hue that progressively becomes more yellow and atrophic. The lesions can become eroded and ulcerated if left untreated.¹

Clinical diagnosis of NL can be challenging due to the similar clinical findings of other granulomatous lesions, such as granuloma annulare and cutaneous sarcoidosis. As reported by Pellicano and colleagues,⁴ dermoscopy has proved to be an excellent tool for differentiating these granulomatous skin lesions. Necrobiosis lipoidica demonstrates elongated serpentine telangiectases overlying a white structureless background, whereas granuloma annulare reveals orange-red structureless peripheral borders.⁵

Treatment of NL is difficult; patients often are refractory. Tight control of blood glucose alone has not been proven to cure NL. The mainstay of treatment is topical and intralesional corticosteroids at the active borders of the lesions. Tumor necrosis factor α inhibitors have shown some success, though recurrence has been reported.⁶ Other treatments, such as topical tretinoin and topical tacrolimus, may be of some benefit for atrophic NL lesions. Studies also have shown that skin grafting can be of surgical benefit in ulcerative NL with a low rate of recurrence.⁶ Control and management of DM plus lifestyle modifications may play a role in decreasing the severity of NL.⁷ Topical psoralen plus UVA light therapy and other experimental treatments, such as antiplatelet medications,⁸ also have been utilized.

The case of NL presented here was complicated by a superimposed suppurative infection consistent with PV, a rare chronic bacterial infection of the skin that presents with vegetative plaques. Pyoderma vegetans is most commonly observed in patients with underlying immunosuppression, likely secondary to DM in this case. Pyoderma vegetans is most often caused by S aureus and β-hemolytic streptococci. The clinical presentation of PV reveals verrucous vegetative plaques with pustules and abscesses. The borders of the lesions may be elevated and have a granulomatous appearance, thus complicating clinical diagnosis. There often is foul-smelling, purulent discharge within the plaques.⁹

Histopathology reveals pseudoepitheliomatous hyperplasia with abscesses and sinus tracts. An acute or chronic granulomatous inflammatory infiltrate may be observed. Basophilic fungus like granules are not seen within specimens of PV, which helps differentiate the disease from botryomycosis.¹⁰

There is no standardized treatment of PV; topical and systemic antibiotics are mainstays.¹⁰ One reported case of PV responded well to acitretin.⁹ Our patient responded well to 3 months of oral antibiotic therapy, followed by topical corticosteroids.

Idiopathic granulomatous mastitis (IGM) is rare during pregnancy; it typically is seen in women of childbearing potential from 6 months to 6 years postpartum.¹ Because of a temporal association with breastfeeding, it is believed that hyperprolactinemia² or an immune response to local lobular secretions might play a role in pathogenesis. Early misdiagnosis as bacterial mastitis is common, prompting multiple antibiotic regimens. When antibiotics fail, patients are worked up for inflammatory breast cancer, given the nonhealing breast nodules. Mammography, ultrasonography, and fine-needle aspiration often are unable to rule out carcinoma, warranting excisional biopsies of nodules. The patient is then referred to rheumatology for potential sarcoidosis or to dermatology for IGM. In either case, the workup should be similar, but additional history focused on behavior and medications is essential in suspected IGM, given the association with hyperprolactinemia.

Because IGM is rare, there are no randomized, placebo-controlled trials of treatment efficacy. In many cases, patients undergo complete mastectomy, which is curative but may be psychologically and physically impactful in young women. In some cases, high-dose corticosteroids have been successful; however, because the IGM process can last longer than 2 years, patients treated in this manner are exposed to steroid morbidities.¹

We report 3 cases of IGM that add to the literature on possible contributing factors, clinical presentations, and treatments for this disease. We also demonstrate that appropriate trigger identification and steroid-sparing agents, specifically methotrexate, can be breast-saving as they can alleviate this debilitating condition, obviating the need for radical surgical intervention.

CASE REPORTS 

Patient 1

A 40-year-old woman with a 4-year history of breastfeeding noted a grape-sized nodule on the left breast that grew to the size of a grapefruit after 2 weeks. Ulceration and drainage periodically occurred, forming pink plaques along the lateral aspects of the breast after healing. Her primary care provider suspected infectious mastitis; she was given an oral antibiotic (cephalexin) and intravenous antibiotics without improvement.

Imaging
Subsequent magnetic resonance imaging revealed a large, irregular, enhancing mass within the outer left breast (6.5 cm at greatest dimension) with additional surrounding amorphous enhancement highly suspicious for malignancy. There also were multiple prominent left axillary lymph nodes, with the largest demonstrating a cortical thickness of 8 mm.

Biopsy
Core breast biopsy showed benign tissue with fat necrosis. Fine-needle aspiration revealed few benign ductal cells and rare histiocytes; because these findings were nondiagnostic and cancer was still a consideration, the patient underwent excisional biopsy.

Histologic sections of breast tissue showed extensive lobulocentric inflammation comprising histiocytes and lymphocytes, with neutrophils admixed and forming microabscesses (Figure 1A). Multinucleated giant cells and single-cell necrosis were seen, but true caseous necrosis was absent (Figure 1B). Duct spaces often contained inflammatory cells or secretions. Special stains for fungal and acid-fast bacterial microorganisms were negative.

Granulomatous lobular mastitis was diagnosed, and the patient was referred to dermatology. On presentation to dermatology, the left breast showed a 6-cm area of firm induration and overlying peau d’orange change to the epidermis (Figure 2A). Based on pathologic analysis, she was worked up for a possible granulomatous etiology. Negative purified protein derivative (tuberculin)(PPD) and a normal chest radiograph ruled out tuberculosis. Normal chest radiography, serum Ca²⁺ and angiotensin-converting enzyme (ACE) levels, and ophthalmology examination ruled out sarcoidosis.

The patient reported she continued breastfeeding her 4-year-old son. Additionally, she had been started on trazadone and buspirone for alcohol abuse recovery, then switched to and maintained on fluoxetine 1 year before developing these symptoms.

Buspirone, fluoxetine, and prolonged breastfeeding all contribute to hyperprolactinemia, a possible trigger of IGM. The patient was therefore advised to stop breastfeeding and to be switched from fluoxetine to a medication that would not increase the prolactin level. She did not require methotrexate treatment because her condition resolved rapidly after breastfeeding and fluoxetine were discontinued (Figure 2B).

Patient 2 

A 40-year-old woman with no history of breastfeeding who gave birth 4.5 years prior presented to her primary care provider with a painful breast lump and rash on the right breast of 2 months’ duration. Infectious mastitis was suspected; she was given cephalexin and clindamycin without improvement of symptoms.

Imaging
Mammography and ultrasonography were nondiagnostic.

Biopsy
Breast biopsy showed tissue with large expanses of histiocytes, neutrophils, lymphocytes, plasma cells, and multinucleated giant cells (Figure 3A). Many discrete granulomas were seen against this mixed inflammatory background, associated with focal fat necrosis (Figure 3B). Special stains were negative for microorganisms. Histologic findings were consistent with granulomatous mastitis.

On presentation to dermatology, the patient was worked up for a possible granulomatous etiology, which included a negative PPD, as well as a normal chest radiograph, serum Ca²⁺ and ACE levels, and ophthalmology examination. Review of symptoms (ROS),medical history, and medication review were unremarkable.

By exclusion, the patient was given a diagnosis of IGM and started on methotrexate (15 mg weekly) with folic acid (1 mg daily). The condition of the right breast improved within 4 weeks of starting methotrexate; however, methotrexate was increased to 30 mg weekly because of occasional flares. The patient remained on methotrexate without further IGM flares for 8 months compared to prior unremitting pain and drainage. She was then tapered from methotrexate over 6 weeks without additional flares.

 

 

Patient 3

A 27-year-old woman who gave birth 2 years prior and discontinued breastfeeding 6 weeks after delivery noted bilateral breast rashes for several months. The lesions were growing in size, tender, and draining. Her primary care provider suspected infectious mastitis and prescribed antibiotics, which were ineffective.

Biopsy
Breast core biopsy showed histologic findings similar to patients 1 and 2, including lobulocentric mixed inflammation, neutrophilic microabscesses, and scattered discrete granulomas. Microorganisms were not found using special stains. Breast cancer was ruled out, and granulomatous mastitis was diagnosed.

Referral to Dermatology
Two years earlier, the patient tested positive for latent tuberculosis and was prescribed a 9-month regimen of isoniazid. At the current presentation, she did not have symptoms of active tuberculosis on ROS (ie, no cough, hemoptysis, weight loss, night sweats); a chest radiograph was normal. Additionally, serum Ca²⁺ and ACE levels as well as an ophthalmology examination were normal, and she was not taking any medications known to increase the prolactin level.

The patient was started on methotrexate (12.5 mg weekly) and folic acid (1 mg daily). She had 1 IGM flare and was given a tapering regimen of prednisone. She received methotrexate for 14 months, tapered during the final 3 months. She has been off methotrexate for 3 years without IGM flares and appears to be in complete remission.

COMMENT

We report 3 cases of IGM, which contribute to the literature on possible presentations, causes, and conservative treatment of this rare connective-tissue disorder.

Differential Diagnosis

The time between recognition of symptoms and diagnosis and treatment of IGM often is prolonged because IGM can present similarly to other disorders, such as infection, breast cancer, tuberculosis, and sarcoidosis. Idiopathic granulomatous mastitis is a diagnosis of exclusion, made after obtaining evidence of granulomatous inflammation on breast biopsy and ruling out other granulomatous disorders, such as tuberculosis and sarcoidosis (Table 1).^3,4

Tuberculosis
A full ROS and a PPD test or T-SPOT.TB test can be helpful in ruling out tuberculosis; because anergy occurs in some patients, tuberculosis should be evaluated in the context of known immunosuppression or human immunodeficiency virus status, or in the case of miliary tuberculosis.

Chest radiography findings classically showing upper lobe infiltrates with cavities in active tuberculosis also should be sought.³ Ziehl-Neelsen staining of 2 sputum specimens, assessed by conventional light microscopy at the time of tissue biopsy has 64% sensitivity and 98% specificity for detecting Mycobacterium tuberculosis; auramine O staining, examined with light-emitting diode fluorescence microscopy, has 73% sensitivity and 93% specificity.⁵

Sarcoidosis
Because more than 90% of sarcoid patients have lung disease, a chest radiograph is used to screen for hilar lymphadenopathy.³ An elevated serum ACE level also can be helpful in diagnosis, but patients do not always have increased ACE, which can occur in other diseases, such as hyperthyroidism and miliary tuberculosis. Sarcoid granulomas can increase active vitamin D production, which in turn increases serum Ca²⁺ in 10% of sarcoid patients. Last, an ophthalmology evaluation should be obtained to rule out anterior or posterior uveitis that can occur in sarcoidosis and initially remain asymptomatic.³ Once these other causes of granulomatous inflammation have been ruled out, a diagnosis of IGM can be made.

Prolactinoma
Prolactinoma is an important cause of hyperprolactinemia that can be screened for based on ROS and the serum prolactin level. Prolactinoma can cause oligomenorrhea or amenorrhea and galactorrhea in 90% and 80% of premenopausal women, respectively, as well as erectile dysfunction and decreased libido in men. Infertility, headache, and visual impairment may be experienced in both sexes.⁴

A normal prolactin level is less than 25 μg/L; more than 25 μg/L but less than 100 μg/L usually is due to certain drugs (eTable),^6-11 estrogen, or idiopathic reasons; and more than 150 μg/L usually is due to prolactinoma.⁵ In many cases, removal of hyperprolactinemia-precipitating factors can resolve disease, as in patient 1. If symptoms continue or precipitating factors are absent, IGM symptom-based treatment should be administered.

Course and Management

Idiopathic granulomatous mastitis is self-limited and usually resolves within 2 years. Therefore, the goal of treatment is to suppress associated pain and drainage until the active inflammatory phase of IGM self-resolves. An established protocol for treating IGM does not exist, but common treatments include corticosteroids, methotrexate, and limited or wide surgical excision (Table 2).^12-16 Before beginning any of these treatments, IGM triggers, such as breastfeeding and drugs that induce hyperprolactinemia, should be removed.

It is important to consider which treatment option is best for limiting disease recurrence and adverse effects (AEs). Keep in mind that the available data are limited, as there are no randomized controlled trials looking at these treatments. Nevertheless, we recommend methotrexate as first line because it resolves granulomatous inflammation symptoms without invasive surgery, while limiting corticosteroid AEs.¹²

With or without concurrent use of corticosteroids, surgical excision typically is the mainstay of treatment. However, surgical excision of IGM lesions can be complicated by abscess formation, fistula, and chronic pyogenic secretions, in addition to a 5% to 50% rate of recurrence of disease.^12-14 Limited excision often is insufficient; therefore, wide local excision, in which negative margins around granulomatous inflammation are obtained, is the surgical modality of choice.¹⁴ Wide local excision can be disfiguring to the breast in young women affected by IGM, making it an undesirable treatment option.

Corticosteroids often have been used to treat IGM, but their efficacy is variable, symptoms can recur upon drug removal, and remarkable AEs can result from long-term use.¹² Additionally, corticosteroid therapy often is used in combination with excision, making it difficult to determine the extent to which corticosteroids or excision are more beneficial. In a prospective observational study, corticosteroid therapy alone resolved 80% of IGM symptoms after 159 days on average. After complete symptom resolution, 23% of patients had disease recurrence.⁹ Observed AEs included gastritis, weight gain, osteoporosis, glucose intolerance, and Cushing syndrome.^12,15

Methotrexate for IGM has not been reviewed in a randomized controlled trial; case reports have shown 83% symptom resolution, with 17% recurrence and limited long-term AEs.¹² Because the active phase of IGM can persist for 2 years, immunosuppressive therapy with limited AEs is necessary. Many AEs can occur when high-dose methotrexate is given for cancer treatment. Low-dose methotrexate has been extensively studied in long-term treatment of rheumatoid arthritis. Adverse effects may include gastrointestinal tract upset and hepatic dysfunction, which are limited when given with folic acid.

Regardless of folic acid cotreatment, stomatitis may occur. Women should use an effective method of birth control because severe birth defects may occur on even low-dose methotrexate.¹⁶

Compared to corticosteroid or surgical treatment, we recommend low-dose methotrexate therapy based on its high efficacy with limited AEs. An occasional mild flare of IGM symptoms with methotrexate is not unusual. If it occurs, corticosteroids can be added and tapered for as long as 2 weeks to speed up resolution of flares while reducing long-term AEs of corticosteroids.

Surgical excision can be performed in cases refractory to all systemic therapies.

CONCLUSION

Idiopathic granulomatous mastitis is a rare granulomatous breast disorder that can have a prolonged time to diagnosis, delaying proper treatment. Many cases self-resolve, but more severe cases can persist for a long period before adequate symptomatic treatment is achieved by methotrexate, corticosteroids, or surgical excision. Before using these therapies, it is important to identify and remove contributing factors, such as long-term breastfeeding and drugs that induce hyperprolactinemia. Improving the rate of IGM diagnosis and treatment would greatly benefit these patients. We report 1 case in which removal of possible precipitating IGM factors led to symptom resolution and 2 cases in which methotrexate was an effective IGM treatment that limited the need for invasive procedures and corticosteroid AEs.

Idiopathic granulomatous mastitis (IGM) is rare during pregnancy; it typically is seen in women of childbearing potential from 6 months to 6 years postpartum.¹ Because of a temporal association with breastfeeding, it is believed that hyperprolactinemia² or an immune response to local lobular secretions might play a role in pathogenesis. Early misdiagnosis as bacterial mastitis is common, prompting multiple antibiotic regimens. When antibiotics fail, patients are worked up for inflammatory breast cancer, given the nonhealing breast nodules. Mammography, ultrasonography, and fine-needle aspiration often are unable to rule out carcinoma, warranting excisional biopsies of nodules. The patient is then referred to rheumatology for potential sarcoidosis or to dermatology for IGM. In either case, the workup should be similar, but additional history focused on behavior and medications is essential in suspected IGM, given the association with hyperprolactinemia.

Because IGM is rare, there are no randomized, placebo-controlled trials of treatment efficacy. In many cases, patients undergo complete mastectomy, which is curative but may be psychologically and physically impactful in young women. In some cases, high-dose corticosteroids have been successful; however, because the IGM process can last longer than 2 years, patients treated in this manner are exposed to steroid morbidities.¹

We report 3 cases of IGM that add to the literature on possible contributing factors, clinical presentations, and treatments for this disease. We also demonstrate that appropriate trigger identification and steroid-sparing agents, specifically methotrexate, can be breast-saving as they can alleviate this debilitating condition, obviating the need for radical surgical intervention.

CASE REPORTS 

Patient 1

A 40-year-old woman with a 4-year history of breastfeeding noted a grape-sized nodule on the left breast that grew to the size of a grapefruit after 2 weeks. Ulceration and drainage periodically occurred, forming pink plaques along the lateral aspects of the breast after healing. Her primary care provider suspected infectious mastitis; she was given an oral antibiotic (cephalexin) and intravenous antibiotics without improvement.

Imaging
Subsequent magnetic resonance imaging revealed a large, irregular, enhancing mass within the outer left breast (6.5 cm at greatest dimension) with additional surrounding amorphous enhancement highly suspicious for malignancy. There also were multiple prominent left axillary lymph nodes, with the largest demonstrating a cortical thickness of 8 mm.

Biopsy
Core breast biopsy showed benign tissue with fat necrosis. Fine-needle aspiration revealed few benign ductal cells and rare histiocytes; because these findings were nondiagnostic and cancer was still a consideration, the patient underwent excisional biopsy.

Histologic sections of breast tissue showed extensive lobulocentric inflammation comprising histiocytes and lymphocytes, with neutrophils admixed and forming microabscesses (Figure 1A). Multinucleated giant cells and single-cell necrosis were seen, but true caseous necrosis was absent (Figure 1B). Duct spaces often contained inflammatory cells or secretions. Special stains for fungal and acid-fast bacterial microorganisms were negative.

Granulomatous lobular mastitis was diagnosed, and the patient was referred to dermatology. On presentation to dermatology, the left breast showed a 6-cm area of firm induration and overlying peau d’orange change to the epidermis (Figure 2A). Based on pathologic analysis, she was worked up for a possible granulomatous etiology. Negative purified protein derivative (tuberculin)(PPD) and a normal chest radiograph ruled out tuberculosis. Normal chest radiography, serum Ca²⁺ and angiotensin-converting enzyme (ACE) levels, and ophthalmology examination ruled out sarcoidosis.

The patient reported she continued breastfeeding her 4-year-old son. Additionally, she had been started on trazadone and buspirone for alcohol abuse recovery, then switched to and maintained on fluoxetine 1 year before developing these symptoms.

Buspirone, fluoxetine, and prolonged breastfeeding all contribute to hyperprolactinemia, a possible trigger of IGM. The patient was therefore advised to stop breastfeeding and to be switched from fluoxetine to a medication that would not increase the prolactin level. She did not require methotrexate treatment because her condition resolved rapidly after breastfeeding and fluoxetine were discontinued (Figure 2B).

Patient 2 

A 40-year-old woman with no history of breastfeeding who gave birth 4.5 years prior presented to her primary care provider with a painful breast lump and rash on the right breast of 2 months’ duration. Infectious mastitis was suspected; she was given cephalexin and clindamycin without improvement of symptoms.

Imaging
Mammography and ultrasonography were nondiagnostic.

Biopsy
Breast biopsy showed tissue with large expanses of histiocytes, neutrophils, lymphocytes, plasma cells, and multinucleated giant cells (Figure 3A). Many discrete granulomas were seen against this mixed inflammatory background, associated with focal fat necrosis (Figure 3B). Special stains were negative for microorganisms. Histologic findings were consistent with granulomatous mastitis.

On presentation to dermatology, the patient was worked up for a possible granulomatous etiology, which included a negative PPD, as well as a normal chest radiograph, serum Ca²⁺ and ACE levels, and ophthalmology examination. Review of symptoms (ROS),medical history, and medication review were unremarkable.

By exclusion, the patient was given a diagnosis of IGM and started on methotrexate (15 mg weekly) with folic acid (1 mg daily). The condition of the right breast improved within 4 weeks of starting methotrexate; however, methotrexate was increased to 30 mg weekly because of occasional flares. The patient remained on methotrexate without further IGM flares for 8 months compared to prior unremitting pain and drainage. She was then tapered from methotrexate over 6 weeks without additional flares.

 

 

Patient 3

A 27-year-old woman who gave birth 2 years prior and discontinued breastfeeding 6 weeks after delivery noted bilateral breast rashes for several months. The lesions were growing in size, tender, and draining. Her primary care provider suspected infectious mastitis and prescribed antibiotics, which were ineffective.

Biopsy
Breast core biopsy showed histologic findings similar to patients 1 and 2, including lobulocentric mixed inflammation, neutrophilic microabscesses, and scattered discrete granulomas. Microorganisms were not found using special stains. Breast cancer was ruled out, and granulomatous mastitis was diagnosed.

Referral to Dermatology
Two years earlier, the patient tested positive for latent tuberculosis and was prescribed a 9-month regimen of isoniazid. At the current presentation, she did not have symptoms of active tuberculosis on ROS (ie, no cough, hemoptysis, weight loss, night sweats); a chest radiograph was normal. Additionally, serum Ca²⁺ and ACE levels as well as an ophthalmology examination were normal, and she was not taking any medications known to increase the prolactin level.

The patient was started on methotrexate (12.5 mg weekly) and folic acid (1 mg daily). She had 1 IGM flare and was given a tapering regimen of prednisone. She received methotrexate for 14 months, tapered during the final 3 months. She has been off methotrexate for 3 years without IGM flares and appears to be in complete remission.

COMMENT

We report 3 cases of IGM, which contribute to the literature on possible presentations, causes, and conservative treatment of this rare connective-tissue disorder.

Differential Diagnosis

The time between recognition of symptoms and diagnosis and treatment of IGM often is prolonged because IGM can present similarly to other disorders, such as infection, breast cancer, tuberculosis, and sarcoidosis. Idiopathic granulomatous mastitis is a diagnosis of exclusion, made after obtaining evidence of granulomatous inflammation on breast biopsy and ruling out other granulomatous disorders, such as tuberculosis and sarcoidosis (Table 1).^3,4

Tuberculosis
A full ROS and a PPD test or T-SPOT.TB test can be helpful in ruling out tuberculosis; because anergy occurs in some patients, tuberculosis should be evaluated in the context of known immunosuppression or human immunodeficiency virus status, or in the case of miliary tuberculosis.

Chest radiography findings classically showing upper lobe infiltrates with cavities in active tuberculosis also should be sought.³ Ziehl-Neelsen staining of 2 sputum specimens, assessed by conventional light microscopy at the time of tissue biopsy has 64% sensitivity and 98% specificity for detecting Mycobacterium tuberculosis; auramine O staining, examined with light-emitting diode fluorescence microscopy, has 73% sensitivity and 93% specificity.⁵

Sarcoidosis
Because more than 90% of sarcoid patients have lung disease, a chest radiograph is used to screen for hilar lymphadenopathy.³ An elevated serum ACE level also can be helpful in diagnosis, but patients do not always have increased ACE, which can occur in other diseases, such as hyperthyroidism and miliary tuberculosis. Sarcoid granulomas can increase active vitamin D production, which in turn increases serum Ca²⁺ in 10% of sarcoid patients. Last, an ophthalmology evaluation should be obtained to rule out anterior or posterior uveitis that can occur in sarcoidosis and initially remain asymptomatic.³ Once these other causes of granulomatous inflammation have been ruled out, a diagnosis of IGM can be made.

Prolactinoma
Prolactinoma is an important cause of hyperprolactinemia that can be screened for based on ROS and the serum prolactin level. Prolactinoma can cause oligomenorrhea or amenorrhea and galactorrhea in 90% and 80% of premenopausal women, respectively, as well as erectile dysfunction and decreased libido in men. Infertility, headache, and visual impairment may be experienced in both sexes.⁴

A normal prolactin level is less than 25 μg/L; more than 25 μg/L but less than 100 μg/L usually is due to certain drugs (eTable),^6-11 estrogen, or idiopathic reasons; and more than 150 μg/L usually is due to prolactinoma.⁵ In many cases, removal of hyperprolactinemia-precipitating factors can resolve disease, as in patient 1. If symptoms continue or precipitating factors are absent, IGM symptom-based treatment should be administered.

Course and Management

Idiopathic granulomatous mastitis is self-limited and usually resolves within 2 years. Therefore, the goal of treatment is to suppress associated pain and drainage until the active inflammatory phase of IGM self-resolves. An established protocol for treating IGM does not exist, but common treatments include corticosteroids, methotrexate, and limited or wide surgical excision (Table 2).^12-16 Before beginning any of these treatments, IGM triggers, such as breastfeeding and drugs that induce hyperprolactinemia, should be removed.

It is important to consider which treatment option is best for limiting disease recurrence and adverse effects (AEs). Keep in mind that the available data are limited, as there are no randomized controlled trials looking at these treatments. Nevertheless, we recommend methotrexate as first line because it resolves granulomatous inflammation symptoms without invasive surgery, while limiting corticosteroid AEs.¹²

With or without concurrent use of corticosteroids, surgical excision typically is the mainstay of treatment. However, surgical excision of IGM lesions can be complicated by abscess formation, fistula, and chronic pyogenic secretions, in addition to a 5% to 50% rate of recurrence of disease.^12-14 Limited excision often is insufficient; therefore, wide local excision, in which negative margins around granulomatous inflammation are obtained, is the surgical modality of choice.¹⁴ Wide local excision can be disfiguring to the breast in young women affected by IGM, making it an undesirable treatment option.

Corticosteroids often have been used to treat IGM, but their efficacy is variable, symptoms can recur upon drug removal, and remarkable AEs can result from long-term use.¹² Additionally, corticosteroid therapy often is used in combination with excision, making it difficult to determine the extent to which corticosteroids or excision are more beneficial. In a prospective observational study, corticosteroid therapy alone resolved 80% of IGM symptoms after 159 days on average. After complete symptom resolution, 23% of patients had disease recurrence.⁹ Observed AEs included gastritis, weight gain, osteoporosis, glucose intolerance, and Cushing syndrome.^12,15

Methotrexate for IGM has not been reviewed in a randomized controlled trial; case reports have shown 83% symptom resolution, with 17% recurrence and limited long-term AEs.¹² Because the active phase of IGM can persist for 2 years, immunosuppressive therapy with limited AEs is necessary. Many AEs can occur when high-dose methotrexate is given for cancer treatment. Low-dose methotrexate has been extensively studied in long-term treatment of rheumatoid arthritis. Adverse effects may include gastrointestinal tract upset and hepatic dysfunction, which are limited when given with folic acid.

Regardless of folic acid cotreatment, stomatitis may occur. Women should use an effective method of birth control because severe birth defects may occur on even low-dose methotrexate.¹⁶

Compared to corticosteroid or surgical treatment, we recommend low-dose methotrexate therapy based on its high efficacy with limited AEs. An occasional mild flare of IGM symptoms with methotrexate is not unusual. If it occurs, corticosteroids can be added and tapered for as long as 2 weeks to speed up resolution of flares while reducing long-term AEs of corticosteroids.

Surgical excision can be performed in cases refractory to all systemic therapies.

CONCLUSION

Idiopathic granulomatous mastitis is a rare granulomatous breast disorder that can have a prolonged time to diagnosis, delaying proper treatment. Many cases self-resolve, but more severe cases can persist for a long period before adequate symptomatic treatment is achieved by methotrexate, corticosteroids, or surgical excision. Before using these therapies, it is important to identify and remove contributing factors, such as long-term breastfeeding and drugs that induce hyperprolactinemia. Improving the rate of IGM diagnosis and treatment would greatly benefit these patients. We report 1 case in which removal of possible precipitating IGM factors led to symptom resolution and 2 cases in which methotrexate was an effective IGM treatment that limited the need for invasive procedures and corticosteroid AEs.

Idiopathic granulomatous mastitis (IGM) is rare during pregnancy; it typically is seen in women of childbearing potential from 6 months to 6 years postpartum.¹ Because of a temporal association with breastfeeding, it is believed that hyperprolactinemia² or an immune response to local lobular secretions might play a role in pathogenesis. Early misdiagnosis as bacterial mastitis is common, prompting multiple antibiotic regimens. When antibiotics fail, patients are worked up for inflammatory breast cancer, given the nonhealing breast nodules. Mammography, ultrasonography, and fine-needle aspiration often are unable to rule out carcinoma, warranting excisional biopsies of nodules. The patient is then referred to rheumatology for potential sarcoidosis or to dermatology for IGM. In either case, the workup should be similar, but additional history focused on behavior and medications is essential in suspected IGM, given the association with hyperprolactinemia.

Because IGM is rare, there are no randomized, placebo-controlled trials of treatment efficacy. In many cases, patients undergo complete mastectomy, which is curative but may be psychologically and physically impactful in young women. In some cases, high-dose corticosteroids have been successful; however, because the IGM process can last longer than 2 years, patients treated in this manner are exposed to steroid morbidities.¹

We report 3 cases of IGM that add to the literature on possible contributing factors, clinical presentations, and treatments for this disease. We also demonstrate that appropriate trigger identification and steroid-sparing agents, specifically methotrexate, can be breast-saving as they can alleviate this debilitating condition, obviating the need for radical surgical intervention.

CASE REPORTS 

Patient 1

A 40-year-old woman with a 4-year history of breastfeeding noted a grape-sized nodule on the left breast that grew to the size of a grapefruit after 2 weeks. Ulceration and drainage periodically occurred, forming pink plaques along the lateral aspects of the breast after healing. Her primary care provider suspected infectious mastitis; she was given an oral antibiotic (cephalexin) and intravenous antibiotics without improvement.

Imaging
Subsequent magnetic resonance imaging revealed a large, irregular, enhancing mass within the outer left breast (6.5 cm at greatest dimension) with additional surrounding amorphous enhancement highly suspicious for malignancy. There also were multiple prominent left axillary lymph nodes, with the largest demonstrating a cortical thickness of 8 mm.

Biopsy
Core breast biopsy showed benign tissue with fat necrosis. Fine-needle aspiration revealed few benign ductal cells and rare histiocytes; because these findings were nondiagnostic and cancer was still a consideration, the patient underwent excisional biopsy.

Histologic sections of breast tissue showed extensive lobulocentric inflammation comprising histiocytes and lymphocytes, with neutrophils admixed and forming microabscesses (Figure 1A). Multinucleated giant cells and single-cell necrosis were seen, but true caseous necrosis was absent (Figure 1B). Duct spaces often contained inflammatory cells or secretions. Special stains for fungal and acid-fast bacterial microorganisms were negative.

Granulomatous lobular mastitis was diagnosed, and the patient was referred to dermatology. On presentation to dermatology, the left breast showed a 6-cm area of firm induration and overlying peau d’orange change to the epidermis (Figure 2A). Based on pathologic analysis, she was worked up for a possible granulomatous etiology. Negative purified protein derivative (tuberculin)(PPD) and a normal chest radiograph ruled out tuberculosis. Normal chest radiography, serum Ca²⁺ and angiotensin-converting enzyme (ACE) levels, and ophthalmology examination ruled out sarcoidosis.

The patient reported she continued breastfeeding her 4-year-old son. Additionally, she had been started on trazadone and buspirone for alcohol abuse recovery, then switched to and maintained on fluoxetine 1 year before developing these symptoms.

Buspirone, fluoxetine, and prolonged breastfeeding all contribute to hyperprolactinemia, a possible trigger of IGM. The patient was therefore advised to stop breastfeeding and to be switched from fluoxetine to a medication that would not increase the prolactin level. She did not require methotrexate treatment because her condition resolved rapidly after breastfeeding and fluoxetine were discontinued (Figure 2B).

Patient 2 

A 40-year-old woman with no history of breastfeeding who gave birth 4.5 years prior presented to her primary care provider with a painful breast lump and rash on the right breast of 2 months’ duration. Infectious mastitis was suspected; she was given cephalexin and clindamycin without improvement of symptoms.

Imaging
Mammography and ultrasonography were nondiagnostic.

Biopsy
Breast biopsy showed tissue with large expanses of histiocytes, neutrophils, lymphocytes, plasma cells, and multinucleated giant cells (Figure 3A). Many discrete granulomas were seen against this mixed inflammatory background, associated with focal fat necrosis (Figure 3B). Special stains were negative for microorganisms. Histologic findings were consistent with granulomatous mastitis.

On presentation to dermatology, the patient was worked up for a possible granulomatous etiology, which included a negative PPD, as well as a normal chest radiograph, serum Ca²⁺ and ACE levels, and ophthalmology examination. Review of symptoms (ROS),medical history, and medication review were unremarkable.

By exclusion, the patient was given a diagnosis of IGM and started on methotrexate (15 mg weekly) with folic acid (1 mg daily). The condition of the right breast improved within 4 weeks of starting methotrexate; however, methotrexate was increased to 30 mg weekly because of occasional flares. The patient remained on methotrexate without further IGM flares for 8 months compared to prior unremitting pain and drainage. She was then tapered from methotrexate over 6 weeks without additional flares.

 

 

Patient 3

A 27-year-old woman who gave birth 2 years prior and discontinued breastfeeding 6 weeks after delivery noted bilateral breast rashes for several months. The lesions were growing in size, tender, and draining. Her primary care provider suspected infectious mastitis and prescribed antibiotics, which were ineffective.

Biopsy
Breast core biopsy showed histologic findings similar to patients 1 and 2, including lobulocentric mixed inflammation, neutrophilic microabscesses, and scattered discrete granulomas. Microorganisms were not found using special stains. Breast cancer was ruled out, and granulomatous mastitis was diagnosed.

Referral to Dermatology
Two years earlier, the patient tested positive for latent tuberculosis and was prescribed a 9-month regimen of isoniazid. At the current presentation, she did not have symptoms of active tuberculosis on ROS (ie, no cough, hemoptysis, weight loss, night sweats); a chest radiograph was normal. Additionally, serum Ca²⁺ and ACE levels as well as an ophthalmology examination were normal, and she was not taking any medications known to increase the prolactin level.

The patient was started on methotrexate (12.5 mg weekly) and folic acid (1 mg daily). She had 1 IGM flare and was given a tapering regimen of prednisone. She received methotrexate for 14 months, tapered during the final 3 months. She has been off methotrexate for 3 years without IGM flares and appears to be in complete remission.

COMMENT

We report 3 cases of IGM, which contribute to the literature on possible presentations, causes, and conservative treatment of this rare connective-tissue disorder.

Differential Diagnosis

The time between recognition of symptoms and diagnosis and treatment of IGM often is prolonged because IGM can present similarly to other disorders, such as infection, breast cancer, tuberculosis, and sarcoidosis. Idiopathic granulomatous mastitis is a diagnosis of exclusion, made after obtaining evidence of granulomatous inflammation on breast biopsy and ruling out other granulomatous disorders, such as tuberculosis and sarcoidosis (Table 1).^3,4

Tuberculosis
A full ROS and a PPD test or T-SPOT.TB test can be helpful in ruling out tuberculosis; because anergy occurs in some patients, tuberculosis should be evaluated in the context of known immunosuppression or human immunodeficiency virus status, or in the case of miliary tuberculosis.

Chest radiography findings classically showing upper lobe infiltrates with cavities in active tuberculosis also should be sought.³ Ziehl-Neelsen staining of 2 sputum specimens, assessed by conventional light microscopy at the time of tissue biopsy has 64% sensitivity and 98% specificity for detecting Mycobacterium tuberculosis; auramine O staining, examined with light-emitting diode fluorescence microscopy, has 73% sensitivity and 93% specificity.⁵

Sarcoidosis
Because more than 90% of sarcoid patients have lung disease, a chest radiograph is used to screen for hilar lymphadenopathy.³ An elevated serum ACE level also can be helpful in diagnosis, but patients do not always have increased ACE, which can occur in other diseases, such as hyperthyroidism and miliary tuberculosis. Sarcoid granulomas can increase active vitamin D production, which in turn increases serum Ca²⁺ in 10% of sarcoid patients. Last, an ophthalmology evaluation should be obtained to rule out anterior or posterior uveitis that can occur in sarcoidosis and initially remain asymptomatic.³ Once these other causes of granulomatous inflammation have been ruled out, a diagnosis of IGM can be made.

Prolactinoma
Prolactinoma is an important cause of hyperprolactinemia that can be screened for based on ROS and the serum prolactin level. Prolactinoma can cause oligomenorrhea or amenorrhea and galactorrhea in 90% and 80% of premenopausal women, respectively, as well as erectile dysfunction and decreased libido in men. Infertility, headache, and visual impairment may be experienced in both sexes.⁴

A normal prolactin level is less than 25 μg/L; more than 25 μg/L but less than 100 μg/L usually is due to certain drugs (eTable),^6-11 estrogen, or idiopathic reasons; and more than 150 μg/L usually is due to prolactinoma.⁵ In many cases, removal of hyperprolactinemia-precipitating factors can resolve disease, as in patient 1. If symptoms continue or precipitating factors are absent, IGM symptom-based treatment should be administered.

Course and Management

Idiopathic granulomatous mastitis is self-limited and usually resolves within 2 years. Therefore, the goal of treatment is to suppress associated pain and drainage until the active inflammatory phase of IGM self-resolves. An established protocol for treating IGM does not exist, but common treatments include corticosteroids, methotrexate, and limited or wide surgical excision (Table 2).^12-16 Before beginning any of these treatments, IGM triggers, such as breastfeeding and drugs that induce hyperprolactinemia, should be removed.

It is important to consider which treatment option is best for limiting disease recurrence and adverse effects (AEs). Keep in mind that the available data are limited, as there are no randomized controlled trials looking at these treatments. Nevertheless, we recommend methotrexate as first line because it resolves granulomatous inflammation symptoms without invasive surgery, while limiting corticosteroid AEs.¹²

With or without concurrent use of corticosteroids, surgical excision typically is the mainstay of treatment. However, surgical excision of IGM lesions can be complicated by abscess formation, fistula, and chronic pyogenic secretions, in addition to a 5% to 50% rate of recurrence of disease.^12-14 Limited excision often is insufficient; therefore, wide local excision, in which negative margins around granulomatous inflammation are obtained, is the surgical modality of choice.¹⁴ Wide local excision can be disfiguring to the breast in young women affected by IGM, making it an undesirable treatment option.

Corticosteroids often have been used to treat IGM, but their efficacy is variable, symptoms can recur upon drug removal, and remarkable AEs can result from long-term use.¹² Additionally, corticosteroid therapy often is used in combination with excision, making it difficult to determine the extent to which corticosteroids or excision are more beneficial. In a prospective observational study, corticosteroid therapy alone resolved 80% of IGM symptoms after 159 days on average. After complete symptom resolution, 23% of patients had disease recurrence.⁹ Observed AEs included gastritis, weight gain, osteoporosis, glucose intolerance, and Cushing syndrome.^12,15

Methotrexate for IGM has not been reviewed in a randomized controlled trial; case reports have shown 83% symptom resolution, with 17% recurrence and limited long-term AEs.¹² Because the active phase of IGM can persist for 2 years, immunosuppressive therapy with limited AEs is necessary. Many AEs can occur when high-dose methotrexate is given for cancer treatment. Low-dose methotrexate has been extensively studied in long-term treatment of rheumatoid arthritis. Adverse effects may include gastrointestinal tract upset and hepatic dysfunction, which are limited when given with folic acid.

Regardless of folic acid cotreatment, stomatitis may occur. Women should use an effective method of birth control because severe birth defects may occur on even low-dose methotrexate.¹⁶

Compared to corticosteroid or surgical treatment, we recommend low-dose methotrexate therapy based on its high efficacy with limited AEs. An occasional mild flare of IGM symptoms with methotrexate is not unusual. If it occurs, corticosteroids can be added and tapered for as long as 2 weeks to speed up resolution of flares while reducing long-term AEs of corticosteroids.

Surgical excision can be performed in cases refractory to all systemic therapies.

CONCLUSION

Idiopathic granulomatous mastitis is a rare granulomatous breast disorder that can have a prolonged time to diagnosis, delaying proper treatment. Many cases self-resolve, but more severe cases can persist for a long period before adequate symptomatic treatment is achieved by methotrexate, corticosteroids, or surgical excision. Before using these therapies, it is important to identify and remove contributing factors, such as long-term breastfeeding and drugs that induce hyperprolactinemia. Improving the rate of IGM diagnosis and treatment would greatly benefit these patients. We report 1 case in which removal of possible precipitating IGM factors led to symptom resolution and 2 cases in which methotrexate was an effective IGM treatment that limited the need for invasive procedures and corticosteroid AEs.

En coup de sabre (ECDS) is a rare subtype of linear scleroderma that is limited to the hemiface in a unilateral distribution. The lesional skin first exhibits contraction and stiffness that lead to characteristic fibrotic plaques with associated linear alopecia.¹ The pansclerotic plaques are ivory in color with hyperpigmented to violaceous borders extending as a paramedian band on the frontoparietal scalp.^2,3 The skin lesions bear resemblance to the stroke of the sabre sword, giving the condition its unique name. Many patients initially present with concerns of frontal scalp alopecia.³ Linear morphea, including the ECDS subtype, is predominantly seen in children and women, usually presenting within the first 2 decades of life.^1,4

The differential diagnoses of ECDS include focal dermal hypoplasia, steroid atrophy, localized morphea, and lupus profundus.⁵ En coup de sabre should be distinguished from progressive hemifacial atrophy (PHA)(also known as Parry-Romberg syndrome).⁶ Progressive hemifacial atrophy presents as unilateral atrophy of the face involving skin, subcutaneous tissue, muscle, and underlying bone in the distribution of the trigeminal nerve.¹ Both PHA and ECDS exist on a spectrum of linear scleroderma and may coexist in the same patient.⁶

There is a strong association with extracutaneous neurologic involvement, including seizures, ocular abnormalities, trigeminal neuralgia, and headache.^7-10 One study examining ECDS and PHA demonstrated that 44% (19/43) of patients who underwent central nervous system imaging had abnormal findings.¹¹ The majority of patients had magnetic resonance imaging with or without contrast, computed tomography, or both. The most common findings on T2-weighted images were white matter hyperintensities, mostly in subcortical and periventricular regions. The findings were bilateral in 61% (11/18) of patients and ipsilateral to the lesion in 33% (6/18) of patients.¹¹ We present a case of ECDS masquerading as alopecia in a 77-year-old woman.

Case Report

A 77-year-old white woman presented with a chief concern of hair loss on the scalp that had been present since 12 years of age. During her adult life, the scalp lesion remained unchanged with no associated symptoms. Her medical history was remarkable for hypertension and non–insulin-dependent diabetes mellitus. The patient denied any history of seizure disorders, facial paralysis, or neurologic deficits. Physical examination revealed a 13.6-cm linear, alopecic, hyperpigmented plaque extending from the left forehead and temporal scalp to the posterior occipital scalp with notable loss of underlying subcutaneous tissue (Figure). The left temporal scalp and forehead demonstrated pronounced atrophy with overlying telangiectases. Minimal depression was observed in the temporal scalp and forehead. The left cranial bone was easily palpable with no underlying subcutaneous tissue present. Computed tomography of the brain revealed no underlying skull or soft tissue abnormalities. No treatment was initiated, as the lesion was of minimal concern to the patient.

Comment

Etiology and Presentation                                                                                                                                                                                                                                      
En coup de sabre is a rare subtype of linear morphea that involves the frontoparietal scalp and forehead.^7,12,13 It manifests as a solitary, linear, fibrous plaque that involves the skin, underlying muscle, and bone.⁷ Although most cases present with a single lesion, multiple lesions can occur.⁸ The exact etiology of this disease remains to be determined but is characterized by thickening and hardening of the skin secondary to increased collagen production.⁷ The incidence of linear morphea ranges from 0.4 to 2.7 cases per 100,000 individuals and is more prevalent in white patients and women.¹⁴ Linear morphea is commonly found in children. Children are more likely to have linear morphea on the face, which can lead to lifelong disfigurement.² Although the disease peaks in the fourth decade of life for adults, most pediatric cases are diagnosed between 2 and 14 years of age.^14-16

Pathogenesis                                                                                                                                                                                                                                                       
Clinical and histopathological data suggest that a complex interaction among the vasculature, extracellular matrix, and immune system plays a role in the pathogenesis of the disease. Similar to scleroderma, the CD4 helper T cell may be involved in the fibrotic changes that occur within these lesions.¹⁷ Early in the disease process, T_H1 and T_H17 inflammatory pathways predominate. The late fibrotic changes seen in scleroderma are more associated with a shift to the T_H2 inflammatory pathway.¹⁷ Infection with Borrelia burgdorferi has been implicated abroad, but a large-scale study confirming Borrelia as a pathologic factor within morphea lesions has not been completed to date.^18-20 Some authors believe early lesions of ECDS mimic erythema chronica migrans, with the late lesions resembling acrodermatitis chronica atrophicans.²⁰

Histopathology                                                                                                                                                                                                                                     
Histopathologic findings of morphea tend to vary depending on the stage of the disease. The 2 stages of morphea can be differentiated by the degree of inflammation present histologically.^14,21 The early phase of morphea primarily affects the connective and subcutaneous tissue surrounding eccrine sweat glands.^14,21 A dense dermal and subcutaneous perivascular lymphocytic infiltrate with a mixture of lymphocytes, plasma cells, and histiocytes is commonly observed.⁵ Later stages of the disease demonstrate densely packed homogenous collagen with minimal inflammation and loss of eccrine glands and blood vessels.^14,21 The adipose tissue is generally replaced by sclerotic collagen, giving the biopsy a squared-off appearance.^5,14

Management                                                                                                                                                                                                                                                            
En coup de sabre presents a treatment challenge. In active lesions, topical or intralesional corticosteroids are considered treatment of choice.⁵ Methotrexate has proven useful in the treatment of acute and deep forms of linear morphea. A study examining methotrexate in juvenile localized scleroderma, with the majority of patients having the linear subtype, revealed that methotrexate is both efficacious and well tolerated.²² Other reports in the literature reveal efficacy with the use of intravenous corticosteroids and methotrexate combination therapy for treatment of morphea.^23,24 A longitudinal prospective study examining the use of high-dose methotrexate and oral corticosteroids for the treatment of localized scleroderma yielded positive results, with patients showing clinical improvement within 2 months of initiation of combination therapy.²⁵ Other treatments include excimer laser; calcipotriene and tacrolimus; and surgical approaches such as autologous fat grafting, grafting with muscle flaps, and tissue inserts.^21,26-31 In addition, patients can choose to forego therapy, as was the case with our patient.

Conclusion

En coup de sabre is a rare subtype of linear scleroderma that is limited to the ipsilateral scalp and face predominately in children and women. Neurologic involvement is common and should prompt a comprehensive neurologic workup in patients suspected to have ECDS or PHA. Current treatment recommendations include topical, intralesional, and oral corticosteroids; methotrexate; and surgical grafts. Although ECDS is a rare entity, more intensive research is needed on the exact pathophysiology and effective treatment options that focus on improving the cosmetic outcome in these patients. Cosmesis is the primary concern in patients with ECDS and should be managed early and appropriately to prevent long-term psychological sequelae.

En coup de sabre (ECDS) is a rare subtype of linear scleroderma that is limited to the hemiface in a unilateral distribution. The lesional skin first exhibits contraction and stiffness that lead to characteristic fibrotic plaques with associated linear alopecia.¹ The pansclerotic plaques are ivory in color with hyperpigmented to violaceous borders extending as a paramedian band on the frontoparietal scalp.^2,3 The skin lesions bear resemblance to the stroke of the sabre sword, giving the condition its unique name. Many patients initially present with concerns of frontal scalp alopecia.³ Linear morphea, including the ECDS subtype, is predominantly seen in children and women, usually presenting within the first 2 decades of life.^1,4

The differential diagnoses of ECDS include focal dermal hypoplasia, steroid atrophy, localized morphea, and lupus profundus.⁵ En coup de sabre should be distinguished from progressive hemifacial atrophy (PHA)(also known as Parry-Romberg syndrome).⁶ Progressive hemifacial atrophy presents as unilateral atrophy of the face involving skin, subcutaneous tissue, muscle, and underlying bone in the distribution of the trigeminal nerve.¹ Both PHA and ECDS exist on a spectrum of linear scleroderma and may coexist in the same patient.⁶

There is a strong association with extracutaneous neurologic involvement, including seizures, ocular abnormalities, trigeminal neuralgia, and headache.^7-10 One study examining ECDS and PHA demonstrated that 44% (19/43) of patients who underwent central nervous system imaging had abnormal findings.¹¹ The majority of patients had magnetic resonance imaging with or without contrast, computed tomography, or both. The most common findings on T2-weighted images were white matter hyperintensities, mostly in subcortical and periventricular regions. The findings were bilateral in 61% (11/18) of patients and ipsilateral to the lesion in 33% (6/18) of patients.¹¹ We present a case of ECDS masquerading as alopecia in a 77-year-old woman.

Case Report

A 77-year-old white woman presented with a chief concern of hair loss on the scalp that had been present since 12 years of age. During her adult life, the scalp lesion remained unchanged with no associated symptoms. Her medical history was remarkable for hypertension and non–insulin-dependent diabetes mellitus. The patient denied any history of seizure disorders, facial paralysis, or neurologic deficits. Physical examination revealed a 13.6-cm linear, alopecic, hyperpigmented plaque extending from the left forehead and temporal scalp to the posterior occipital scalp with notable loss of underlying subcutaneous tissue (Figure). The left temporal scalp and forehead demonstrated pronounced atrophy with overlying telangiectases. Minimal depression was observed in the temporal scalp and forehead. The left cranial bone was easily palpable with no underlying subcutaneous tissue present. Computed tomography of the brain revealed no underlying skull or soft tissue abnormalities. No treatment was initiated, as the lesion was of minimal concern to the patient.

Comment

Etiology and Presentation                                                                                                                                                                                                                                      
En coup de sabre is a rare subtype of linear morphea that involves the frontoparietal scalp and forehead.^7,12,13 It manifests as a solitary, linear, fibrous plaque that involves the skin, underlying muscle, and bone.⁷ Although most cases present with a single lesion, multiple lesions can occur.⁸ The exact etiology of this disease remains to be determined but is characterized by thickening and hardening of the skin secondary to increased collagen production.⁷ The incidence of linear morphea ranges from 0.4 to 2.7 cases per 100,000 individuals and is more prevalent in white patients and women.¹⁴ Linear morphea is commonly found in children. Children are more likely to have linear morphea on the face, which can lead to lifelong disfigurement.² Although the disease peaks in the fourth decade of life for adults, most pediatric cases are diagnosed between 2 and 14 years of age.^14-16

Pathogenesis                                                                                                                                                                                                                                                       
Clinical and histopathological data suggest that a complex interaction among the vasculature, extracellular matrix, and immune system plays a role in the pathogenesis of the disease. Similar to scleroderma, the CD4 helper T cell may be involved in the fibrotic changes that occur within these lesions.¹⁷ Early in the disease process, T_H1 and T_H17 inflammatory pathways predominate. The late fibrotic changes seen in scleroderma are more associated with a shift to the T_H2 inflammatory pathway.¹⁷ Infection with Borrelia burgdorferi has been implicated abroad, but a large-scale study confirming Borrelia as a pathologic factor within morphea lesions has not been completed to date.^18-20 Some authors believe early lesions of ECDS mimic erythema chronica migrans, with the late lesions resembling acrodermatitis chronica atrophicans.²⁰

Histopathology                                                                                                                                                                                                                                     
Histopathologic findings of morphea tend to vary depending on the stage of the disease. The 2 stages of morphea can be differentiated by the degree of inflammation present histologically.^14,21 The early phase of morphea primarily affects the connective and subcutaneous tissue surrounding eccrine sweat glands.^14,21 A dense dermal and subcutaneous perivascular lymphocytic infiltrate with a mixture of lymphocytes, plasma cells, and histiocytes is commonly observed.⁵ Later stages of the disease demonstrate densely packed homogenous collagen with minimal inflammation and loss of eccrine glands and blood vessels.^14,21 The adipose tissue is generally replaced by sclerotic collagen, giving the biopsy a squared-off appearance.^5,14

Management                                                                                                                                                                                                                                                            
En coup de sabre presents a treatment challenge. In active lesions, topical or intralesional corticosteroids are considered treatment of choice.⁵ Methotrexate has proven useful in the treatment of acute and deep forms of linear morphea. A study examining methotrexate in juvenile localized scleroderma, with the majority of patients having the linear subtype, revealed that methotrexate is both efficacious and well tolerated.²² Other reports in the literature reveal efficacy with the use of intravenous corticosteroids and methotrexate combination therapy for treatment of morphea.^23,24 A longitudinal prospective study examining the use of high-dose methotrexate and oral corticosteroids for the treatment of localized scleroderma yielded positive results, with patients showing clinical improvement within 2 months of initiation of combination therapy.²⁵ Other treatments include excimer laser; calcipotriene and tacrolimus; and surgical approaches such as autologous fat grafting, grafting with muscle flaps, and tissue inserts.^21,26-31 In addition, patients can choose to forego therapy, as was the case with our patient.

Conclusion

En coup de sabre is a rare subtype of linear scleroderma that is limited to the ipsilateral scalp and face predominately in children and women. Neurologic involvement is common and should prompt a comprehensive neurologic workup in patients suspected to have ECDS or PHA. Current treatment recommendations include topical, intralesional, and oral corticosteroids; methotrexate; and surgical grafts. Although ECDS is a rare entity, more intensive research is needed on the exact pathophysiology and effective treatment options that focus on improving the cosmetic outcome in these patients. Cosmesis is the primary concern in patients with ECDS and should be managed early and appropriately to prevent long-term psychological sequelae.

En coup de sabre (ECDS) is a rare subtype of linear scleroderma that is limited to the hemiface in a unilateral distribution. The lesional skin first exhibits contraction and stiffness that lead to characteristic fibrotic plaques with associated linear alopecia.¹ The pansclerotic plaques are ivory in color with hyperpigmented to violaceous borders extending as a paramedian band on the frontoparietal scalp.^2,3 The skin lesions bear resemblance to the stroke of the sabre sword, giving the condition its unique name. Many patients initially present with concerns of frontal scalp alopecia.³ Linear morphea, including the ECDS subtype, is predominantly seen in children and women, usually presenting within the first 2 decades of life.^1,4

The differential diagnoses of ECDS include focal dermal hypoplasia, steroid atrophy, localized morphea, and lupus profundus.⁵ En coup de sabre should be distinguished from progressive hemifacial atrophy (PHA)(also known as Parry-Romberg syndrome).⁶ Progressive hemifacial atrophy presents as unilateral atrophy of the face involving skin, subcutaneous tissue, muscle, and underlying bone in the distribution of the trigeminal nerve.¹ Both PHA and ECDS exist on a spectrum of linear scleroderma and may coexist in the same patient.⁶

There is a strong association with extracutaneous neurologic involvement, including seizures, ocular abnormalities, trigeminal neuralgia, and headache.^7-10 One study examining ECDS and PHA demonstrated that 44% (19/43) of patients who underwent central nervous system imaging had abnormal findings.¹¹ The majority of patients had magnetic resonance imaging with or without contrast, computed tomography, or both. The most common findings on T2-weighted images were white matter hyperintensities, mostly in subcortical and periventricular regions. The findings were bilateral in 61% (11/18) of patients and ipsilateral to the lesion in 33% (6/18) of patients.¹¹ We present a case of ECDS masquerading as alopecia in a 77-year-old woman.

Case Report

A 77-year-old white woman presented with a chief concern of hair loss on the scalp that had been present since 12 years of age. During her adult life, the scalp lesion remained unchanged with no associated symptoms. Her medical history was remarkable for hypertension and non–insulin-dependent diabetes mellitus. The patient denied any history of seizure disorders, facial paralysis, or neurologic deficits. Physical examination revealed a 13.6-cm linear, alopecic, hyperpigmented plaque extending from the left forehead and temporal scalp to the posterior occipital scalp with notable loss of underlying subcutaneous tissue (Figure). The left temporal scalp and forehead demonstrated pronounced atrophy with overlying telangiectases. Minimal depression was observed in the temporal scalp and forehead. The left cranial bone was easily palpable with no underlying subcutaneous tissue present. Computed tomography of the brain revealed no underlying skull or soft tissue abnormalities. No treatment was initiated, as the lesion was of minimal concern to the patient.

Comment

Etiology and Presentation                                                                                                                                                                                                                                      
En coup de sabre is a rare subtype of linear morphea that involves the frontoparietal scalp and forehead.^7,12,13 It manifests as a solitary, linear, fibrous plaque that involves the skin, underlying muscle, and bone.⁷ Although most cases present with a single lesion, multiple lesions can occur.⁸ The exact etiology of this disease remains to be determined but is characterized by thickening and hardening of the skin secondary to increased collagen production.⁷ The incidence of linear morphea ranges from 0.4 to 2.7 cases per 100,000 individuals and is more prevalent in white patients and women.¹⁴ Linear morphea is commonly found in children. Children are more likely to have linear morphea on the face, which can lead to lifelong disfigurement.² Although the disease peaks in the fourth decade of life for adults, most pediatric cases are diagnosed between 2 and 14 years of age.^14-16

Pathogenesis                                                                                                                                                                                                                                                       
Clinical and histopathological data suggest that a complex interaction among the vasculature, extracellular matrix, and immune system plays a role in the pathogenesis of the disease. Similar to scleroderma, the CD4 helper T cell may be involved in the fibrotic changes that occur within these lesions.¹⁷ Early in the disease process, T_H1 and T_H17 inflammatory pathways predominate. The late fibrotic changes seen in scleroderma are more associated with a shift to the T_H2 inflammatory pathway.¹⁷ Infection with Borrelia burgdorferi has been implicated abroad, but a large-scale study confirming Borrelia as a pathologic factor within morphea lesions has not been completed to date.^18-20 Some authors believe early lesions of ECDS mimic erythema chronica migrans, with the late lesions resembling acrodermatitis chronica atrophicans.²⁰

Histopathology                                                                                                                                                                                                                                     
Histopathologic findings of morphea tend to vary depending on the stage of the disease. The 2 stages of morphea can be differentiated by the degree of inflammation present histologically.^14,21 The early phase of morphea primarily affects the connective and subcutaneous tissue surrounding eccrine sweat glands.^14,21 A dense dermal and subcutaneous perivascular lymphocytic infiltrate with a mixture of lymphocytes, plasma cells, and histiocytes is commonly observed.⁵ Later stages of the disease demonstrate densely packed homogenous collagen with minimal inflammation and loss of eccrine glands and blood vessels.^14,21 The adipose tissue is generally replaced by sclerotic collagen, giving the biopsy a squared-off appearance.^5,14

Management                                                                                                                                                                                                                                                            
En coup de sabre presents a treatment challenge. In active lesions, topical or intralesional corticosteroids are considered treatment of choice.⁵ Methotrexate has proven useful in the treatment of acute and deep forms of linear morphea. A study examining methotrexate in juvenile localized scleroderma, with the majority of patients having the linear subtype, revealed that methotrexate is both efficacious and well tolerated.²² Other reports in the literature reveal efficacy with the use of intravenous corticosteroids and methotrexate combination therapy for treatment of morphea.^23,24 A longitudinal prospective study examining the use of high-dose methotrexate and oral corticosteroids for the treatment of localized scleroderma yielded positive results, with patients showing clinical improvement within 2 months of initiation of combination therapy.²⁵ Other treatments include excimer laser; calcipotriene and tacrolimus; and surgical approaches such as autologous fat grafting, grafting with muscle flaps, and tissue inserts.^21,26-31 In addition, patients can choose to forego therapy, as was the case with our patient.

Conclusion

En coup de sabre is a rare subtype of linear scleroderma that is limited to the ipsilateral scalp and face predominately in children and women. Neurologic involvement is common and should prompt a comprehensive neurologic workup in patients suspected to have ECDS or PHA. Current treatment recommendations include topical, intralesional, and oral corticosteroids; methotrexate; and surgical grafts. Although ECDS is a rare entity, more intensive research is needed on the exact pathophysiology and effective treatment options that focus on improving the cosmetic outcome in these patients. Cosmesis is the primary concern in patients with ECDS and should be managed early and appropriately to prevent long-term psychological sequelae.

Pityriasis amiantacea (PA) is characterized by adherence of hair shafts proximally.¹ It has been associated with dermatologic conditions and rarely with medications. We describe a woman who developed PA following a bone marrow transplant with melphalan conditioning. We also review drug-induced PA and disorders that have been linked to this condition.

Case Report

A 67-year-old woman with a history of multiple myeloma was treated with 7 courses of chemotherapy (cyclophosphamide, bortezomib, prednisone). One month later, the patient underwent a bone marrow transplant with melphalan conditioning due to residual plasma cell myeloma. Following the transplant, she developed complete scalp alopecia. Prior to and following transplant, the patient’s hair care regimen included washing her hair and scalp every other day with over-the-counter “natural” shampoos. During drug-induced alopecia, the hair washing became less frequent.

The patient left the hospital 4 weeks posttransplant; her hair had started to regrow, but its appearance was altered. Posttransplant, the patient was maintained on bortezomib every other week and zoledronate once per month. She continued to develop multiple lesions in the scalp hairs during the following 4 months.

Eight months posttransplant she presented for evaluation of the scalp hair. Clinical examination showed hairs that were entwined together proximally, resulting in matting of the hair (Figure 1). A diagnosis of PA was established based on the clinical examination.

Treatment included mineral oil application to the scalp under occlusion each evening, followed by morning washing with coal tar 0.5%, salicylic acid 6%, or ketoconazole 2% shampoo in a repeating sequential manner. Within 1 month there was complete resolution of the scalp condition (Figure 2).

Comment

Clinical Presentation                                                                                                                                                                                                                                                                                                                                                             
Pityriasis amiantacea is characterized by thick excessive scale of the scalp¹; it was initially described by Alibert² in 1832. He described the gross appearance of the scales as resembling the feathers of young birds, which naturalists dub “amiante” or asbestoslike.^1,2 In 1917, Gougerot³ explored infectious etiologies of this condition by describing cases of impetigo that transitioned into PA.¹ Later, in 1929, Photinos⁴ described fungal origins of PA, giving credence to “tinea amiantacea.”¹ However, more recent analyses failed to isolate fungus.^5-7 As such, pityriasis (scaling) amiantacea is the more appropriate term, as emphasized by Brown⁸ in 1948. The cause of PA remains unclear; it is hypothesized that the condition is a reaction to underlying inflammatory dermatoses, though concurrent bacterial or fungal infection may be present.^5,9

Prevalence                                                                                                                                                                                                                                                                                                                                                                             
Pityriasis amiantacea is considered to be most prevalent in pediatric patients and young adults; it is more common in females.^1,9,10 In a review of 85 PA patients, more than 80% were women (n=69), and the mean age at presentation was 23.8 years. Approximately half of these patients had widespread scalp lesions (n=42); however, focal localized lesions were common.⁹ No hereditary patterns have been described, though 3 pairs of the 10 patients with PA in Ring and Kaplan’s⁷ review were siblings.

Histopathology                                                                                                                                                                                                                                                                                                                                                                 
Submission of hair specimens to histopathology usually is not performed since the diagnosis often is established based on the clinical presentation.⁵ However, submitted specimens have demonstrated spongiosis and parakeratosis along with reduction in the size of the sebaceous glands.^1,9 Additionally, follicular keratosis that surrounds the hair shafts with a sheath of horn is present.⁹ Acanthosis and migration of lymphocytes into the epidermis also have been found.¹ Often, Staphylococcus aureus isolates are detected.^9,13
 

Differential Diagnosis                                                                                                                                                                                                                                                                                                                                                                   
The clinical differential diagnosis of PA includes hair casts,¹¹ pediculosis,¹⁴ and tinea capitis.¹² In PA, thick scales surround hair shafts and thus bind down tufts of hair.⁹ In patients with pediculosis, nits are attached to the hair shaft at an angle and do not entirely envelop the hair shaft.¹⁴ In addition, PA may be complicated by impetiginization; bacteria often are found in the keratin surrounding the hair shaft and represent either normal flora or secondary infection.^1,15 It has been speculated that microbial biofilms from S aureus and Staphylococcus epidermidis promote agglomeration of hair shafts and adherent scale.¹⁶ Bona fide dermatophyte infection of the scalp also may be concurrently present.¹²

Treatment                                                                                                                                                                                                                                                                                                                                                                                     
Our treatment included occlusion with mineral oil to loosen the scales from the scalp in tandem with shampoos traditionally used in patients with seborrheic dermatitis or psoriasis. Timely treatment is important to prevent scarring alopecia.^13,17 Pityriasis amiantacea may be treatment resistant, and there are no specific therapeutic guidelines; rather, therapy should be targeted at the suspected underlying condition.¹⁷ Treatment generally includes keratolytic agents, such as salicylic acid.¹⁸ These agents allow enhanced penetration of other topical agents.¹⁹ Topical antifungal shampoos such as ketoconazole and ciclopirox are recommended,¹⁸ though other topical agents, such as coal tar and zinc pyrithione, also may benefit patients.¹³ Topical corticosteroids may be used if the condition is linked with psoriasis.¹³ Systemic antibiotics are added if S aureus superinfection is suspected.⁹

A single report described successful management of a patient with severe refractory PA who was treated with the tumor necrosis factor (TNF) α inhibitor infliximab.¹³ A 47-year-old woman presented with thick adherent scale on the scalp. She was treated with coal tar for 18 months but showed no improvement; the patient was subsequently prescribed salicylic acid 10%, clobetasol solution, and coal tar shampoo. After 3 months, when no improvement was observed, the patient was offered infliximab but declined. For 6 years the patient was treated with salicylic acid 20%, clobetasol (foam, lotion, shampoo, and solution), and coal tar shampoo without improvement. She then consented to infliximab therapy; after 3 infusions at weeks 0, 2, and 6, she demonstrated notable improvement. The patient was maintained on infliximab every 8 weeks.¹³

Pathogenesis                                                                                                                                                                                                                                                                                                                                                                                 
The pathogenesis of PA has yet to be definitively established, and the condition is usually idiopathic. In addition to bacterial or fungal etiologies,^3,4 PA has been linked to medications (Table 1)^16,20,21 and systemic conditions (Table 2).^{1,3,5,7-10,12,22-25}

A PubMed search of articles indexed for MEDLINE using the search terms amiantacea, bone, drug, hair marrow, malignancy, melphalan, pityriasis, tinea, and transplant yielded 4 patients—2 men and 2 women (including our patient)—with possible drug-induced PA (Table 1)^16,20,21; however, the onset after 2 years of medication (TNF-α inhibitors) or resolution while still receiving the agent (vemurafenib) makes the drug-induced linkage weak. The patients ranged in age from 21 to 67 years, with the median age being 37.5 years. Medications included melphalan, TNF-α inhibitors (adalimumab, infliximab),^16,21 and vemurafenib²⁰; it is interesting that infliximab was the medication associated with eliciting PA in 1 patient yet was an effective therapy in another patient with treatment-resistant PA. The onset of PA occurred between 1 month (melphalan) and 24 months (TNF-α inhibitors) after drug initiation. The patients’ associated diseases included Crohn disease,^16,21 metastatic melanoma,²⁰ and multiple myeloma.

Other conditions have been described in patients with PA (Table 2). Indeed, PA may be a manifestation of an underlying inflammatory skin disease.⁹ In addition to dermatologic conditions, procedures or malignancy may be associated with the disease, as demonstrated in our patient. Most commonly, PA is seen in association with psoriasis and seborrheic dermatitis; atopic dermatitis, bacterial infection, fungal infection, lichen planus, and neurodermatitis also have been associated with PA.^{1,3,5,7-10,12,18,22-25}

Conclusion

Pityriasis amiantacea is a benign condition affecting the scalp hair. Albeit uncommon, it may appear in patients treated with medications such as melphalan, TNF-α inhibitors, and vemurafenib. In addition, it has been described in individuals with dermatologic conditions, systemic procedures, or underlying malignancy. Our patient developed PA following a bone marrow transplant after receiving conditioning with melphalan.

Pityriasis amiantacea (PA) is characterized by adherence of hair shafts proximally.¹ It has been associated with dermatologic conditions and rarely with medications. We describe a woman who developed PA following a bone marrow transplant with melphalan conditioning. We also review drug-induced PA and disorders that have been linked to this condition.

Case Report

A 67-year-old woman with a history of multiple myeloma was treated with 7 courses of chemotherapy (cyclophosphamide, bortezomib, prednisone). One month later, the patient underwent a bone marrow transplant with melphalan conditioning due to residual plasma cell myeloma. Following the transplant, she developed complete scalp alopecia. Prior to and following transplant, the patient’s hair care regimen included washing her hair and scalp every other day with over-the-counter “natural” shampoos. During drug-induced alopecia, the hair washing became less frequent.

The patient left the hospital 4 weeks posttransplant; her hair had started to regrow, but its appearance was altered. Posttransplant, the patient was maintained on bortezomib every other week and zoledronate once per month. She continued to develop multiple lesions in the scalp hairs during the following 4 months.

Eight months posttransplant she presented for evaluation of the scalp hair. Clinical examination showed hairs that were entwined together proximally, resulting in matting of the hair (Figure 1). A diagnosis of PA was established based on the clinical examination.

Treatment included mineral oil application to the scalp under occlusion each evening, followed by morning washing with coal tar 0.5%, salicylic acid 6%, or ketoconazole 2% shampoo in a repeating sequential manner. Within 1 month there was complete resolution of the scalp condition (Figure 2).

Comment

Clinical Presentation                                                                                                                                                                                                                                                                                                                                                             
Pityriasis amiantacea is characterized by thick excessive scale of the scalp¹; it was initially described by Alibert² in 1832. He described the gross appearance of the scales as resembling the feathers of young birds, which naturalists dub “amiante” or asbestoslike.^1,2 In 1917, Gougerot³ explored infectious etiologies of this condition by describing cases of impetigo that transitioned into PA.¹ Later, in 1929, Photinos⁴ described fungal origins of PA, giving credence to “tinea amiantacea.”¹ However, more recent analyses failed to isolate fungus.^5-7 As such, pityriasis (scaling) amiantacea is the more appropriate term, as emphasized by Brown⁸ in 1948. The cause of PA remains unclear; it is hypothesized that the condition is a reaction to underlying inflammatory dermatoses, though concurrent bacterial or fungal infection may be present.^5,9

Prevalence                                                                                                                                                                                                                                                                                                                                                                             
Pityriasis amiantacea is considered to be most prevalent in pediatric patients and young adults; it is more common in females.^1,9,10 In a review of 85 PA patients, more than 80% were women (n=69), and the mean age at presentation was 23.8 years. Approximately half of these patients had widespread scalp lesions (n=42); however, focal localized lesions were common.⁹ No hereditary patterns have been described, though 3 pairs of the 10 patients with PA in Ring and Kaplan’s⁷ review were siblings.

Histopathology                                                                                                                                                                                                                                                                                                                                                                 
Submission of hair specimens to histopathology usually is not performed since the diagnosis often is established based on the clinical presentation.⁵ However, submitted specimens have demonstrated spongiosis and parakeratosis along with reduction in the size of the sebaceous glands.^1,9 Additionally, follicular keratosis that surrounds the hair shafts with a sheath of horn is present.⁹ Acanthosis and migration of lymphocytes into the epidermis also have been found.¹ Often, Staphylococcus aureus isolates are detected.^9,13
 

Differential Diagnosis                                                                                                                                                                                                                                                                                                                                                                   
The clinical differential diagnosis of PA includes hair casts,¹¹ pediculosis,¹⁴ and tinea capitis.¹² In PA, thick scales surround hair shafts and thus bind down tufts of hair.⁹ In patients with pediculosis, nits are attached to the hair shaft at an angle and do not entirely envelop the hair shaft.¹⁴ In addition, PA may be complicated by impetiginization; bacteria often are found in the keratin surrounding the hair shaft and represent either normal flora or secondary infection.^1,15 It has been speculated that microbial biofilms from S aureus and Staphylococcus epidermidis promote agglomeration of hair shafts and adherent scale.¹⁶ Bona fide dermatophyte infection of the scalp also may be concurrently present.¹²

Treatment                                                                                                                                                                                                                                                                                                                                                                                     
Our treatment included occlusion with mineral oil to loosen the scales from the scalp in tandem with shampoos traditionally used in patients with seborrheic dermatitis or psoriasis. Timely treatment is important to prevent scarring alopecia.^13,17 Pityriasis amiantacea may be treatment resistant, and there are no specific therapeutic guidelines; rather, therapy should be targeted at the suspected underlying condition.¹⁷ Treatment generally includes keratolytic agents, such as salicylic acid.¹⁸ These agents allow enhanced penetration of other topical agents.¹⁹ Topical antifungal shampoos such as ketoconazole and ciclopirox are recommended,¹⁸ though other topical agents, such as coal tar and zinc pyrithione, also may benefit patients.¹³ Topical corticosteroids may be used if the condition is linked with psoriasis.¹³ Systemic antibiotics are added if S aureus superinfection is suspected.⁹

A single report described successful management of a patient with severe refractory PA who was treated with the tumor necrosis factor (TNF) α inhibitor infliximab.¹³ A 47-year-old woman presented with thick adherent scale on the scalp. She was treated with coal tar for 18 months but showed no improvement; the patient was subsequently prescribed salicylic acid 10%, clobetasol solution, and coal tar shampoo. After 3 months, when no improvement was observed, the patient was offered infliximab but declined. For 6 years the patient was treated with salicylic acid 20%, clobetasol (foam, lotion, shampoo, and solution), and coal tar shampoo without improvement. She then consented to infliximab therapy; after 3 infusions at weeks 0, 2, and 6, she demonstrated notable improvement. The patient was maintained on infliximab every 8 weeks.¹³

Pathogenesis                                                                                                                                                                                                                                                                                                                                                                                 
The pathogenesis of PA has yet to be definitively established, and the condition is usually idiopathic. In addition to bacterial or fungal etiologies,^3,4 PA has been linked to medications (Table 1)^16,20,21 and systemic conditions (Table 2).^{1,3,5,7-10,12,22-25}

A PubMed search of articles indexed for MEDLINE using the search terms amiantacea, bone, drug, hair marrow, malignancy, melphalan, pityriasis, tinea, and transplant yielded 4 patients—2 men and 2 women (including our patient)—with possible drug-induced PA (Table 1)^16,20,21; however, the onset after 2 years of medication (TNF-α inhibitors) or resolution while still receiving the agent (vemurafenib) makes the drug-induced linkage weak. The patients ranged in age from 21 to 67 years, with the median age being 37.5 years. Medications included melphalan, TNF-α inhibitors (adalimumab, infliximab),^16,21 and vemurafenib²⁰; it is interesting that infliximab was the medication associated with eliciting PA in 1 patient yet was an effective therapy in another patient with treatment-resistant PA. The onset of PA occurred between 1 month (melphalan) and 24 months (TNF-α inhibitors) after drug initiation. The patients’ associated diseases included Crohn disease,^16,21 metastatic melanoma,²⁰ and multiple myeloma.

Other conditions have been described in patients with PA (Table 2). Indeed, PA may be a manifestation of an underlying inflammatory skin disease.⁹ In addition to dermatologic conditions, procedures or malignancy may be associated with the disease, as demonstrated in our patient. Most commonly, PA is seen in association with psoriasis and seborrheic dermatitis; atopic dermatitis, bacterial infection, fungal infection, lichen planus, and neurodermatitis also have been associated with PA.^{1,3,5,7-10,12,18,22-25}

Conclusion

Pityriasis amiantacea is a benign condition affecting the scalp hair. Albeit uncommon, it may appear in patients treated with medications such as melphalan, TNF-α inhibitors, and vemurafenib. In addition, it has been described in individuals with dermatologic conditions, systemic procedures, or underlying malignancy. Our patient developed PA following a bone marrow transplant after receiving conditioning with melphalan.

Pityriasis amiantacea (PA) is characterized by adherence of hair shafts proximally.¹ It has been associated with dermatologic conditions and rarely with medications. We describe a woman who developed PA following a bone marrow transplant with melphalan conditioning. We also review drug-induced PA and disorders that have been linked to this condition.

Case Report

A 67-year-old woman with a history of multiple myeloma was treated with 7 courses of chemotherapy (cyclophosphamide, bortezomib, prednisone). One month later, the patient underwent a bone marrow transplant with melphalan conditioning due to residual plasma cell myeloma. Following the transplant, she developed complete scalp alopecia. Prior to and following transplant, the patient’s hair care regimen included washing her hair and scalp every other day with over-the-counter “natural” shampoos. During drug-induced alopecia, the hair washing became less frequent.

The patient left the hospital 4 weeks posttransplant; her hair had started to regrow, but its appearance was altered. Posttransplant, the patient was maintained on bortezomib every other week and zoledronate once per month. She continued to develop multiple lesions in the scalp hairs during the following 4 months.

Eight months posttransplant she presented for evaluation of the scalp hair. Clinical examination showed hairs that were entwined together proximally, resulting in matting of the hair (Figure 1). A diagnosis of PA was established based on the clinical examination.

Treatment included mineral oil application to the scalp under occlusion each evening, followed by morning washing with coal tar 0.5%, salicylic acid 6%, or ketoconazole 2% shampoo in a repeating sequential manner. Within 1 month there was complete resolution of the scalp condition (Figure 2).

Comment

Clinical Presentation                                                                                                                                                                                                                                                                                                                                                             
Pityriasis amiantacea is characterized by thick excessive scale of the scalp¹; it was initially described by Alibert² in 1832. He described the gross appearance of the scales as resembling the feathers of young birds, which naturalists dub “amiante” or asbestoslike.^1,2 In 1917, Gougerot³ explored infectious etiologies of this condition by describing cases of impetigo that transitioned into PA.¹ Later, in 1929, Photinos⁴ described fungal origins of PA, giving credence to “tinea amiantacea.”¹ However, more recent analyses failed to isolate fungus.^5-7 As such, pityriasis (scaling) amiantacea is the more appropriate term, as emphasized by Brown⁸ in 1948. The cause of PA remains unclear; it is hypothesized that the condition is a reaction to underlying inflammatory dermatoses, though concurrent bacterial or fungal infection may be present.^5,9

Prevalence                                                                                                                                                                                                                                                                                                                                                                             
Pityriasis amiantacea is considered to be most prevalent in pediatric patients and young adults; it is more common in females.^1,9,10 In a review of 85 PA patients, more than 80% were women (n=69), and the mean age at presentation was 23.8 years. Approximately half of these patients had widespread scalp lesions (n=42); however, focal localized lesions were common.⁹ No hereditary patterns have been described, though 3 pairs of the 10 patients with PA in Ring and Kaplan’s⁷ review were siblings.

Histopathology                                                                                                                                                                                                                                                                                                                                                                 
Submission of hair specimens to histopathology usually is not performed since the diagnosis often is established based on the clinical presentation.⁵ However, submitted specimens have demonstrated spongiosis and parakeratosis along with reduction in the size of the sebaceous glands.^1,9 Additionally, follicular keratosis that surrounds the hair shafts with a sheath of horn is present.⁹ Acanthosis and migration of lymphocytes into the epidermis also have been found.¹ Often, Staphylococcus aureus isolates are detected.^9,13
 

Differential Diagnosis                                                                                                                                                                                                                                                                                                                                                                   
The clinical differential diagnosis of PA includes hair casts,¹¹ pediculosis,¹⁴ and tinea capitis.¹² In PA, thick scales surround hair shafts and thus bind down tufts of hair.⁹ In patients with pediculosis, nits are attached to the hair shaft at an angle and do not entirely envelop the hair shaft.¹⁴ In addition, PA may be complicated by impetiginization; bacteria often are found in the keratin surrounding the hair shaft and represent either normal flora or secondary infection.^1,15 It has been speculated that microbial biofilms from S aureus and Staphylococcus epidermidis promote agglomeration of hair shafts and adherent scale.¹⁶ Bona fide dermatophyte infection of the scalp also may be concurrently present.¹²

Treatment                                                                                                                                                                                                                                                                                                                                                                                     
Our treatment included occlusion with mineral oil to loosen the scales from the scalp in tandem with shampoos traditionally used in patients with seborrheic dermatitis or psoriasis. Timely treatment is important to prevent scarring alopecia.^13,17 Pityriasis amiantacea may be treatment resistant, and there are no specific therapeutic guidelines; rather, therapy should be targeted at the suspected underlying condition.¹⁷ Treatment generally includes keratolytic agents, such as salicylic acid.¹⁸ These agents allow enhanced penetration of other topical agents.¹⁹ Topical antifungal shampoos such as ketoconazole and ciclopirox are recommended,¹⁸ though other topical agents, such as coal tar and zinc pyrithione, also may benefit patients.¹³ Topical corticosteroids may be used if the condition is linked with psoriasis.¹³ Systemic antibiotics are added if S aureus superinfection is suspected.⁹

A single report described successful management of a patient with severe refractory PA who was treated with the tumor necrosis factor (TNF) α inhibitor infliximab.¹³ A 47-year-old woman presented with thick adherent scale on the scalp. She was treated with coal tar for 18 months but showed no improvement; the patient was subsequently prescribed salicylic acid 10%, clobetasol solution, and coal tar shampoo. After 3 months, when no improvement was observed, the patient was offered infliximab but declined. For 6 years the patient was treated with salicylic acid 20%, clobetasol (foam, lotion, shampoo, and solution), and coal tar shampoo without improvement. She then consented to infliximab therapy; after 3 infusions at weeks 0, 2, and 6, she demonstrated notable improvement. The patient was maintained on infliximab every 8 weeks.¹³

Pathogenesis                                                                                                                                                                                                                                                                                                                                                                                 
The pathogenesis of PA has yet to be definitively established, and the condition is usually idiopathic. In addition to bacterial or fungal etiologies,^3,4 PA has been linked to medications (Table 1)^16,20,21 and systemic conditions (Table 2).^{1,3,5,7-10,12,22-25}

A PubMed search of articles indexed for MEDLINE using the search terms amiantacea, bone, drug, hair marrow, malignancy, melphalan, pityriasis, tinea, and transplant yielded 4 patients—2 men and 2 women (including our patient)—with possible drug-induced PA (Table 1)^16,20,21; however, the onset after 2 years of medication (TNF-α inhibitors) or resolution while still receiving the agent (vemurafenib) makes the drug-induced linkage weak. The patients ranged in age from 21 to 67 years, with the median age being 37.5 years. Medications included melphalan, TNF-α inhibitors (adalimumab, infliximab),^16,21 and vemurafenib²⁰; it is interesting that infliximab was the medication associated with eliciting PA in 1 patient yet was an effective therapy in another patient with treatment-resistant PA. The onset of PA occurred between 1 month (melphalan) and 24 months (TNF-α inhibitors) after drug initiation. The patients’ associated diseases included Crohn disease,^16,21 metastatic melanoma,²⁰ and multiple myeloma.

Other conditions have been described in patients with PA (Table 2). Indeed, PA may be a manifestation of an underlying inflammatory skin disease.⁹ In addition to dermatologic conditions, procedures or malignancy may be associated with the disease, as demonstrated in our patient. Most commonly, PA is seen in association with psoriasis and seborrheic dermatitis; atopic dermatitis, bacterial infection, fungal infection, lichen planus, and neurodermatitis also have been associated with PA.^{1,3,5,7-10,12,18,22-25}

Conclusion

Pityriasis amiantacea is a benign condition affecting the scalp hair. Albeit uncommon, it may appear in patients treated with medications such as melphalan, TNF-α inhibitors, and vemurafenib. In addition, it has been described in individuals with dermatologic conditions, systemic procedures, or underlying malignancy. Our patient developed PA following a bone marrow transplant after receiving conditioning with melphalan.

Abstract

Primary heart sarcomas are rare and represent 20% of all primary cardiac tumors. Symptoms depend on which chambers and cardiac structures are involved. Angiosarcoma is one of the most common and the most aggressive types of primary heart sarcomas. Typically, these tumors are found in the right atrium, however, cardiac angiosarcomas may involve any part of the heart. Most of these tumors are diagnosed in advanced stages and the patient prognosis is poor. Most tumors are diagnosed using echocardiography. Computed tomography (CT) and magnetic resonance imaging (MRI) provide useful information on tumor size and location for planning surgery, which is the only treatment shown to increase survival.  We present the case of a 69-year-old woman who presented to the emergency department with hypotension, dyspnea and progressive shortness of breath. After adequate resuscitation, a cardiac mass was identified and surgery was successfully performed. Pathology confirmed a grade 2 primary heart angiosarcoma. Following surgery, the patient was admitted to the intensive care unit and later died secondary to multi-organ system failure.

Introduction

Primary heart angiosarcoma is an aggressive and usually fatal cardiac neoplasm (1). Angiosarcomas can originate at any location in the heart (2, 3), but these tumors typically reside in the right atrium and frequently cause nonspecific symptoms such as dyspnea, cough, heart failure, and arrhythmias. (2) Surgery followed by chemotherapy is the typical approach to these tumors. (4)

We present the case of a 69-year-old woman who presented to the emergency department with hypotension and severe dyspnea.

Case Report

The patient was a 69-year-old woman with a medical history of diabetes. A week before seeking care in the emergency department, she experienced a general feeling of unwellness, dyspnea, and mild respiratory distress. She reported these symptoms had become more and more severe in the last 24 hours and were accompanied by acute chest pain and progressive shortness of breath.

On clinical examination, the patient was hypotensive, had tachypnea and tachycardia, and was hypoxic. Cardiac auscultation detected a systolic murmur in the apex, and auscultation of the lungs revealed crackles and rales, especially at the bases of the lungs. The remainder of her clinical examination was unremarkable.  She had sinus tachycardia on an electrocardiogram. A chest X-ray showed a left atrial enlargement along with some patchy opacities in the middle and lower zones of the lungs, along with Kerley B lines suggestive of pulmonary edema.

With these findings, and after adequate resuscitation, a contrast-enhanced computed tomography (CT) scan detected a filling defect in the left atrium suggestive of a large intra-cardiac mass with a thick and hyper-enhanced interatrial septum. Bilateral pleural effusions also were evident, (Figure 1A) hence an echocardiogram was requested and it confirmed the presence of a 30 x 29 x 40 mm lobulated highly mobile mass in the left atrium. 

After a cardiothoracic consultation, cardiac magnetic resonance imaging (MRI) was performed. The findings showed the presence of a 58 x 45 x 6 mm well-circumscribed hyperemic mass on the anterior leaflet of the mitral valve and a second 10 x 10 x 6 mm smaller mass firmly adhered to the posterior leaflet of the mitral valve. 

The patient, who was hypotensive and hypoxic, was admitted to the hospital for surgical treatment.

Following sternotomy and cardiopulmonary bypass, a right atriotomy was performed using a trans-septal approach. The large left atrial mass was firmly adhered to the endocardium at the level of the anterior leaflet of the mitral valve and the interatrial septum. The mass had a grey and whitish appearance with some bluish necrotic patches, (Figure 1B, 2B, 3B). 

The patient had a complicated postoperative course in the Intensive Care Unit (ICU) and needed inotropic support and vasoactive agents. A postop echocardiogram indicated appropriate left ventricle systolic function, nonetheless, the patient persisted in a hypotensive status that caused refractory shock and ultimately provoked severe organ dysfunction that led to the patient’s death.

Discussion

Primary heart sarcomas are extremely rare malignant neoplasms derived from mesenchymal cells, (1) with an incidence ranging from 0.001% to 0.28% at autopsy. 

Cardiac angiosarcomas (CA) account for one-third of all primary heart sarcomas (4) and usually develop as gray-brown masses with hemorrhagic patches in the right atrium of male patients. The tumors are filled with vascular channels and their cells are positive for CD34 and factor VIII. (5)  Left-sided cardiac angiosarcoma can cause heart failure early in the disease process, but the tumors tend to be more circumscribed, less infiltrative, and associated with better overall survival. (6, 7) Most patients are asymptomatic early in their disease, (2) making the diagnosis even more difficult and worsening its already poor prognosis. (1) The preference of cardiac angiosarcomas for the right heart often leads to a presentation with right-sided congestive heart failure. (2) At later stages, symptoms depend on the structures compromised and range from mild dyspnea on exertion to cardiogenic shock. (8) Cardiac angiosarcomas tend to have a notable intracavitary element, and in some cases may intermittently compromise a cardiac valve, thereby simulating a stenosis or regurgitation. (2, 7)

Our patient presented with acute cardiac failure, pulmonary edema and severe valve dysfunction due to a mass in the left atrium. The tumor had a vascular supply and showed positivity for CD34.

Most patients with cardiac angiosarcoma have metastases, typically to the lung, at diagnosis. (1) Several decades ago, cardiac angiosarcoma was mainly diagnosed postmortem. (1) Now, it can be suspected when cardiomegaly or pleural effusions are seen on chest x-rays (8). Echocardiography is the most useful diagnostic tool, (2) however, CT and MRI can provide useful information on tumor size, invasion and localization. (2, 9) This imaging combination generally provides an excellent anatomic description for preoperative planning. (1, 9)

In our patient, progressive dyspnea was the main symptom and after a prompt evaluation an intracardiac mass was identified as the cause of severe cardiac dysfunction. Because of this finding and the clinical condition of the patient, surgery was planned.

Complete resection of the tumor is the treatment of choice, and is the only therapy currently seen to influence survival. (8) But because of the highly aggressive behavior and a high incidence of systemic metastases with cardiac angiosarcomas, a complete surgical resection is often hampered. (1)  Cardiac angiosarcoma carries a grim prognosis as these tumors are universally fatal with a mean survival time of several months after initial presentation even after successful surgery. (2) Chemotherapy is recommended after surgery, even when clear surgical margins are obtained because of the high probability of missed microscopic disease. (1, 2)

High clinical suspicion together with an appropriate history, a thorough physical examination, and precise complementary tests are vital for timely diagnosis and proper treatment.

Authors and Affiliations
Santiago A. Endara: Department of General Surgery, Division of Cardiothoracic Surgery, Hospital Metropolitano, Quito, Ecuador, MD
Gerardo A. Dávalos: Department of General Surgery, Division of Cardiothoracic Surgery, Hospital Metropolitano, Quito, Ecuador, MD
Patricia M. Pontón: Hospital Metropolitano, Quito, Ecuador. Department of Internal Medicine Division of Pathology, MD
Gabriel A. Molina: Pontificia Universidad Católica del Ecuador (PUCE), Quito, Ecuador. PGY4 General Surgery Resident, MD
Daniel L. Mogrovejo: Pontificia Universidad Católica del Ecuador (PUCE), Quito, Ecuador. PGY1 General Surgery Resident, MD
Corresponding Author Info:
Santiago A. Endara, Hospital Metropolitano, Av. Mariana de Jesus Oe 7/47 y Conclina, Edificio Diagnostico 2000 tercer piso 3/3, Quito, Ecuador, + 593 9 98416157
Email: drsantiagoendara@gmail.com

Abstract

Primary heart sarcomas are rare and represent 20% of all primary cardiac tumors. Symptoms depend on which chambers and cardiac structures are involved. Angiosarcoma is one of the most common and the most aggressive types of primary heart sarcomas. Typically, these tumors are found in the right atrium, however, cardiac angiosarcomas may involve any part of the heart. Most of these tumors are diagnosed in advanced stages and the patient prognosis is poor. Most tumors are diagnosed using echocardiography. Computed tomography (CT) and magnetic resonance imaging (MRI) provide useful information on tumor size and location for planning surgery, which is the only treatment shown to increase survival.  We present the case of a 69-year-old woman who presented to the emergency department with hypotension, dyspnea and progressive shortness of breath. After adequate resuscitation, a cardiac mass was identified and surgery was successfully performed. Pathology confirmed a grade 2 primary heart angiosarcoma. Following surgery, the patient was admitted to the intensive care unit and later died secondary to multi-organ system failure.

Introduction

Primary heart angiosarcoma is an aggressive and usually fatal cardiac neoplasm (1). Angiosarcomas can originate at any location in the heart (2, 3), but these tumors typically reside in the right atrium and frequently cause nonspecific symptoms such as dyspnea, cough, heart failure, and arrhythmias. (2) Surgery followed by chemotherapy is the typical approach to these tumors. (4)

We present the case of a 69-year-old woman who presented to the emergency department with hypotension and severe dyspnea.

Case Report

The patient was a 69-year-old woman with a medical history of diabetes. A week before seeking care in the emergency department, she experienced a general feeling of unwellness, dyspnea, and mild respiratory distress. She reported these symptoms had become more and more severe in the last 24 hours and were accompanied by acute chest pain and progressive shortness of breath.

On clinical examination, the patient was hypotensive, had tachypnea and tachycardia, and was hypoxic. Cardiac auscultation detected a systolic murmur in the apex, and auscultation of the lungs revealed crackles and rales, especially at the bases of the lungs. The remainder of her clinical examination was unremarkable.  She had sinus tachycardia on an electrocardiogram. A chest X-ray showed a left atrial enlargement along with some patchy opacities in the middle and lower zones of the lungs, along with Kerley B lines suggestive of pulmonary edema.

With these findings, and after adequate resuscitation, a contrast-enhanced computed tomography (CT) scan detected a filling defect in the left atrium suggestive of a large intra-cardiac mass with a thick and hyper-enhanced interatrial septum. Bilateral pleural effusions also were evident, (Figure 1A) hence an echocardiogram was requested and it confirmed the presence of a 30 x 29 x 40 mm lobulated highly mobile mass in the left atrium. 

After a cardiothoracic consultation, cardiac magnetic resonance imaging (MRI) was performed. The findings showed the presence of a 58 x 45 x 6 mm well-circumscribed hyperemic mass on the anterior leaflet of the mitral valve and a second 10 x 10 x 6 mm smaller mass firmly adhered to the posterior leaflet of the mitral valve. 

The patient, who was hypotensive and hypoxic, was admitted to the hospital for surgical treatment.

Following sternotomy and cardiopulmonary bypass, a right atriotomy was performed using a trans-septal approach. The large left atrial mass was firmly adhered to the endocardium at the level of the anterior leaflet of the mitral valve and the interatrial septum. The mass had a grey and whitish appearance with some bluish necrotic patches, (Figure 1B, 2B, 3B). 

The patient had a complicated postoperative course in the Intensive Care Unit (ICU) and needed inotropic support and vasoactive agents. A postop echocardiogram indicated appropriate left ventricle systolic function, nonetheless, the patient persisted in a hypotensive status that caused refractory shock and ultimately provoked severe organ dysfunction that led to the patient’s death.

Discussion

Primary heart sarcomas are extremely rare malignant neoplasms derived from mesenchymal cells, (1) with an incidence ranging from 0.001% to 0.28% at autopsy. 

Cardiac angiosarcomas (CA) account for one-third of all primary heart sarcomas (4) and usually develop as gray-brown masses with hemorrhagic patches in the right atrium of male patients. The tumors are filled with vascular channels and their cells are positive for CD34 and factor VIII. (5)  Left-sided cardiac angiosarcoma can cause heart failure early in the disease process, but the tumors tend to be more circumscribed, less infiltrative, and associated with better overall survival. (6, 7) Most patients are asymptomatic early in their disease, (2) making the diagnosis even more difficult and worsening its already poor prognosis. (1) The preference of cardiac angiosarcomas for the right heart often leads to a presentation with right-sided congestive heart failure. (2) At later stages, symptoms depend on the structures compromised and range from mild dyspnea on exertion to cardiogenic shock. (8) Cardiac angiosarcomas tend to have a notable intracavitary element, and in some cases may intermittently compromise a cardiac valve, thereby simulating a stenosis or regurgitation. (2, 7)

Our patient presented with acute cardiac failure, pulmonary edema and severe valve dysfunction due to a mass in the left atrium. The tumor had a vascular supply and showed positivity for CD34.

Most patients with cardiac angiosarcoma have metastases, typically to the lung, at diagnosis. (1) Several decades ago, cardiac angiosarcoma was mainly diagnosed postmortem. (1) Now, it can be suspected when cardiomegaly or pleural effusions are seen on chest x-rays (8). Echocardiography is the most useful diagnostic tool, (2) however, CT and MRI can provide useful information on tumor size, invasion and localization. (2, 9) This imaging combination generally provides an excellent anatomic description for preoperative planning. (1, 9)

In our patient, progressive dyspnea was the main symptom and after a prompt evaluation an intracardiac mass was identified as the cause of severe cardiac dysfunction. Because of this finding and the clinical condition of the patient, surgery was planned.

Complete resection of the tumor is the treatment of choice, and is the only therapy currently seen to influence survival. (8) But because of the highly aggressive behavior and a high incidence of systemic metastases with cardiac angiosarcomas, a complete surgical resection is often hampered. (1)  Cardiac angiosarcoma carries a grim prognosis as these tumors are universally fatal with a mean survival time of several months after initial presentation even after successful surgery. (2) Chemotherapy is recommended after surgery, even when clear surgical margins are obtained because of the high probability of missed microscopic disease. (1, 2)

High clinical suspicion together with an appropriate history, a thorough physical examination, and precise complementary tests are vital for timely diagnosis and proper treatment.

Authors and Affiliations
Santiago A. Endara: Department of General Surgery, Division of Cardiothoracic Surgery, Hospital Metropolitano, Quito, Ecuador, MD
Gerardo A. Dávalos: Department of General Surgery, Division of Cardiothoracic Surgery, Hospital Metropolitano, Quito, Ecuador, MD
Patricia M. Pontón: Hospital Metropolitano, Quito, Ecuador. Department of Internal Medicine Division of Pathology, MD
Gabriel A. Molina: Pontificia Universidad Católica del Ecuador (PUCE), Quito, Ecuador. PGY4 General Surgery Resident, MD
Daniel L. Mogrovejo: Pontificia Universidad Católica del Ecuador (PUCE), Quito, Ecuador. PGY1 General Surgery Resident, MD
Corresponding Author Info:
Santiago A. Endara, Hospital Metropolitano, Av. Mariana de Jesus Oe 7/47 y Conclina, Edificio Diagnostico 2000 tercer piso 3/3, Quito, Ecuador, + 593 9 98416157
Email: drsantiagoendara@gmail.com

Abstract

Primary heart sarcomas are rare and represent 20% of all primary cardiac tumors. Symptoms depend on which chambers and cardiac structures are involved. Angiosarcoma is one of the most common and the most aggressive types of primary heart sarcomas. Typically, these tumors are found in the right atrium, however, cardiac angiosarcomas may involve any part of the heart. Most of these tumors are diagnosed in advanced stages and the patient prognosis is poor. Most tumors are diagnosed using echocardiography. Computed tomography (CT) and magnetic resonance imaging (MRI) provide useful information on tumor size and location for planning surgery, which is the only treatment shown to increase survival.  We present the case of a 69-year-old woman who presented to the emergency department with hypotension, dyspnea and progressive shortness of breath. After adequate resuscitation, a cardiac mass was identified and surgery was successfully performed. Pathology confirmed a grade 2 primary heart angiosarcoma. Following surgery, the patient was admitted to the intensive care unit and later died secondary to multi-organ system failure.

Introduction

Primary heart angiosarcoma is an aggressive and usually fatal cardiac neoplasm (1). Angiosarcomas can originate at any location in the heart (2, 3), but these tumors typically reside in the right atrium and frequently cause nonspecific symptoms such as dyspnea, cough, heart failure, and arrhythmias. (2) Surgery followed by chemotherapy is the typical approach to these tumors. (4)

We present the case of a 69-year-old woman who presented to the emergency department with hypotension and severe dyspnea.

Case Report

The patient was a 69-year-old woman with a medical history of diabetes. A week before seeking care in the emergency department, she experienced a general feeling of unwellness, dyspnea, and mild respiratory distress. She reported these symptoms had become more and more severe in the last 24 hours and were accompanied by acute chest pain and progressive shortness of breath.

On clinical examination, the patient was hypotensive, had tachypnea and tachycardia, and was hypoxic. Cardiac auscultation detected a systolic murmur in the apex, and auscultation of the lungs revealed crackles and rales, especially at the bases of the lungs. The remainder of her clinical examination was unremarkable.  She had sinus tachycardia on an electrocardiogram. A chest X-ray showed a left atrial enlargement along with some patchy opacities in the middle and lower zones of the lungs, along with Kerley B lines suggestive of pulmonary edema.

With these findings, and after adequate resuscitation, a contrast-enhanced computed tomography (CT) scan detected a filling defect in the left atrium suggestive of a large intra-cardiac mass with a thick and hyper-enhanced interatrial septum. Bilateral pleural effusions also were evident, (Figure 1A) hence an echocardiogram was requested and it confirmed the presence of a 30 x 29 x 40 mm lobulated highly mobile mass in the left atrium. 

After a cardiothoracic consultation, cardiac magnetic resonance imaging (MRI) was performed. The findings showed the presence of a 58 x 45 x 6 mm well-circumscribed hyperemic mass on the anterior leaflet of the mitral valve and a second 10 x 10 x 6 mm smaller mass firmly adhered to the posterior leaflet of the mitral valve. 

The patient, who was hypotensive and hypoxic, was admitted to the hospital for surgical treatment.

Following sternotomy and cardiopulmonary bypass, a right atriotomy was performed using a trans-septal approach. The large left atrial mass was firmly adhered to the endocardium at the level of the anterior leaflet of the mitral valve and the interatrial septum. The mass had a grey and whitish appearance with some bluish necrotic patches, (Figure 1B, 2B, 3B). 

The patient had a complicated postoperative course in the Intensive Care Unit (ICU) and needed inotropic support and vasoactive agents. A postop echocardiogram indicated appropriate left ventricle systolic function, nonetheless, the patient persisted in a hypotensive status that caused refractory shock and ultimately provoked severe organ dysfunction that led to the patient’s death.

Discussion

Primary heart sarcomas are extremely rare malignant neoplasms derived from mesenchymal cells, (1) with an incidence ranging from 0.001% to 0.28% at autopsy. 

Cardiac angiosarcomas (CA) account for one-third of all primary heart sarcomas (4) and usually develop as gray-brown masses with hemorrhagic patches in the right atrium of male patients. The tumors are filled with vascular channels and their cells are positive for CD34 and factor VIII. (5)  Left-sided cardiac angiosarcoma can cause heart failure early in the disease process, but the tumors tend to be more circumscribed, less infiltrative, and associated with better overall survival. (6, 7) Most patients are asymptomatic early in their disease, (2) making the diagnosis even more difficult and worsening its already poor prognosis. (1) The preference of cardiac angiosarcomas for the right heart often leads to a presentation with right-sided congestive heart failure. (2) At later stages, symptoms depend on the structures compromised and range from mild dyspnea on exertion to cardiogenic shock. (8) Cardiac angiosarcomas tend to have a notable intracavitary element, and in some cases may intermittently compromise a cardiac valve, thereby simulating a stenosis or regurgitation. (2, 7)

Our patient presented with acute cardiac failure, pulmonary edema and severe valve dysfunction due to a mass in the left atrium. The tumor had a vascular supply and showed positivity for CD34.

Most patients with cardiac angiosarcoma have metastases, typically to the lung, at diagnosis. (1) Several decades ago, cardiac angiosarcoma was mainly diagnosed postmortem. (1) Now, it can be suspected when cardiomegaly or pleural effusions are seen on chest x-rays (8). Echocardiography is the most useful diagnostic tool, (2) however, CT and MRI can provide useful information on tumor size, invasion and localization. (2, 9) This imaging combination generally provides an excellent anatomic description for preoperative planning. (1, 9)

In our patient, progressive dyspnea was the main symptom and after a prompt evaluation an intracardiac mass was identified as the cause of severe cardiac dysfunction. Because of this finding and the clinical condition of the patient, surgery was planned.

Complete resection of the tumor is the treatment of choice, and is the only therapy currently seen to influence survival. (8) But because of the highly aggressive behavior and a high incidence of systemic metastases with cardiac angiosarcomas, a complete surgical resection is often hampered. (1)  Cardiac angiosarcoma carries a grim prognosis as these tumors are universally fatal with a mean survival time of several months after initial presentation even after successful surgery. (2) Chemotherapy is recommended after surgery, even when clear surgical margins are obtained because of the high probability of missed microscopic disease. (1, 2)

High clinical suspicion together with an appropriate history, a thorough physical examination, and precise complementary tests are vital for timely diagnosis and proper treatment.

Authors and Affiliations
Santiago A. Endara: Department of General Surgery, Division of Cardiothoracic Surgery, Hospital Metropolitano, Quito, Ecuador, MD
Gerardo A. Dávalos: Department of General Surgery, Division of Cardiothoracic Surgery, Hospital Metropolitano, Quito, Ecuador, MD
Patricia M. Pontón: Hospital Metropolitano, Quito, Ecuador. Department of Internal Medicine Division of Pathology, MD
Gabriel A. Molina: Pontificia Universidad Católica del Ecuador (PUCE), Quito, Ecuador. PGY4 General Surgery Resident, MD
Daniel L. Mogrovejo: Pontificia Universidad Católica del Ecuador (PUCE), Quito, Ecuador. PGY1 General Surgery Resident, MD
Corresponding Author Info:
Santiago A. Endara, Hospital Metropolitano, Av. Mariana de Jesus Oe 7/47 y Conclina, Edificio Diagnostico 2000 tercer piso 3/3, Quito, Ecuador, + 593 9 98416157
Email: drsantiagoendara@gmail.com

To the Editor:

Although relatively uncommon, hypersensitivity reactions to tattoo pigment are on the rise due to the increasing popularity and prevalence of tattoos.¹ Multiple adverse events have been described in association with tattoos, including inflammatory, infectious, and neoplastic responses.² An id reaction (also known as autoeczematization or autosensitization) develops distant to an initial site of infection or sensitization. We describe a unique case of an id reaction and subsequent development of prurigo nodules associated with contact allergy to red tattoo ink.

A 40-year-old woman was referred to the New York University Skin and Cancer Unit (New York, New York) for evaluation of a pruritic eruption arising on and near sites of tattooed skin on the right foot and right upper arm of 8 months’ duration. The patient reported that she had obtained a polychromatic tattoo on the right dorsal foot 9 months prior to the current presentation. Approximately 1 month later, she developed pruritic papulonodular lesions localized to the red-pigmented areas of the tattoo. Concomitantly, the patient developed a similar eruption confined to areas of red pigment in a polychromatic tattoo on the right upper arm that she had obtained 10 years prior. She was treated with intralesional triamcinolone to several of the lesions on the right dorsal foot with some benefit; however, a few days later she developed a generalized, erythematous, pruritic eruption on the back, abdomen, arms, and legs. Her medical history was remarkable only for mild iron-deficiency anemia. She had no known drug allergies or history of atopy and was not taking any medications prior to the onset of the eruption.

Skin examination revealed multiple, well-demarcated, eczematous papulonodules with surrounding erythema confined to the red-pigmented areas of the tattoo on the right dorsal foot, with several similar lesions on the surrounding nontattooed skin (Figure 1). Linear, well-demarcated, eczematous, hyperpigmented plaques also were noted on the red-pigmented areas of the tattoo on the patient’s right upper arm (Figure 2). Eczematous plaques and scattered excoriations were noted on the back, abdomen, flanks, arms, and legs.

Patch testing with the North American Standard Series, metal series, and samples of the red pigments used in the tattoo on the foot were negative. A punch biopsy of a lesion on the dorsal right foot showed a psoriasiform spongiotic dermatitis with eosinophils (Figure 3). Periodic acid–Schiff staining with diastase failed to reveal fungal hyphae. The histologic findings were consistent with allergic contact dermatitis. A punch biopsy of the eczematous reaction on nontattooed skin on the trunk demonstrated a perivascular dermatitis with eosinophils and subtle spongiosis consistent with an id reaction.

The patient was treated with fluocinonide ointment for several months with no effect. Subsequently, she received several short courses of oral prednisone, after which the affected areas of the tattoo on the arm and foot flattened and the id reaction resolved; however, after several months, the red-pigmented areas of the tattoo on the foot again became elevated and pruritic, and the patient developed widespread prurigo nodules on nontattooed skin on the trunk, arms, and legs. She was subsequently referred to a laser specialist for a trial of fractional laser treatment to cautiously remove the red tattoo pigment. After 2 treatments, the pruritus improved and the papular lesions appeared slightly flatter; however, the prurigo nodules remained. The tattoo on the patient’s foot was surgically removed; however, the prurigo nodules remained. Ultimately, the lesions cleared with a several-month course of mycophenolate mofetil.

Systemic allergic reactions to tattoo ink are rare but can cause considerable morbidity. An id reaction, also known as autoeczematization or autosensitization, is a reaction that develops distant to an initial site of infection or sensitization. Although the pathogenesis of this reaction is not certain, it has been hypothesized that autoimmunity to skin antigens might play a role.³ Autologous epidermal cells are thought to become antigenic in the presence of acute inflammation at the primary cutaneous site. These antigenic autologous epidermal cells are postulated to enter the circulation and cause secondary eczematous lesions at distant sites. This proposed mechanism is supported by the development of positive skin reactions to autologous extracts of epidermal scaling in patients with active id reaction.³

Hematogenous dissemination of cytokines has been implicated in id reactions.⁴ Keratinocytes produce cytokines in response to conditions that are known to trigger id reactions.⁵ Epidermal cytokines released from the primary site of sensitization are thought to heighten sensitivity at distant skin areas.⁴ These cytokines regulate both cell-mediated and humoral cutaneous immune responses. Increased levels of activated HLA-DR isotype–positive T cells in patients with active autoeczemization favors a cellular-mediated immune mechanism. The presence of activated antigen-specific T cells also supports the role of allergic contact dermatitis in triggering id reactions.⁶

Allergic contact dermatitis is the most common hypersensitivity reaction to tattoo ink, with red pigments representing the most common cause of tattoo-related allergic contact dermatitis. Historically, cinnabar (mercuric sulfide) has been the most common red pigment to cause allergic contact dermatitis.⁷ More recently, mercury-free organic pigments (eg, azo dyes) have been used in polychromatic tattoos due to their ability to retain color over long periods of time⁸; however, these organic red tattoo pigments also have been implicated in allergic reactions.^8-11 The composition of these new organic red tattoo pigments varies, but chemical analysis has revealed a mixture of aromatic azo compounds (eg, quinacridone),¹⁰ heavy metals (eg, aluminum, lead, cadmium, chromium, cobalt, iron, titanium),^9,12 and intermediate reactive compounds (eg, naphthalene, 2-naphthol, chlorobenzene, benzene).⁸ Allergic contact dermatitis to red tattoo ink is well documented^8,13; however, a PubMed search of articles indexed for MEDLINE using the terms tattoo and dermatitis, tattoo and allergy, tattoo and autosensitization, tattoo and id reaction, and tattoo and autoeczematization yielded only 3 other reports of a concomitant id reaction.^11,14,15

The diagnosis of id reaction associated with allergic contact dermatitis is made on the basis of clinical history, physical examination, and histopathology. Patch testing usually is not positive in cases of tattoo allergy; it is thought that the allergen is a tattoo ink byproduct possibly caused by photoinduced or metabolic change of the tattoo pigment and a haptenization process.^1,8,16 Histologically, variable reaction patterns, including eczematous, lichenoid, granulomatous, and pseudolymphomatous reactions have been reported in association with delayed-type inflammatory reactions to tattoo pigments, but the lichenoid pattern is most commonly observed.⁸

Treatment options for allergic contact dermatitis to tattoo ink include topical, intralesional, and oral steroids; topical calcineurin inhibitors; and surgical excision of the tattoo. Q-switched lasers—ruby, Nd:YAG, and alexandrite—are the gold standard for removing tattoo pigments¹⁷; however, these lasers remove tattoo pigment by selective photothermolysis, resulting in extracellular extravasation of pigment, which can precipitate a heightened immune response that can lead to localized and generalized allergic reactions.¹⁸ Therefore, Q-switched lasers should be avoided in the setting of an allergic reaction to tattoo ink. Fractional ablative laser resurfacing may be a safer alternative for removal of tattoos in the setting of an allergic reaction.¹⁷ Further studies are needed to confirm the safety and efficacy of this modality for allergic tattoo ink removal.^17,18

Our case illustrates a rare cause of id reaction and the subsequent development of prurigo nodules associated with contact allergy to red tattoo ink. We present this case to raise awareness of the potential health and iatrogenic risks associated with tattoo placement. Further investigation of these color additives is warranted to better elucidate ink components responsible for these cutaneous allergic reactions.

Acknowledgments
We would like to thank Vitaly Terushkin, MD (West Orange, New Jersey, and New York, New York), and Arielle Kauvar, MD (New York, New York), for their contributions to the patient’s clinical care.

To the Editor:

Although relatively uncommon, hypersensitivity reactions to tattoo pigment are on the rise due to the increasing popularity and prevalence of tattoos.¹ Multiple adverse events have been described in association with tattoos, including inflammatory, infectious, and neoplastic responses.² An id reaction (also known as autoeczematization or autosensitization) develops distant to an initial site of infection or sensitization. We describe a unique case of an id reaction and subsequent development of prurigo nodules associated with contact allergy to red tattoo ink.

A 40-year-old woman was referred to the New York University Skin and Cancer Unit (New York, New York) for evaluation of a pruritic eruption arising on and near sites of tattooed skin on the right foot and right upper arm of 8 months’ duration. The patient reported that she had obtained a polychromatic tattoo on the right dorsal foot 9 months prior to the current presentation. Approximately 1 month later, she developed pruritic papulonodular lesions localized to the red-pigmented areas of the tattoo. Concomitantly, the patient developed a similar eruption confined to areas of red pigment in a polychromatic tattoo on the right upper arm that she had obtained 10 years prior. She was treated with intralesional triamcinolone to several of the lesions on the right dorsal foot with some benefit; however, a few days later she developed a generalized, erythematous, pruritic eruption on the back, abdomen, arms, and legs. Her medical history was remarkable only for mild iron-deficiency anemia. She had no known drug allergies or history of atopy and was not taking any medications prior to the onset of the eruption.

Skin examination revealed multiple, well-demarcated, eczematous papulonodules with surrounding erythema confined to the red-pigmented areas of the tattoo on the right dorsal foot, with several similar lesions on the surrounding nontattooed skin (Figure 1). Linear, well-demarcated, eczematous, hyperpigmented plaques also were noted on the red-pigmented areas of the tattoo on the patient’s right upper arm (Figure 2). Eczematous plaques and scattered excoriations were noted on the back, abdomen, flanks, arms, and legs.

Patch testing with the North American Standard Series, metal series, and samples of the red pigments used in the tattoo on the foot were negative. A punch biopsy of a lesion on the dorsal right foot showed a psoriasiform spongiotic dermatitis with eosinophils (Figure 3). Periodic acid–Schiff staining with diastase failed to reveal fungal hyphae. The histologic findings were consistent with allergic contact dermatitis. A punch biopsy of the eczematous reaction on nontattooed skin on the trunk demonstrated a perivascular dermatitis with eosinophils and subtle spongiosis consistent with an id reaction.

The patient was treated with fluocinonide ointment for several months with no effect. Subsequently, she received several short courses of oral prednisone, after which the affected areas of the tattoo on the arm and foot flattened and the id reaction resolved; however, after several months, the red-pigmented areas of the tattoo on the foot again became elevated and pruritic, and the patient developed widespread prurigo nodules on nontattooed skin on the trunk, arms, and legs. She was subsequently referred to a laser specialist for a trial of fractional laser treatment to cautiously remove the red tattoo pigment. After 2 treatments, the pruritus improved and the papular lesions appeared slightly flatter; however, the prurigo nodules remained. The tattoo on the patient’s foot was surgically removed; however, the prurigo nodules remained. Ultimately, the lesions cleared with a several-month course of mycophenolate mofetil.

Systemic allergic reactions to tattoo ink are rare but can cause considerable morbidity. An id reaction, also known as autoeczematization or autosensitization, is a reaction that develops distant to an initial site of infection or sensitization. Although the pathogenesis of this reaction is not certain, it has been hypothesized that autoimmunity to skin antigens might play a role.³ Autologous epidermal cells are thought to become antigenic in the presence of acute inflammation at the primary cutaneous site. These antigenic autologous epidermal cells are postulated to enter the circulation and cause secondary eczematous lesions at distant sites. This proposed mechanism is supported by the development of positive skin reactions to autologous extracts of epidermal scaling in patients with active id reaction.³

Hematogenous dissemination of cytokines has been implicated in id reactions.⁴ Keratinocytes produce cytokines in response to conditions that are known to trigger id reactions.⁵ Epidermal cytokines released from the primary site of sensitization are thought to heighten sensitivity at distant skin areas.⁴ These cytokines regulate both cell-mediated and humoral cutaneous immune responses. Increased levels of activated HLA-DR isotype–positive T cells in patients with active autoeczemization favors a cellular-mediated immune mechanism. The presence of activated antigen-specific T cells also supports the role of allergic contact dermatitis in triggering id reactions.⁶

Allergic contact dermatitis is the most common hypersensitivity reaction to tattoo ink, with red pigments representing the most common cause of tattoo-related allergic contact dermatitis. Historically, cinnabar (mercuric sulfide) has been the most common red pigment to cause allergic contact dermatitis.⁷ More recently, mercury-free organic pigments (eg, azo dyes) have been used in polychromatic tattoos due to their ability to retain color over long periods of time⁸; however, these organic red tattoo pigments also have been implicated in allergic reactions.^8-11 The composition of these new organic red tattoo pigments varies, but chemical analysis has revealed a mixture of aromatic azo compounds (eg, quinacridone),¹⁰ heavy metals (eg, aluminum, lead, cadmium, chromium, cobalt, iron, titanium),^9,12 and intermediate reactive compounds (eg, naphthalene, 2-naphthol, chlorobenzene, benzene).⁸ Allergic contact dermatitis to red tattoo ink is well documented^8,13; however, a PubMed search of articles indexed for MEDLINE using the terms tattoo and dermatitis, tattoo and allergy, tattoo and autosensitization, tattoo and id reaction, and tattoo and autoeczematization yielded only 3 other reports of a concomitant id reaction.^11,14,15

The diagnosis of id reaction associated with allergic contact dermatitis is made on the basis of clinical history, physical examination, and histopathology. Patch testing usually is not positive in cases of tattoo allergy; it is thought that the allergen is a tattoo ink byproduct possibly caused by photoinduced or metabolic change of the tattoo pigment and a haptenization process.^1,8,16 Histologically, variable reaction patterns, including eczematous, lichenoid, granulomatous, and pseudolymphomatous reactions have been reported in association with delayed-type inflammatory reactions to tattoo pigments, but the lichenoid pattern is most commonly observed.⁸

Treatment options for allergic contact dermatitis to tattoo ink include topical, intralesional, and oral steroids; topical calcineurin inhibitors; and surgical excision of the tattoo. Q-switched lasers—ruby, Nd:YAG, and alexandrite—are the gold standard for removing tattoo pigments¹⁷; however, these lasers remove tattoo pigment by selective photothermolysis, resulting in extracellular extravasation of pigment, which can precipitate a heightened immune response that can lead to localized and generalized allergic reactions.¹⁸ Therefore, Q-switched lasers should be avoided in the setting of an allergic reaction to tattoo ink. Fractional ablative laser resurfacing may be a safer alternative for removal of tattoos in the setting of an allergic reaction.¹⁷ Further studies are needed to confirm the safety and efficacy of this modality for allergic tattoo ink removal.^17,18

Our case illustrates a rare cause of id reaction and the subsequent development of prurigo nodules associated with contact allergy to red tattoo ink. We present this case to raise awareness of the potential health and iatrogenic risks associated with tattoo placement. Further investigation of these color additives is warranted to better elucidate ink components responsible for these cutaneous allergic reactions.

Acknowledgments
We would like to thank Vitaly Terushkin, MD (West Orange, New Jersey, and New York, New York), and Arielle Kauvar, MD (New York, New York), for their contributions to the patient’s clinical care.

To the Editor:

Although relatively uncommon, hypersensitivity reactions to tattoo pigment are on the rise due to the increasing popularity and prevalence of tattoos.¹ Multiple adverse events have been described in association with tattoos, including inflammatory, infectious, and neoplastic responses.² An id reaction (also known as autoeczematization or autosensitization) develops distant to an initial site of infection or sensitization. We describe a unique case of an id reaction and subsequent development of prurigo nodules associated with contact allergy to red tattoo ink.

A 40-year-old woman was referred to the New York University Skin and Cancer Unit (New York, New York) for evaluation of a pruritic eruption arising on and near sites of tattooed skin on the right foot and right upper arm of 8 months’ duration. The patient reported that she had obtained a polychromatic tattoo on the right dorsal foot 9 months prior to the current presentation. Approximately 1 month later, she developed pruritic papulonodular lesions localized to the red-pigmented areas of the tattoo. Concomitantly, the patient developed a similar eruption confined to areas of red pigment in a polychromatic tattoo on the right upper arm that she had obtained 10 years prior. She was treated with intralesional triamcinolone to several of the lesions on the right dorsal foot with some benefit; however, a few days later she developed a generalized, erythematous, pruritic eruption on the back, abdomen, arms, and legs. Her medical history was remarkable only for mild iron-deficiency anemia. She had no known drug allergies or history of atopy and was not taking any medications prior to the onset of the eruption.

Skin examination revealed multiple, well-demarcated, eczematous papulonodules with surrounding erythema confined to the red-pigmented areas of the tattoo on the right dorsal foot, with several similar lesions on the surrounding nontattooed skin (Figure 1). Linear, well-demarcated, eczematous, hyperpigmented plaques also were noted on the red-pigmented areas of the tattoo on the patient’s right upper arm (Figure 2). Eczematous plaques and scattered excoriations were noted on the back, abdomen, flanks, arms, and legs.

Patch testing with the North American Standard Series, metal series, and samples of the red pigments used in the tattoo on the foot were negative. A punch biopsy of a lesion on the dorsal right foot showed a psoriasiform spongiotic dermatitis with eosinophils (Figure 3). Periodic acid–Schiff staining with diastase failed to reveal fungal hyphae. The histologic findings were consistent with allergic contact dermatitis. A punch biopsy of the eczematous reaction on nontattooed skin on the trunk demonstrated a perivascular dermatitis with eosinophils and subtle spongiosis consistent with an id reaction.

The patient was treated with fluocinonide ointment for several months with no effect. Subsequently, she received several short courses of oral prednisone, after which the affected areas of the tattoo on the arm and foot flattened and the id reaction resolved; however, after several months, the red-pigmented areas of the tattoo on the foot again became elevated and pruritic, and the patient developed widespread prurigo nodules on nontattooed skin on the trunk, arms, and legs. She was subsequently referred to a laser specialist for a trial of fractional laser treatment to cautiously remove the red tattoo pigment. After 2 treatments, the pruritus improved and the papular lesions appeared slightly flatter; however, the prurigo nodules remained. The tattoo on the patient’s foot was surgically removed; however, the prurigo nodules remained. Ultimately, the lesions cleared with a several-month course of mycophenolate mofetil.

Systemic allergic reactions to tattoo ink are rare but can cause considerable morbidity. An id reaction, also known as autoeczematization or autosensitization, is a reaction that develops distant to an initial site of infection or sensitization. Although the pathogenesis of this reaction is not certain, it has been hypothesized that autoimmunity to skin antigens might play a role.³ Autologous epidermal cells are thought to become antigenic in the presence of acute inflammation at the primary cutaneous site. These antigenic autologous epidermal cells are postulated to enter the circulation and cause secondary eczematous lesions at distant sites. This proposed mechanism is supported by the development of positive skin reactions to autologous extracts of epidermal scaling in patients with active id reaction.³

Hematogenous dissemination of cytokines has been implicated in id reactions.⁴ Keratinocytes produce cytokines in response to conditions that are known to trigger id reactions.⁵ Epidermal cytokines released from the primary site of sensitization are thought to heighten sensitivity at distant skin areas.⁴ These cytokines regulate both cell-mediated and humoral cutaneous immune responses. Increased levels of activated HLA-DR isotype–positive T cells in patients with active autoeczemization favors a cellular-mediated immune mechanism. The presence of activated antigen-specific T cells also supports the role of allergic contact dermatitis in triggering id reactions.⁶

Allergic contact dermatitis is the most common hypersensitivity reaction to tattoo ink, with red pigments representing the most common cause of tattoo-related allergic contact dermatitis. Historically, cinnabar (mercuric sulfide) has been the most common red pigment to cause allergic contact dermatitis.⁷ More recently, mercury-free organic pigments (eg, azo dyes) have been used in polychromatic tattoos due to their ability to retain color over long periods of time⁸; however, these organic red tattoo pigments also have been implicated in allergic reactions.^8-11 The composition of these new organic red tattoo pigments varies, but chemical analysis has revealed a mixture of aromatic azo compounds (eg, quinacridone),¹⁰ heavy metals (eg, aluminum, lead, cadmium, chromium, cobalt, iron, titanium),^9,12 and intermediate reactive compounds (eg, naphthalene, 2-naphthol, chlorobenzene, benzene).⁸ Allergic contact dermatitis to red tattoo ink is well documented^8,13; however, a PubMed search of articles indexed for MEDLINE using the terms tattoo and dermatitis, tattoo and allergy, tattoo and autosensitization, tattoo and id reaction, and tattoo and autoeczematization yielded only 3 other reports of a concomitant id reaction.^11,14,15

The diagnosis of id reaction associated with allergic contact dermatitis is made on the basis of clinical history, physical examination, and histopathology. Patch testing usually is not positive in cases of tattoo allergy; it is thought that the allergen is a tattoo ink byproduct possibly caused by photoinduced or metabolic change of the tattoo pigment and a haptenization process.^1,8,16 Histologically, variable reaction patterns, including eczematous, lichenoid, granulomatous, and pseudolymphomatous reactions have been reported in association with delayed-type inflammatory reactions to tattoo pigments, but the lichenoid pattern is most commonly observed.⁸

Treatment options for allergic contact dermatitis to tattoo ink include topical, intralesional, and oral steroids; topical calcineurin inhibitors; and surgical excision of the tattoo. Q-switched lasers—ruby, Nd:YAG, and alexandrite—are the gold standard for removing tattoo pigments¹⁷; however, these lasers remove tattoo pigment by selective photothermolysis, resulting in extracellular extravasation of pigment, which can precipitate a heightened immune response that can lead to localized and generalized allergic reactions.¹⁸ Therefore, Q-switched lasers should be avoided in the setting of an allergic reaction to tattoo ink. Fractional ablative laser resurfacing may be a safer alternative for removal of tattoos in the setting of an allergic reaction.¹⁷ Further studies are needed to confirm the safety and efficacy of this modality for allergic tattoo ink removal.^17,18

Our case illustrates a rare cause of id reaction and the subsequent development of prurigo nodules associated with contact allergy to red tattoo ink. We present this case to raise awareness of the potential health and iatrogenic risks associated with tattoo placement. Further investigation of these color additives is warranted to better elucidate ink components responsible for these cutaneous allergic reactions.

Acknowledgments
We would like to thank Vitaly Terushkin, MD (West Orange, New Jersey, and New York, New York), and Arielle Kauvar, MD (New York, New York), for their contributions to the patient’s clinical care.

Intravascular large B-cell lymphoma (IVBCL) is an aggressive and systemically disseminated disease that affects the elderly, with a median age of diagnosis around 70 years and no gender predilection. It is a rare subtype of extranodal diffuse large B-cell lymphoma (DLBCL) characterized by selective growth of neoplastic cells within blood vessel lumen without any obvious extravascular tumor mass. Hence, an absence of marked lymphadenopathy and heterogeneous clinical presentation make it difficult to diagnose accurately and timely, with roughly half of the cases found postmortem in previous case reports.^1,2 The exact incidence of this disease is not known, but more recently, the accuracy of diagnosis of this type of lymphoma has improved with random skin and bone marrow biopsy.^1,2 We present here a clinical case of this disease with an atypical presentation followed by a detailed review of its clinical aspects.

Case presentation and summary

A 43-year-old white woman with a history of hypothyroidism and recurrent ovarian cysts presented to clinic with 3 months of loss of appetite, abdominal distension, pelvic pain, and progressive lower-extremity swelling. A physical examination was notable for marked abdominal distension, diffuse lower abdominal tenderness, and pitting lower-extremity edema. No skin rash or any other cutaneous abnormality was noted on exam. Laboratory test results revealed a lactate dehydrogenase (LDH) level of 1652 U/L and a CA-125 level of 50 U/mL (reference range, 0-35 U/mL). No significant beta-human chorionic gonadotropin and alpha-fetoprotein levels were detected. Computed-tomographic (CT) imaging revealed small bilateral pleural effusions and gallbladder wall thickening with abdominal wall edema, but it was otherwise unrevealing. An echocardiogram showed normal cardiac structure and function, with a left ventricular ejection fraction of 60%. No protein was detected in the patient’s urine, and thyroid function tests were unrevealing. Doppler ultrasound studies of her lower extremities and abdomen revealed no thrombosis. Given the patient’s continued pelvic pain, history of ovarian cysts, and elevation in CA-125, she underwent a laparoscopic total abdominal hysterectomy and bilateral salpingoopherectomy.

Histologic examination revealed neoplastic cells involving only the vascular lumina of the cervix, endomyometrium, bilateral fallopian tubes, and bilateral ovaries (Figure 1). Immunohistochemistry stains were positive for CD5, CD20, PAX-5, CD45, BCL-2, and BCL-6 and focally positive for CD10. Peripheral smear showed pseudo-Pelger–Huet cells with 5% atypical lymphoma cells (Figure 2). Complete staging with positron-emission and CT (PET–CT) imaging revealed no metabolic activity, and a bone marrow biopsy showed trilineage hematopoiesis with adequate maturation and less than 5% of the marrow involved with large B-cell lymphoma cells. A diagnosis of IVBCL was made.

Discussion

Clinical presentation

The clinical manifestation of this disease is highly variable, and virtually any organ can be involved. Besides causing constitutional symptoms, including fatigue, B symptoms, and decline in performance status, heterogeneity of the clinical presentation depends on the organ system involved. One of the exceptional features of this disease is the difference in clinical presentation based on the geographical origin of the patient.^2-4

Western-variant IVBCL has a higher frequency of CNS and skin involvement, whereas Asian-variant IVBCL shows preferential involvement of bone marrow with hemophagocytosis, hepatosplenomegaly, and thrombocytopenia. However, these 2 clinical variants have no difference in clinical outcome, except with the cutaneous-variant kind.²⁴ A retrospective case series of 38 Western-variant IVBCL cases showed that 55% of patients had B symptoms with poor performance status.³ Brain and skin were the organs that were most frequently involved, with 68% of patients having involvement of at least 1 of those organs. Ten patients in this case series had disease that was exclusively limited to the skin and described as a “cutaneous variant” of IVBCL.³

Similarly, a retrospective case series of 96 cases of Asian-variant IVBCL showed B symptoms in 76% of patients, with predominant bone marrow involvement in 75% of patients, accompanied by hemophagocytosis in 66% and hepatosplenomegaly and anemia/thrombocytopenia in 77% and 84% of the patients, respectively.⁴ This difference in clinical presentation might have existed as a result of ethnic difference associated with production of inflammatory cytokines, including interferon gamma, tumor necrosis factor-alpha, interlukin-1 beta, and soluble interlukin-2 receptor, with levels of soluble interlukin-2 receptor found to be significantly higher in Asian patients than non-Asian patients.²
 

Diagnosis

Involved organ biopsy is mandatory for establishing the diagnosis of IVBCL. Laboratory findings are nonspecific, with the most common abnormality being increased serum LDH and beta-2 microglobulin levels observed in 80% to 90% or more of patients. Despite its intravascular growth pattern, IVBCL was associated with peripheral blood involvement in only 5% to 9% of patients.¹

Staging

Clinical staging work-up suggested for IVBCL patients by International Extranodal lymphoma study group in 2005 included physical examination (with emphasis on nervous system and skin), routine blood studies, peripheral blood smear, total body CT scan with contrast or PET–CT scan, MRI brain with contrast, CSF cytology, and bone marrow or organ biopsy.¹ The role of fluorodeoxyglucose-PET scan is controversial but can be helpful to detect unexpected locations for biopsy and to assess treatment response.^5,6

Morphology and immunophenotyping

In general, IVBCL histopathology shows large neoplastic lymphoid cells with large nuclei along with one or more nucleoli and scant cytoplasm within blood vessel lumen. Immunophenotypically, IVBCL cells mostly express nongerminal B-cell–associated markers with CD79a (100%), CD20 (96%), MUM-IRF4 (95%), CD5 (38%), and CD10 (12%) expressions. IVBCL cells have been demonstrated to lack cell surface protein CD29 and CD54 critical to transvascular migration. Similarly, aberrant expression of proteins such as CD11a and CXCR3 allows lymphoma cells to be attracted to endothelial cells, which might explain their intravascular confinement.⁷
 

Genetics

No pathognomic cytogenetic abnormalities have been reported in IVBCL to date, and the genetic features of this disease are not yet completely understood.^2,7
 

Management

IVBCL is considered a stage IV disseminated disease with an International Prognostic Index score of high-intermediate to high in most cases. Half of the patients with IVBCL who were treated with anthracycline-based chemotherapy relapsed and died within 18 months of diagnosis. One third of the relapses involved the CNS, thereby highlighting the importance of prophylactic CNS-directed Intrathecal therapy in an induction treatment regimen.^2-4 Ferreri and colleagues reported in their case series response rates of about 60%, with an overall survival (OS) of 3 years of 30% in patients who were treated with anthracycline-based chemotherapy. A multivariate analysis of the entire series showed cutaneous variant of the disease to be an independent favorable prognostic factor for OS.³

In the Murase and colleagues case series, the authors reported 67% response rates and a median OS of 13 months with CHOP (cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone) or CHOP-like regimens. Multivariate analysis showed older age, thrombocytopenia, and absence of anthracycline-based chemotherapy to be an independent negative prognostic factor for OS.⁴ Another retrospective analysis by Shimada and colleagues of 106 patients with IVBCL showed improved outcome with the addition of rituximab to CHOP-based chemotherapy (R-CHOP). Complete response rate (CR), 2-year progression-free survival, and OS were significantly higher for patients in rituximab-chemotherapy group than for those in the chemotherapy-alone group (CR, 82% vs 51%, respectively, P = .001; PFS, 56% vs 27%; OS, 66% vs 46%, P = .001), thereby establishing rituximab with CHOP-based therapy as induction therapy for IVBCL patients.⁸

The role of high-dose chemotherapy followed by ASCT could also be used as consolidation therapy to improve clinical outcomes as reported in 7 patients, showing durable remission after transplant in these 2 case series.^3,4 Another retrospective analysis of 6 patients with IVBCL who were treated with 6 cycles of R-CHOP as induction therapy and consolidated with ASCT reported all patients to be alive and in complete remission after a median follow-up of 56 months.⁹ Based on the retrospective case series data by Kato and colleagues and considering that more than 80% of the patients with IVBCL were in the high-risk International Prognostic Index group, ASCT in first remission might be a useful treatment option for durable remission; however, because the median age for the diagnosis of IVBCL is about 70 years, ASCT may not be a realistic option for all patients.
 

Conclusions

IVBCL is a rare, aggressive, and distinct type of DLBCL with complex constellations of symptoms requiring strong clinical suspicion to establish this challenging diagnosis. Rituximab with anthracycline-based therapy along with prophylactic CNS-directed therapy followed by consolidative ASCT may lead to long-term remission. More research is needed into the genetic features of this disease to better understand its pathogenesis and potential targets for treatment.

Intravascular large B-cell lymphoma (IVBCL) is an aggressive and systemically disseminated disease that affects the elderly, with a median age of diagnosis around 70 years and no gender predilection. It is a rare subtype of extranodal diffuse large B-cell lymphoma (DLBCL) characterized by selective growth of neoplastic cells within blood vessel lumen without any obvious extravascular tumor mass. Hence, an absence of marked lymphadenopathy and heterogeneous clinical presentation make it difficult to diagnose accurately and timely, with roughly half of the cases found postmortem in previous case reports.^1,2 The exact incidence of this disease is not known, but more recently, the accuracy of diagnosis of this type of lymphoma has improved with random skin and bone marrow biopsy.^1,2 We present here a clinical case of this disease with an atypical presentation followed by a detailed review of its clinical aspects.

Case presentation and summary

A 43-year-old white woman with a history of hypothyroidism and recurrent ovarian cysts presented to clinic with 3 months of loss of appetite, abdominal distension, pelvic pain, and progressive lower-extremity swelling. A physical examination was notable for marked abdominal distension, diffuse lower abdominal tenderness, and pitting lower-extremity edema. No skin rash or any other cutaneous abnormality was noted on exam. Laboratory test results revealed a lactate dehydrogenase (LDH) level of 1652 U/L and a CA-125 level of 50 U/mL (reference range, 0-35 U/mL). No significant beta-human chorionic gonadotropin and alpha-fetoprotein levels were detected. Computed-tomographic (CT) imaging revealed small bilateral pleural effusions and gallbladder wall thickening with abdominal wall edema, but it was otherwise unrevealing. An echocardiogram showed normal cardiac structure and function, with a left ventricular ejection fraction of 60%. No protein was detected in the patient’s urine, and thyroid function tests were unrevealing. Doppler ultrasound studies of her lower extremities and abdomen revealed no thrombosis. Given the patient’s continued pelvic pain, history of ovarian cysts, and elevation in CA-125, she underwent a laparoscopic total abdominal hysterectomy and bilateral salpingoopherectomy.

Histologic examination revealed neoplastic cells involving only the vascular lumina of the cervix, endomyometrium, bilateral fallopian tubes, and bilateral ovaries (Figure 1). Immunohistochemistry stains were positive for CD5, CD20, PAX-5, CD45, BCL-2, and BCL-6 and focally positive for CD10. Peripheral smear showed pseudo-Pelger–Huet cells with 5% atypical lymphoma cells (Figure 2). Complete staging with positron-emission and CT (PET–CT) imaging revealed no metabolic activity, and a bone marrow biopsy showed trilineage hematopoiesis with adequate maturation and less than 5% of the marrow involved with large B-cell lymphoma cells. A diagnosis of IVBCL was made.

Discussion

Clinical presentation

The clinical manifestation of this disease is highly variable, and virtually any organ can be involved. Besides causing constitutional symptoms, including fatigue, B symptoms, and decline in performance status, heterogeneity of the clinical presentation depends on the organ system involved. One of the exceptional features of this disease is the difference in clinical presentation based on the geographical origin of the patient.^2-4

Western-variant IVBCL has a higher frequency of CNS and skin involvement, whereas Asian-variant IVBCL shows preferential involvement of bone marrow with hemophagocytosis, hepatosplenomegaly, and thrombocytopenia. However, these 2 clinical variants have no difference in clinical outcome, except with the cutaneous-variant kind.²⁴ A retrospective case series of 38 Western-variant IVBCL cases showed that 55% of patients had B symptoms with poor performance status.³ Brain and skin were the organs that were most frequently involved, with 68% of patients having involvement of at least 1 of those organs. Ten patients in this case series had disease that was exclusively limited to the skin and described as a “cutaneous variant” of IVBCL.³

Similarly, a retrospective case series of 96 cases of Asian-variant IVBCL showed B symptoms in 76% of patients, with predominant bone marrow involvement in 75% of patients, accompanied by hemophagocytosis in 66% and hepatosplenomegaly and anemia/thrombocytopenia in 77% and 84% of the patients, respectively.⁴ This difference in clinical presentation might have existed as a result of ethnic difference associated with production of inflammatory cytokines, including interferon gamma, tumor necrosis factor-alpha, interlukin-1 beta, and soluble interlukin-2 receptor, with levels of soluble interlukin-2 receptor found to be significantly higher in Asian patients than non-Asian patients.²
 

Diagnosis

Involved organ biopsy is mandatory for establishing the diagnosis of IVBCL. Laboratory findings are nonspecific, with the most common abnormality being increased serum LDH and beta-2 microglobulin levels observed in 80% to 90% or more of patients. Despite its intravascular growth pattern, IVBCL was associated with peripheral blood involvement in only 5% to 9% of patients.¹

Staging

Clinical staging work-up suggested for IVBCL patients by International Extranodal lymphoma study group in 2005 included physical examination (with emphasis on nervous system and skin), routine blood studies, peripheral blood smear, total body CT scan with contrast or PET–CT scan, MRI brain with contrast, CSF cytology, and bone marrow or organ biopsy.¹ The role of fluorodeoxyglucose-PET scan is controversial but can be helpful to detect unexpected locations for biopsy and to assess treatment response.^5,6

Morphology and immunophenotyping

In general, IVBCL histopathology shows large neoplastic lymphoid cells with large nuclei along with one or more nucleoli and scant cytoplasm within blood vessel lumen. Immunophenotypically, IVBCL cells mostly express nongerminal B-cell–associated markers with CD79a (100%), CD20 (96%), MUM-IRF4 (95%), CD5 (38%), and CD10 (12%) expressions. IVBCL cells have been demonstrated to lack cell surface protein CD29 and CD54 critical to transvascular migration. Similarly, aberrant expression of proteins such as CD11a and CXCR3 allows lymphoma cells to be attracted to endothelial cells, which might explain their intravascular confinement.⁷
 

Genetics

No pathognomic cytogenetic abnormalities have been reported in IVBCL to date, and the genetic features of this disease are not yet completely understood.^2,7
 

Management

IVBCL is considered a stage IV disseminated disease with an International Prognostic Index score of high-intermediate to high in most cases. Half of the patients with IVBCL who were treated with anthracycline-based chemotherapy relapsed and died within 18 months of diagnosis. One third of the relapses involved the CNS, thereby highlighting the importance of prophylactic CNS-directed Intrathecal therapy in an induction treatment regimen.^2-4 Ferreri and colleagues reported in their case series response rates of about 60%, with an overall survival (OS) of 3 years of 30% in patients who were treated with anthracycline-based chemotherapy. A multivariate analysis of the entire series showed cutaneous variant of the disease to be an independent favorable prognostic factor for OS.³

In the Murase and colleagues case series, the authors reported 67% response rates and a median OS of 13 months with CHOP (cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone) or CHOP-like regimens. Multivariate analysis showed older age, thrombocytopenia, and absence of anthracycline-based chemotherapy to be an independent negative prognostic factor for OS.⁴ Another retrospective analysis by Shimada and colleagues of 106 patients with IVBCL showed improved outcome with the addition of rituximab to CHOP-based chemotherapy (R-CHOP). Complete response rate (CR), 2-year progression-free survival, and OS were significantly higher for patients in rituximab-chemotherapy group than for those in the chemotherapy-alone group (CR, 82% vs 51%, respectively, P = .001; PFS, 56% vs 27%; OS, 66% vs 46%, P = .001), thereby establishing rituximab with CHOP-based therapy as induction therapy for IVBCL patients.⁸

The role of high-dose chemotherapy followed by ASCT could also be used as consolidation therapy to improve clinical outcomes as reported in 7 patients, showing durable remission after transplant in these 2 case series.^3,4 Another retrospective analysis of 6 patients with IVBCL who were treated with 6 cycles of R-CHOP as induction therapy and consolidated with ASCT reported all patients to be alive and in complete remission after a median follow-up of 56 months.⁹ Based on the retrospective case series data by Kato and colleagues and considering that more than 80% of the patients with IVBCL were in the high-risk International Prognostic Index group, ASCT in first remission might be a useful treatment option for durable remission; however, because the median age for the diagnosis of IVBCL is about 70 years, ASCT may not be a realistic option for all patients.
 

Conclusions

IVBCL is a rare, aggressive, and distinct type of DLBCL with complex constellations of symptoms requiring strong clinical suspicion to establish this challenging diagnosis. Rituximab with anthracycline-based therapy along with prophylactic CNS-directed therapy followed by consolidative ASCT may lead to long-term remission. More research is needed into the genetic features of this disease to better understand its pathogenesis and potential targets for treatment.

Intravascular large B-cell lymphoma (IVBCL) is an aggressive and systemically disseminated disease that affects the elderly, with a median age of diagnosis around 70 years and no gender predilection. It is a rare subtype of extranodal diffuse large B-cell lymphoma (DLBCL) characterized by selective growth of neoplastic cells within blood vessel lumen without any obvious extravascular tumor mass. Hence, an absence of marked lymphadenopathy and heterogeneous clinical presentation make it difficult to diagnose accurately and timely, with roughly half of the cases found postmortem in previous case reports.^1,2 The exact incidence of this disease is not known, but more recently, the accuracy of diagnosis of this type of lymphoma has improved with random skin and bone marrow biopsy.^1,2 We present here a clinical case of this disease with an atypical presentation followed by a detailed review of its clinical aspects.

Case presentation and summary

A 43-year-old white woman with a history of hypothyroidism and recurrent ovarian cysts presented to clinic with 3 months of loss of appetite, abdominal distension, pelvic pain, and progressive lower-extremity swelling. A physical examination was notable for marked abdominal distension, diffuse lower abdominal tenderness, and pitting lower-extremity edema. No skin rash or any other cutaneous abnormality was noted on exam. Laboratory test results revealed a lactate dehydrogenase (LDH) level of 1652 U/L and a CA-125 level of 50 U/mL (reference range, 0-35 U/mL). No significant beta-human chorionic gonadotropin and alpha-fetoprotein levels were detected. Computed-tomographic (CT) imaging revealed small bilateral pleural effusions and gallbladder wall thickening with abdominal wall edema, but it was otherwise unrevealing. An echocardiogram showed normal cardiac structure and function, with a left ventricular ejection fraction of 60%. No protein was detected in the patient’s urine, and thyroid function tests were unrevealing. Doppler ultrasound studies of her lower extremities and abdomen revealed no thrombosis. Given the patient’s continued pelvic pain, history of ovarian cysts, and elevation in CA-125, she underwent a laparoscopic total abdominal hysterectomy and bilateral salpingoopherectomy.

Histologic examination revealed neoplastic cells involving only the vascular lumina of the cervix, endomyometrium, bilateral fallopian tubes, and bilateral ovaries (Figure 1). Immunohistochemistry stains were positive for CD5, CD20, PAX-5, CD45, BCL-2, and BCL-6 and focally positive for CD10. Peripheral smear showed pseudo-Pelger–Huet cells with 5% atypical lymphoma cells (Figure 2). Complete staging with positron-emission and CT (PET–CT) imaging revealed no metabolic activity, and a bone marrow biopsy showed trilineage hematopoiesis with adequate maturation and less than 5% of the marrow involved with large B-cell lymphoma cells. A diagnosis of IVBCL was made.

Discussion

Clinical presentation

The clinical manifestation of this disease is highly variable, and virtually any organ can be involved. Besides causing constitutional symptoms, including fatigue, B symptoms, and decline in performance status, heterogeneity of the clinical presentation depends on the organ system involved. One of the exceptional features of this disease is the difference in clinical presentation based on the geographical origin of the patient.^2-4

Western-variant IVBCL has a higher frequency of CNS and skin involvement, whereas Asian-variant IVBCL shows preferential involvement of bone marrow with hemophagocytosis, hepatosplenomegaly, and thrombocytopenia. However, these 2 clinical variants have no difference in clinical outcome, except with the cutaneous-variant kind.²⁴ A retrospective case series of 38 Western-variant IVBCL cases showed that 55% of patients had B symptoms with poor performance status.³ Brain and skin were the organs that were most frequently involved, with 68% of patients having involvement of at least 1 of those organs. Ten patients in this case series had disease that was exclusively limited to the skin and described as a “cutaneous variant” of IVBCL.³

Similarly, a retrospective case series of 96 cases of Asian-variant IVBCL showed B symptoms in 76% of patients, with predominant bone marrow involvement in 75% of patients, accompanied by hemophagocytosis in 66% and hepatosplenomegaly and anemia/thrombocytopenia in 77% and 84% of the patients, respectively.⁴ This difference in clinical presentation might have existed as a result of ethnic difference associated with production of inflammatory cytokines, including interferon gamma, tumor necrosis factor-alpha, interlukin-1 beta, and soluble interlukin-2 receptor, with levels of soluble interlukin-2 receptor found to be significantly higher in Asian patients than non-Asian patients.²
 

Diagnosis

Involved organ biopsy is mandatory for establishing the diagnosis of IVBCL. Laboratory findings are nonspecific, with the most common abnormality being increased serum LDH and beta-2 microglobulin levels observed in 80% to 90% or more of patients. Despite its intravascular growth pattern, IVBCL was associated with peripheral blood involvement in only 5% to 9% of patients.¹

Staging

Clinical staging work-up suggested for IVBCL patients by International Extranodal lymphoma study group in 2005 included physical examination (with emphasis on nervous system and skin), routine blood studies, peripheral blood smear, total body CT scan with contrast or PET–CT scan, MRI brain with contrast, CSF cytology, and bone marrow or organ biopsy.¹ The role of fluorodeoxyglucose-PET scan is controversial but can be helpful to detect unexpected locations for biopsy and to assess treatment response.^5,6

Morphology and immunophenotyping

In general, IVBCL histopathology shows large neoplastic lymphoid cells with large nuclei along with one or more nucleoli and scant cytoplasm within blood vessel lumen. Immunophenotypically, IVBCL cells mostly express nongerminal B-cell–associated markers with CD79a (100%), CD20 (96%), MUM-IRF4 (95%), CD5 (38%), and CD10 (12%) expressions. IVBCL cells have been demonstrated to lack cell surface protein CD29 and CD54 critical to transvascular migration. Similarly, aberrant expression of proteins such as CD11a and CXCR3 allows lymphoma cells to be attracted to endothelial cells, which might explain their intravascular confinement.⁷
 

Genetics

No pathognomic cytogenetic abnormalities have been reported in IVBCL to date, and the genetic features of this disease are not yet completely understood.^2,7
 

Management

IVBCL is considered a stage IV disseminated disease with an International Prognostic Index score of high-intermediate to high in most cases. Half of the patients with IVBCL who were treated with anthracycline-based chemotherapy relapsed and died within 18 months of diagnosis. One third of the relapses involved the CNS, thereby highlighting the importance of prophylactic CNS-directed Intrathecal therapy in an induction treatment regimen.^2-4 Ferreri and colleagues reported in their case series response rates of about 60%, with an overall survival (OS) of 3 years of 30% in patients who were treated with anthracycline-based chemotherapy. A multivariate analysis of the entire series showed cutaneous variant of the disease to be an independent favorable prognostic factor for OS.³

In the Murase and colleagues case series, the authors reported 67% response rates and a median OS of 13 months with CHOP (cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone) or CHOP-like regimens. Multivariate analysis showed older age, thrombocytopenia, and absence of anthracycline-based chemotherapy to be an independent negative prognostic factor for OS.⁴ Another retrospective analysis by Shimada and colleagues of 106 patients with IVBCL showed improved outcome with the addition of rituximab to CHOP-based chemotherapy (R-CHOP). Complete response rate (CR), 2-year progression-free survival, and OS were significantly higher for patients in rituximab-chemotherapy group than for those in the chemotherapy-alone group (CR, 82% vs 51%, respectively, P = .001; PFS, 56% vs 27%; OS, 66% vs 46%, P = .001), thereby establishing rituximab with CHOP-based therapy as induction therapy for IVBCL patients.⁸

The role of high-dose chemotherapy followed by ASCT could also be used as consolidation therapy to improve clinical outcomes as reported in 7 patients, showing durable remission after transplant in these 2 case series.^3,4 Another retrospective analysis of 6 patients with IVBCL who were treated with 6 cycles of R-CHOP as induction therapy and consolidated with ASCT reported all patients to be alive and in complete remission after a median follow-up of 56 months.⁹ Based on the retrospective case series data by Kato and colleagues and considering that more than 80% of the patients with IVBCL were in the high-risk International Prognostic Index group, ASCT in first remission might be a useful treatment option for durable remission; however, because the median age for the diagnosis of IVBCL is about 70 years, ASCT may not be a realistic option for all patients.
 

Conclusions

IVBCL is a rare, aggressive, and distinct type of DLBCL with complex constellations of symptoms requiring strong clinical suspicion to establish this challenging diagnosis. Rituximab with anthracycline-based therapy along with prophylactic CNS-directed therapy followed by consolidative ASCT may lead to long-term remission. More research is needed into the genetic features of this disease to better understand its pathogenesis and potential targets for treatment.

About 12.4% of women in the United States will be diagnosed with breast cancer at some point in their lifetime.¹ A percentage of these women will develop metastatic disease and are estimated to have a 5-year survival rate of 22%.² There have been meaningful improvements in survival because of earlier detection and more effective systemic therapies. Patients with hormone-sensitive disease often respond to endocrine therapy, and this frequently represents front-line treatment for these patients, resulting in palliation of symptoms while maintaining quality of life.

However, endocrine resistance inevitably occurs, and a great deal of research has been focused on developing strategies to combat resistance. One mechanism of endocrine resistance is though the Cyclin-dependent kinases 4 and 6 (CDK4/6) complexes. Among the most promising of the strategies to prevent resistance are the CDK4/6 inhibitors. There are now 3 approved CDK4/6 inhibitor drugs that can be used in combination with endocrine therapy, 1 of which can also be used as a single agent. When used in combination with endocrine therapy, the use of CDK 4/6 inhibitors has significantly improved progression-free survival (PFS) in patients with hormone-sensitive HER2-negative metastatic breast cancer by inhibiting cellular division and growth.³ In postmenopausal women, endocrine therapy plus CDK4/6 inhibitors are the preferred first-line regimen for metastatic disease.

Since the approval of palbociclib by the US Food and Drug Administration in 2015, the most common hematologic lab abnormalities are anemia, leukopenia, neutropenia, and thrombocytopenia. The most common nonhematologic adverse events (AEs) are fatigue, infection, nausea, and stomatitis. Hepatic toxicity has not been commonly observed. We report here the case of a 57-year-old woman on palbociclib and fulvestrant who developed significant elevation of liver function tests after starting palbociclib, suggesting a possible drug-induced liver injury from palbociclib.
 

Case presentation and summary

A 57-year-old woman with history of hypothyroidism and hypertension presented in May 2016 with a lump in her right breast and back pain. The lump was biopsied and revealed invasive ductal carcinoma, moderately differentiated, estrogen receptor (ER) positive 100%, progesterone receptor (PR) positive 95%, and HER2 negative. A positron emission tomography (PET)–computed tomography (CT) scan and magnetic resonance imaging showed bone metastasis at several vertebral levels, and the results of a bone biopsy confirmed metastatic adenocarcinoma of breast origin, ER positive 60%, PR positive 40%, and HER2 negative. No liver lesions were seen on imaging, but there was suggestion of fatty liver. She was started on letrozole 2.5 mg daily in July 2016 while undergoing kyphoplasty and subsequent radiation. A restaging PET scan revealed progression of disease on letrozole, with possible new rib lesion and progression in the breast. No liver disease was noted. Therapy was changed to fulvestrant and palbociclib. Fulvestrant was started in March 2017 with standard dosing of 500 mg intramuscular on days 1, 15, and 29, and then once a month thereafter. Her first cycle of palbociclib was started on April 5, dosed at 125 mg by mouth daily for 21 days, followed by 7 days off, repeated every 28 days (all dates hereinafter fell within 2017, unless otherwise stipulated).

Labs checked on April 28 and May 26 were unremarkable. A restaging CT scan of the chest, abdomen, and pelvis was done on June 21 after completion of 3 cycles of fulvestrant and palbociclib. There was no evidence of liver metastases, only the fatty infiltration of the liver that had been seen previously. On June 23, 2017, lab results showed a transaminitis with an alanine aminotransferase (ALT) level of 446 IU/L (reference range 10-33 IU/L) and aspartate aminotransferase (AST) level of 183 IU/L (reference range 0-32 IU/L).

The patient’s liver enzyme levels continued to increase and peaked on July 3 at ALT >700 IU/L and AST at 421 IU/L. Her total bilirubin and alkaline phosphatase levels remained within normal limits. She had received her final dose of fulvestrant on May 31 and had taken her last dose of palbociclib on June 20, 2017. She had no history of elevated liver enzymes or liver disease, although the initial PET scan done at diagnosis had suggested hepatic steatosis. She said she had not recently used antibiotics, alcohol, or over-the-counter medications or supplements. There was no family history of liver problems, inflammatory bowel disease, or gastrointestinal malignancy. The only other medications she had taken recently were denosumab, levothyroxine for hypothyroidism, and amlodipine for hypertension. She was seen by hepatology for evaluation of acute hepatitis. Other etiologies for her elevated liver enzymes were ruled out, and she was diagnosed with a drug-induced liver injury from one of her anticancer medications. Her treatments with fulvestrant and palbociclib were held, and the results of her liver function tests normalized by September 2017.

Fulvestrant was restarted on August 24, and her lab results remained normal through November of that year, when restaging scans showed progression with new axillary adenopathy suspicious for metastasis. Imaging also showed a 1.6-cm hepatic lesion suggestive of a focal area of fat deposition or atypical hemangioma without definitive evidence of metastasis. Follow-up imaging was recommended. She was therefore rechallenged with palbociclib at a reduced dose of 100 mg by mouth daily and received the first dose on November 30. On December 8, repeat labs again showed elevated liver function tests (ALT, 285 IU/L; AST, 112 IU/L). Treatment with palbociclib was discontinued on December 10. Because the patient was not able to tolerate palbociclib, and fulvestrant alone was not controlling the disease, she was started on an alternate endocrine therapy with tamoxifen on December 26. The patient’s liver function tests normalized again by January 2018.
 

Discussion

The use of targeted therapies has changed the landscape of oncologic treatments. Several studies have evaluated the safety and efficacy of palbociclib in combination with endocrine therapy. The Palbociclib Ongoing Trials in the Management of Breast Cancer (PALOMA)-1 study, an open-label, randomized, phase-2 trial involving patients with newly diagnosed metastatic hormone sensitive HER2-negative breast cancer, demonstrated that palbociclib in combination with letrozole was associated with significantly longer PFS than letrozole alone.⁴ These results were later confirmed in the larger PALOMA-2 study, a randomized, double-blind, phase-3 trial that evaluated 666 postmenopausal patients with no prior systemic therapy. In that study, median PFS for the palbociclib–letrozole group was 24.8 months, compared with 14.5 months for the letrozole-alone group (hazard ratio [HR] for disease progression or death, 0.58 [0.46–0.72], P < .001).⁵ The most recent PALOMA-3 study, a phase-3 trial involving 521 patients with advanced hormone receptor–positive, HER2-negative breast cancer that had progressed during initial endocrine therapy, evaluated the efficacy of combined palbociclib and fulvestrant in a randomized, double-blind, placebo-controlled, parallel-group trial. The result was that the palbociclib–fulvestrant combination resulted in longer median PFS of 9.2 months, compared with 3.8 months with fulvestrant alone (P < .001).⁶

These trials also monitored the number of AEs as secondary aims. The most commonly reported AEs in the PALOMA trials for those patients in the palbociclib group were hematologic, with neutropenia being the most common, followed by leukopenia, anemia, and thrombocytopenia. The most common nonhematologic AEs reported in the palbociclib-fulvestrant group were fatigue, nausea, and headache. Elevated liver function tests were a rare but reported AE in 7.2% of the palbociclib-treated patients in the PALOMA-1 study.⁷ In the PALOMA-2 study, ALT and AST elevations were reported as AEs (all grades) in 9.9% and 9.7% of palbociclib-treated patients, respectively.⁵ In the PALOMA-3 study, there was 1 fatal serious AE of hepatic failure with grade 5 disease progression in the palbociclib group; however, the patient’s medical history included progressive liver metastasis and disease progression.⁶ A pooled safety analysis conducted across all PALOMA studies demonstrated that grade 3/4 AST and ALT elevations occurred in 3.3% and 2.3% of palbociclib-treated patients, respectively, again highlighting a reported but rare occurrence.⁸

The patient described in the present case report started on combination fulvestrant and palbociclib after her disease showed progression on letrozole. She developed an increase in transaminases after completing 3 cycles of palbociclib. Liver function tests increased nearly 12 weeks after beginning her first cycle of the CDK 4/6 inhibitor. Staging scans of the patient demonstrated fatty liver. It is not known if her fatty liver contributed to her transminitis; however, her baseline labs showed normal liver function tests, and they did not increase until after therapy with fulvestrant–palbociclib was started. It might have been that her fatty liver caused her to be at higher risk of transaminitis with administration of palbociclib, although we cannot be certain. Her lab results remained normal while she was on fulvestrant alone, and the liver function test results increased only after palbociclib was started, making this drug the more likely culprit.

Both events of increased liver enzymes occurred within a week of the last palbociclib dose; however, we note that hepatotoxicity developed at a faster rate when the patient was rechallenged with palbociclib at a lower dose, with elevated liver function tests increasing 1 week after restarting treatment as opposed to the first episode that occurred after 3 cycles of the palbociclib. After discontinuation of the medication, liver function tests again normalized, suggesting that palbociclib was most likely the causative agent. In addition, the degree of elevated liver enzymes was less severe on re-exposure at the lower dose of 100 mg, which raises the possibility that there could be a dose-dependent association between palbociclib and hepatotoxicity. There have been few case reports of increased liver enzymes associated with palbociclib, and it is only recently that this association has been more recognized. A meta-analysis by Zaw and colleagues has demonstrated that CDK 4/6 inhibitor–based regimens are associated with a higher risk of elevated AST and ALT; however, their relation with dose dependence was not described. In particular, they found that CDK 4/6 inhibitors increased the risk of high-grade, elevated ALT with a relative risk of 4.33 (95% confidence interval, 2.15-8.71; P < .0001). The meta-analysis also included other CDK 4/6 inhibitors such as abemaciclib and ribociclib, which have been more commonly associated with liver toxicity than palbociclib has.⁹ Our case report highlights the specific association between palbociclib and elevated liver enzymes.

In conclusion, this case report illustrates that our patient’s elevated liver enzymes were likely related to palbociclib. This is further supported by the fact that this AE occurred twice, both times after palbociclib exposure. In each instance, liver enzymes normalized after discontinuation of palbociclib. One cannot entirely rule out that fulvestrant might have been the culprit medication, but the patient’s normal hepatic panel for several months after starting fulvestrant suggests that is less likely. This case report is indicative of an uncommon complication in the treatment of metastatic breast cancer, one that is starting to gain more recognition, and we must think of palbociclib as a possible cause of drug-induced liver injury when targeted CDK 4/6–based regimens are used.

About 12.4% of women in the United States will be diagnosed with breast cancer at some point in their lifetime.¹ A percentage of these women will develop metastatic disease and are estimated to have a 5-year survival rate of 22%.² There have been meaningful improvements in survival because of earlier detection and more effective systemic therapies. Patients with hormone-sensitive disease often respond to endocrine therapy, and this frequently represents front-line treatment for these patients, resulting in palliation of symptoms while maintaining quality of life.

However, endocrine resistance inevitably occurs, and a great deal of research has been focused on developing strategies to combat resistance. One mechanism of endocrine resistance is though the Cyclin-dependent kinases 4 and 6 (CDK4/6) complexes. Among the most promising of the strategies to prevent resistance are the CDK4/6 inhibitors. There are now 3 approved CDK4/6 inhibitor drugs that can be used in combination with endocrine therapy, 1 of which can also be used as a single agent. When used in combination with endocrine therapy, the use of CDK 4/6 inhibitors has significantly improved progression-free survival (PFS) in patients with hormone-sensitive HER2-negative metastatic breast cancer by inhibiting cellular division and growth.³ In postmenopausal women, endocrine therapy plus CDK4/6 inhibitors are the preferred first-line regimen for metastatic disease.

Since the approval of palbociclib by the US Food and Drug Administration in 2015, the most common hematologic lab abnormalities are anemia, leukopenia, neutropenia, and thrombocytopenia. The most common nonhematologic adverse events (AEs) are fatigue, infection, nausea, and stomatitis. Hepatic toxicity has not been commonly observed. We report here the case of a 57-year-old woman on palbociclib and fulvestrant who developed significant elevation of liver function tests after starting palbociclib, suggesting a possible drug-induced liver injury from palbociclib.
 

Case presentation and summary

A 57-year-old woman with history of hypothyroidism and hypertension presented in May 2016 with a lump in her right breast and back pain. The lump was biopsied and revealed invasive ductal carcinoma, moderately differentiated, estrogen receptor (ER) positive 100%, progesterone receptor (PR) positive 95%, and HER2 negative. A positron emission tomography (PET)–computed tomography (CT) scan and magnetic resonance imaging showed bone metastasis at several vertebral levels, and the results of a bone biopsy confirmed metastatic adenocarcinoma of breast origin, ER positive 60%, PR positive 40%, and HER2 negative. No liver lesions were seen on imaging, but there was suggestion of fatty liver. She was started on letrozole 2.5 mg daily in July 2016 while undergoing kyphoplasty and subsequent radiation. A restaging PET scan revealed progression of disease on letrozole, with possible new rib lesion and progression in the breast. No liver disease was noted. Therapy was changed to fulvestrant and palbociclib. Fulvestrant was started in March 2017 with standard dosing of 500 mg intramuscular on days 1, 15, and 29, and then once a month thereafter. Her first cycle of palbociclib was started on April 5, dosed at 125 mg by mouth daily for 21 days, followed by 7 days off, repeated every 28 days (all dates hereinafter fell within 2017, unless otherwise stipulated).

Labs checked on April 28 and May 26 were unremarkable. A restaging CT scan of the chest, abdomen, and pelvis was done on June 21 after completion of 3 cycles of fulvestrant and palbociclib. There was no evidence of liver metastases, only the fatty infiltration of the liver that had been seen previously. On June 23, 2017, lab results showed a transaminitis with an alanine aminotransferase (ALT) level of 446 IU/L (reference range 10-33 IU/L) and aspartate aminotransferase (AST) level of 183 IU/L (reference range 0-32 IU/L).

The patient’s liver enzyme levels continued to increase and peaked on July 3 at ALT >700 IU/L and AST at 421 IU/L. Her total bilirubin and alkaline phosphatase levels remained within normal limits. She had received her final dose of fulvestrant on May 31 and had taken her last dose of palbociclib on June 20, 2017. She had no history of elevated liver enzymes or liver disease, although the initial PET scan done at diagnosis had suggested hepatic steatosis. She said she had not recently used antibiotics, alcohol, or over-the-counter medications or supplements. There was no family history of liver problems, inflammatory bowel disease, or gastrointestinal malignancy. The only other medications she had taken recently were denosumab, levothyroxine for hypothyroidism, and amlodipine for hypertension. She was seen by hepatology for evaluation of acute hepatitis. Other etiologies for her elevated liver enzymes were ruled out, and she was diagnosed with a drug-induced liver injury from one of her anticancer medications. Her treatments with fulvestrant and palbociclib were held, and the results of her liver function tests normalized by September 2017.

Fulvestrant was restarted on August 24, and her lab results remained normal through November of that year, when restaging scans showed progression with new axillary adenopathy suspicious for metastasis. Imaging also showed a 1.6-cm hepatic lesion suggestive of a focal area of fat deposition or atypical hemangioma without definitive evidence of metastasis. Follow-up imaging was recommended. She was therefore rechallenged with palbociclib at a reduced dose of 100 mg by mouth daily and received the first dose on November 30. On December 8, repeat labs again showed elevated liver function tests (ALT, 285 IU/L; AST, 112 IU/L). Treatment with palbociclib was discontinued on December 10. Because the patient was not able to tolerate palbociclib, and fulvestrant alone was not controlling the disease, she was started on an alternate endocrine therapy with tamoxifen on December 26. The patient’s liver function tests normalized again by January 2018.
 

Discussion

The use of targeted therapies has changed the landscape of oncologic treatments. Several studies have evaluated the safety and efficacy of palbociclib in combination with endocrine therapy. The Palbociclib Ongoing Trials in the Management of Breast Cancer (PALOMA)-1 study, an open-label, randomized, phase-2 trial involving patients with newly diagnosed metastatic hormone sensitive HER2-negative breast cancer, demonstrated that palbociclib in combination with letrozole was associated with significantly longer PFS than letrozole alone.⁴ These results were later confirmed in the larger PALOMA-2 study, a randomized, double-blind, phase-3 trial that evaluated 666 postmenopausal patients with no prior systemic therapy. In that study, median PFS for the palbociclib–letrozole group was 24.8 months, compared with 14.5 months for the letrozole-alone group (hazard ratio [HR] for disease progression or death, 0.58 [0.46–0.72], P < .001).⁵ The most recent PALOMA-3 study, a phase-3 trial involving 521 patients with advanced hormone receptor–positive, HER2-negative breast cancer that had progressed during initial endocrine therapy, evaluated the efficacy of combined palbociclib and fulvestrant in a randomized, double-blind, placebo-controlled, parallel-group trial. The result was that the palbociclib–fulvestrant combination resulted in longer median PFS of 9.2 months, compared with 3.8 months with fulvestrant alone (P < .001).⁶

These trials also monitored the number of AEs as secondary aims. The most commonly reported AEs in the PALOMA trials for those patients in the palbociclib group were hematologic, with neutropenia being the most common, followed by leukopenia, anemia, and thrombocytopenia. The most common nonhematologic AEs reported in the palbociclib-fulvestrant group were fatigue, nausea, and headache. Elevated liver function tests were a rare but reported AE in 7.2% of the palbociclib-treated patients in the PALOMA-1 study.⁷ In the PALOMA-2 study, ALT and AST elevations were reported as AEs (all grades) in 9.9% and 9.7% of palbociclib-treated patients, respectively.⁵ In the PALOMA-3 study, there was 1 fatal serious AE of hepatic failure with grade 5 disease progression in the palbociclib group; however, the patient’s medical history included progressive liver metastasis and disease progression.⁶ A pooled safety analysis conducted across all PALOMA studies demonstrated that grade 3/4 AST and ALT elevations occurred in 3.3% and 2.3% of palbociclib-treated patients, respectively, again highlighting a reported but rare occurrence.⁸

The patient described in the present case report started on combination fulvestrant and palbociclib after her disease showed progression on letrozole. She developed an increase in transaminases after completing 3 cycles of palbociclib. Liver function tests increased nearly 12 weeks after beginning her first cycle of the CDK 4/6 inhibitor. Staging scans of the patient demonstrated fatty liver. It is not known if her fatty liver contributed to her transminitis; however, her baseline labs showed normal liver function tests, and they did not increase until after therapy with fulvestrant–palbociclib was started. It might have been that her fatty liver caused her to be at higher risk of transaminitis with administration of palbociclib, although we cannot be certain. Her lab results remained normal while she was on fulvestrant alone, and the liver function test results increased only after palbociclib was started, making this drug the more likely culprit.

Both events of increased liver enzymes occurred within a week of the last palbociclib dose; however, we note that hepatotoxicity developed at a faster rate when the patient was rechallenged with palbociclib at a lower dose, with elevated liver function tests increasing 1 week after restarting treatment as opposed to the first episode that occurred after 3 cycles of the palbociclib. After discontinuation of the medication, liver function tests again normalized, suggesting that palbociclib was most likely the causative agent. In addition, the degree of elevated liver enzymes was less severe on re-exposure at the lower dose of 100 mg, which raises the possibility that there could be a dose-dependent association between palbociclib and hepatotoxicity. There have been few case reports of increased liver enzymes associated with palbociclib, and it is only recently that this association has been more recognized. A meta-analysis by Zaw and colleagues has demonstrated that CDK 4/6 inhibitor–based regimens are associated with a higher risk of elevated AST and ALT; however, their relation with dose dependence was not described. In particular, they found that CDK 4/6 inhibitors increased the risk of high-grade, elevated ALT with a relative risk of 4.33 (95% confidence interval, 2.15-8.71; P < .0001). The meta-analysis also included other CDK 4/6 inhibitors such as abemaciclib and ribociclib, which have been more commonly associated with liver toxicity than palbociclib has.⁹ Our case report highlights the specific association between palbociclib and elevated liver enzymes.

In conclusion, this case report illustrates that our patient’s elevated liver enzymes were likely related to palbociclib. This is further supported by the fact that this AE occurred twice, both times after palbociclib exposure. In each instance, liver enzymes normalized after discontinuation of palbociclib. One cannot entirely rule out that fulvestrant might have been the culprit medication, but the patient’s normal hepatic panel for several months after starting fulvestrant suggests that is less likely. This case report is indicative of an uncommon complication in the treatment of metastatic breast cancer, one that is starting to gain more recognition, and we must think of palbociclib as a possible cause of drug-induced liver injury when targeted CDK 4/6–based regimens are used.

About 12.4% of women in the United States will be diagnosed with breast cancer at some point in their lifetime.¹ A percentage of these women will develop metastatic disease and are estimated to have a 5-year survival rate of 22%.² There have been meaningful improvements in survival because of earlier detection and more effective systemic therapies. Patients with hormone-sensitive disease often respond to endocrine therapy, and this frequently represents front-line treatment for these patients, resulting in palliation of symptoms while maintaining quality of life.

However, endocrine resistance inevitably occurs, and a great deal of research has been focused on developing strategies to combat resistance. One mechanism of endocrine resistance is though the Cyclin-dependent kinases 4 and 6 (CDK4/6) complexes. Among the most promising of the strategies to prevent resistance are the CDK4/6 inhibitors. There are now 3 approved CDK4/6 inhibitor drugs that can be used in combination with endocrine therapy, 1 of which can also be used as a single agent. When used in combination with endocrine therapy, the use of CDK 4/6 inhibitors has significantly improved progression-free survival (PFS) in patients with hormone-sensitive HER2-negative metastatic breast cancer by inhibiting cellular division and growth.³ In postmenopausal women, endocrine therapy plus CDK4/6 inhibitors are the preferred first-line regimen for metastatic disease.

Since the approval of palbociclib by the US Food and Drug Administration in 2015, the most common hematologic lab abnormalities are anemia, leukopenia, neutropenia, and thrombocytopenia. The most common nonhematologic adverse events (AEs) are fatigue, infection, nausea, and stomatitis. Hepatic toxicity has not been commonly observed. We report here the case of a 57-year-old woman on palbociclib and fulvestrant who developed significant elevation of liver function tests after starting palbociclib, suggesting a possible drug-induced liver injury from palbociclib.
 

Case presentation and summary

A 57-year-old woman with history of hypothyroidism and hypertension presented in May 2016 with a lump in her right breast and back pain. The lump was biopsied and revealed invasive ductal carcinoma, moderately differentiated, estrogen receptor (ER) positive 100%, progesterone receptor (PR) positive 95%, and HER2 negative. A positron emission tomography (PET)–computed tomography (CT) scan and magnetic resonance imaging showed bone metastasis at several vertebral levels, and the results of a bone biopsy confirmed metastatic adenocarcinoma of breast origin, ER positive 60%, PR positive 40%, and HER2 negative. No liver lesions were seen on imaging, but there was suggestion of fatty liver. She was started on letrozole 2.5 mg daily in July 2016 while undergoing kyphoplasty and subsequent radiation. A restaging PET scan revealed progression of disease on letrozole, with possible new rib lesion and progression in the breast. No liver disease was noted. Therapy was changed to fulvestrant and palbociclib. Fulvestrant was started in March 2017 with standard dosing of 500 mg intramuscular on days 1, 15, and 29, and then once a month thereafter. Her first cycle of palbociclib was started on April 5, dosed at 125 mg by mouth daily for 21 days, followed by 7 days off, repeated every 28 days (all dates hereinafter fell within 2017, unless otherwise stipulated).

Labs checked on April 28 and May 26 were unremarkable. A restaging CT scan of the chest, abdomen, and pelvis was done on June 21 after completion of 3 cycles of fulvestrant and palbociclib. There was no evidence of liver metastases, only the fatty infiltration of the liver that had been seen previously. On June 23, 2017, lab results showed a transaminitis with an alanine aminotransferase (ALT) level of 446 IU/L (reference range 10-33 IU/L) and aspartate aminotransferase (AST) level of 183 IU/L (reference range 0-32 IU/L).

The patient’s liver enzyme levels continued to increase and peaked on July 3 at ALT >700 IU/L and AST at 421 IU/L. Her total bilirubin and alkaline phosphatase levels remained within normal limits. She had received her final dose of fulvestrant on May 31 and had taken her last dose of palbociclib on June 20, 2017. She had no history of elevated liver enzymes or liver disease, although the initial PET scan done at diagnosis had suggested hepatic steatosis. She said she had not recently used antibiotics, alcohol, or over-the-counter medications or supplements. There was no family history of liver problems, inflammatory bowel disease, or gastrointestinal malignancy. The only other medications she had taken recently were denosumab, levothyroxine for hypothyroidism, and amlodipine for hypertension. She was seen by hepatology for evaluation of acute hepatitis. Other etiologies for her elevated liver enzymes were ruled out, and she was diagnosed with a drug-induced liver injury from one of her anticancer medications. Her treatments with fulvestrant and palbociclib were held, and the results of her liver function tests normalized by September 2017.

Fulvestrant was restarted on August 24, and her lab results remained normal through November of that year, when restaging scans showed progression with new axillary adenopathy suspicious for metastasis. Imaging also showed a 1.6-cm hepatic lesion suggestive of a focal area of fat deposition or atypical hemangioma without definitive evidence of metastasis. Follow-up imaging was recommended. She was therefore rechallenged with palbociclib at a reduced dose of 100 mg by mouth daily and received the first dose on November 30. On December 8, repeat labs again showed elevated liver function tests (ALT, 285 IU/L; AST, 112 IU/L). Treatment with palbociclib was discontinued on December 10. Because the patient was not able to tolerate palbociclib, and fulvestrant alone was not controlling the disease, she was started on an alternate endocrine therapy with tamoxifen on December 26. The patient’s liver function tests normalized again by January 2018.
 

Discussion

The use of targeted therapies has changed the landscape of oncologic treatments. Several studies have evaluated the safety and efficacy of palbociclib in combination with endocrine therapy. The Palbociclib Ongoing Trials in the Management of Breast Cancer (PALOMA)-1 study, an open-label, randomized, phase-2 trial involving patients with newly diagnosed metastatic hormone sensitive HER2-negative breast cancer, demonstrated that palbociclib in combination with letrozole was associated with significantly longer PFS than letrozole alone.⁴ These results were later confirmed in the larger PALOMA-2 study, a randomized, double-blind, phase-3 trial that evaluated 666 postmenopausal patients with no prior systemic therapy. In that study, median PFS for the palbociclib–letrozole group was 24.8 months, compared with 14.5 months for the letrozole-alone group (hazard ratio [HR] for disease progression or death, 0.58 [0.46–0.72], P < .001).⁵ The most recent PALOMA-3 study, a phase-3 trial involving 521 patients with advanced hormone receptor–positive, HER2-negative breast cancer that had progressed during initial endocrine therapy, evaluated the efficacy of combined palbociclib and fulvestrant in a randomized, double-blind, placebo-controlled, parallel-group trial. The result was that the palbociclib–fulvestrant combination resulted in longer median PFS of 9.2 months, compared with 3.8 months with fulvestrant alone (P < .001).⁶

These trials also monitored the number of AEs as secondary aims. The most commonly reported AEs in the PALOMA trials for those patients in the palbociclib group were hematologic, with neutropenia being the most common, followed by leukopenia, anemia, and thrombocytopenia. The most common nonhematologic AEs reported in the palbociclib-fulvestrant group were fatigue, nausea, and headache. Elevated liver function tests were a rare but reported AE in 7.2% of the palbociclib-treated patients in the PALOMA-1 study.⁷ In the PALOMA-2 study, ALT and AST elevations were reported as AEs (all grades) in 9.9% and 9.7% of palbociclib-treated patients, respectively.⁵ In the PALOMA-3 study, there was 1 fatal serious AE of hepatic failure with grade 5 disease progression in the palbociclib group; however, the patient’s medical history included progressive liver metastasis and disease progression.⁶ A pooled safety analysis conducted across all PALOMA studies demonstrated that grade 3/4 AST and ALT elevations occurred in 3.3% and 2.3% of palbociclib-treated patients, respectively, again highlighting a reported but rare occurrence.⁸

The patient described in the present case report started on combination fulvestrant and palbociclib after her disease showed progression on letrozole. She developed an increase in transaminases after completing 3 cycles of palbociclib. Liver function tests increased nearly 12 weeks after beginning her first cycle of the CDK 4/6 inhibitor. Staging scans of the patient demonstrated fatty liver. It is not known if her fatty liver contributed to her transminitis; however, her baseline labs showed normal liver function tests, and they did not increase until after therapy with fulvestrant–palbociclib was started. It might have been that her fatty liver caused her to be at higher risk of transaminitis with administration of palbociclib, although we cannot be certain. Her lab results remained normal while she was on fulvestrant alone, and the liver function test results increased only after palbociclib was started, making this drug the more likely culprit.

Both events of increased liver enzymes occurred within a week of the last palbociclib dose; however, we note that hepatotoxicity developed at a faster rate when the patient was rechallenged with palbociclib at a lower dose, with elevated liver function tests increasing 1 week after restarting treatment as opposed to the first episode that occurred after 3 cycles of the palbociclib. After discontinuation of the medication, liver function tests again normalized, suggesting that palbociclib was most likely the causative agent. In addition, the degree of elevated liver enzymes was less severe on re-exposure at the lower dose of 100 mg, which raises the possibility that there could be a dose-dependent association between palbociclib and hepatotoxicity. There have been few case reports of increased liver enzymes associated with palbociclib, and it is only recently that this association has been more recognized. A meta-analysis by Zaw and colleagues has demonstrated that CDK 4/6 inhibitor–based regimens are associated with a higher risk of elevated AST and ALT; however, their relation with dose dependence was not described. In particular, they found that CDK 4/6 inhibitors increased the risk of high-grade, elevated ALT with a relative risk of 4.33 (95% confidence interval, 2.15-8.71; P < .0001). The meta-analysis also included other CDK 4/6 inhibitors such as abemaciclib and ribociclib, which have been more commonly associated with liver toxicity than palbociclib has.⁹ Our case report highlights the specific association between palbociclib and elevated liver enzymes.

In conclusion, this case report illustrates that our patient’s elevated liver enzymes were likely related to palbociclib. This is further supported by the fact that this AE occurred twice, both times after palbociclib exposure. In each instance, liver enzymes normalized after discontinuation of palbociclib. One cannot entirely rule out that fulvestrant might have been the culprit medication, but the patient’s normal hepatic panel for several months after starting fulvestrant suggests that is less likely. This case report is indicative of an uncommon complication in the treatment of metastatic breast cancer, one that is starting to gain more recognition, and we must think of palbociclib as a possible cause of drug-induced liver injury when targeted CDK 4/6–based regimens are used.

Epidermal growth factor receptor antibodies (EGFR) such as cetuximab have been approved for use as first-line management as well as salvage therapy for head and neck and colorectal cancers. Among the most common expected toxicity is a cutaneous eruption described as acneiform. The presence of a rash has been postulated to predict a more favorable treatment outcome for cancers of the head and neck¹ but not for colorectum.² With more severe drug reactions, patients may require a treatment break, which has been shown to reduce locoregional control and survival, particularly in patients with head and neck cancer.³ This has prompted clinicians to affect rapid therapy to reverse the drug eruption. Given the controversy around rapid and effective reversal of this drug reaction, this report aims to address the current status of clinical management using an actual patient vignette.

Case presentation and summary

The patient was a 57-year-old white man who had been diagnosed with stage 4 T4N0M1 grade 3 cutaneous squamous cell carcinoma (SCC) of the right postauricular soft tissues, with erosion into the right mastoid and biopsy-proven metastatic disease involving the contralateral left supraclavicular fossa and bilateral lungs. His disease became chemotherapy-refractory, and he was referred for palliative local therapy to the base of skull. Because of the size of the tumor (4 cm × 5 cm), he was considered for sensitizing chemotherapy, but cisplatin was not appropriate because of chronic hearing loss.⁴ The patient was recommended sensitizing doses of cetuximab. This EGFR antibody has been shown to offer similar benefits to those seen with cisplatin in the definitive management of head and neck SCC.⁵

The standard loading dose of cetuximab was given at 400 mg/m² intravenously (IV). The following week, the sensitizing dose of 250 mg/m² IV was given along with daily radiotherapy to the target volumes. The weekly dose of cetuximab continued at 250 mg/m². The radiotherapy prescription was for 6,000 cGy in 200 cGy daily fractions, encompassing the gross tumor volume as identified on a computed-tomographic scan with 3-mm cuts. We used a noncoplanar arc radiotherapy beam arrangement because it inherently spreads the dose over a larger volume of normal tissue while conformally delivering its largest dose to the gross tumor volume. As such, a volume of the patient’s oropharynx and oral cavity was included within the radiotherapy dose penumbra. After receiving 3 weekly doses of cetuximab (1 loading dose and 2 weekly sensitizing doses) and 2,000 cGy of radiotherapy, the patient developed a robust grade 2 cutaneous eruption delimited to the face, with few scattered lesions on the upper anterior chest. He was seen in the medical oncology department and was prescribed doxycycline 100 mg orally twice daily and topical clindamycin 2% ointment twice daily.

In the radiation oncology clinic, his drug therapy was manipulated. His cetuximab cutaneous reaction was a grade 2, manifested by moderate erythema with nonconfluent moist desquamation. Because of concern that the patient would develop oral candida, which would further delay his therapy, the oral and topical antibiotics were discontinued, as was the oral prednisone. He was prescribed triamcinolone cream 0.1% to be applied to the facial and few chest wall areas twice daily and an oncology mouth rinse to address early nonconfluent mucositis. The accompanying images show the extent of the patient’s cetuximab cutaneous reaction at baseline before treatment initiation (Figure 1), at 4 days after the intervention (Figure 2), and again at 6 days after the intervention (Figure 3). The patient consented to having his photographs taken and understood that they would be used for educational and research publication purposes.

Discussion

A cetuximab-induced rash is common. In a 2011 meta-analysis quantifying grades 1 to 4 in severity, about 75% of patients treated with an EGFR inhibitor experienced a rash. Most of the rashes were lower than grade 3, and the drug was either dose-reduced or temporarily held, but it was not generally discontinued.⁶ Of note is that in a nonselected survey of medical oncologists who were prescribing cetuximab, 76% reported holding the drug owing to rash severity, 60% reported dose reductions for a drug rash, and 32% reported changing the drug because of rash severity.⁷

In the initial pharmaceutical registration trial, 76% to 88% of patients who received cetuximab developed a rash, 17% of which were at least grade 3. The pharma recommendations for managing the drug rash include a drug delay for up to 2 weeks for a rash of grade 3 or less and to terminate use of the drug if there is no clinical improvement after 2 weeks.⁸ Biopsies of the rash confirm a suppurative inflammatory reaction separate from an infectious acne reaction,⁹ resulting in a recommendation to treat with topical steroid therapy. In some circumstances, the drug reaction can become infected or involve the paronychia, often related to Staphylococcus aureus.¹⁰ Despite what would otherwise be a problem addressed by anti-inflammatory medical therapy, the clinical appearance of the rash marked by pustules, coupled with the relative immunosuppressed state of a cancer patient, has prompted medical oncologists to prescribe antibiotic therapy.

To address the many single-institutional reports on management of the EGFR rash, several guidelines have been published. The earliest guideline – after a report that concurrent cetuximab and radiotherapy was superior to radiotherapy alone in locally advanced head and neck cancer, which documented a 23% incidence of at least grade 3 cutaneous toxicity in the cetuximab arm¹ – attempted to score the severity of the rash according to the National Cancer Institute’s (NCI) Common Terminology Criteria for Adverse Events (CTCAE). Under those criteria, the authors defined grade 2 toxicity as moderate to brisk erythema with patchy moist desquamation, mostly confined to skin folds and creases. Grade 3 toxicity was described as moist desquamation other than skin folds and creases with bleeding induced by minor trauma, and grade 4 skin toxicity was defined as skin necrosis or ulceration of full thickness dermis with spontaneous bleeding from the involved site. The authors went on to describe a grade-related treatment algorithm that included gently washing the skin, keeping it dry, and using topical anti-inflammatory agents, including steroids. Antibiotics should be used in the presence of a suspected infection after culturing the area, and grade 4 toxicity should be referred to a wound care center.¹¹

In a consensus statement from the National Comprehensive Cancer Network, the authors noted that most management recommendations were anecdotal. They recommended against the use of astringents and other drying agents because they exacerbate pain. The ultimate choice of topical steroids or antibiotics was based entirely on subjective judgement given the absence of prospective data.¹²

A Spanish consensus conference report argued against any prophylaxis against a skin reaction, other than keeping the skin clean and dry.¹³ The authors of the report recommended against washing the affected skin more than twice a day to avoid excess drying, and they advocated for moisturizers and debridement of skin crusting with hydrogels to reduce superinfection and bleeding.¹³ The authors also noted that some guidelines have suggested that topical steroids might exacerbate a skin rash,¹⁴ but they concluded that topical steroids are beneficial as long as they are used for less than 2 weeks. Any use of antibiotics should be based on clear evidence of an infection.¹³

In the first modification of the NCI’s CTCAE rash grading scale, an international panel addressed the increasing number of reports in the literature suggesting that the previous toxicity scale was possibly inadequate in its recommendations for appropriate treatment. The initial scale had defined only the skin reaction and not what therapy should be administered; therefore, in the update, the descriptions for grades 1 and 2 toxicity remained unchanged, but oral antibiotics were recommended for grade 3 lesion, and parenteral antibiotics with skin grafting were required with grade 4 toxicity.¹⁵

An Asian expert panel suggested modifying the bioradiation dermatitis scale, defining a grade 3 dermatitis as >50% moist desquamation of the involved field with formation of confluent lesions because of treatment. They recommended both topical and oral therapy, wound care, and possible hospitalization in severe cases. The panel suggested topical and systemic steroids and antibiotics.¹⁶

Finally, in an Italian consensus report, the members again modified the skin toxicity grading and were notably more aggressive in terms of their management recommendations. They defined grade 2 toxicity as pustules or papules covering 10% to 30% of the body surface area, with potential pruritus or tenderness. They also noted the psychosocial impact of skin toxicities on patients and the limits to their activities of daily living. They recommended vitamin K1 (menadione) cream, topical antibiotics, topical intermediate potency steroids, and oral antibiotic therapy for up to 4 weeks for grade 2 toxicity. Despite this aggressive treatment course, the authors admitted that the utility of topical steroids and antibiotics was unknown. They defined grade 3 toxicity as pustules or papules covering more than 30% of the body surface area, with signs of possible pruritus and tenderness. Activities of daily living and self-care were affected, and there was evidence of a superinfection. The panel suggested use of antibiotics pending culture results, oral prednisone, antihistamines, and oral analgesics. Topical therapy was not included.¹⁷ It is noteworthy that only the Italian panel recommended the use of vitamin K1 cream. In a prospective randomized, double-blinded, placebo-controlled phase 2 trial of 30 patients, menadione exhibited no clinical benefit in terms of reducing the severity of cetuximab skin lesions.¹⁸

Figure 4 illustrates our institutional approach to treating cetuximab rash based on a combination of the Spanish and NCI approaches.

Epidermal growth factor receptor antibodies (EGFR) such as cetuximab have been approved for use as first-line management as well as salvage therapy for head and neck and colorectal cancers. Among the most common expected toxicity is a cutaneous eruption described as acneiform. The presence of a rash has been postulated to predict a more favorable treatment outcome for cancers of the head and neck¹ but not for colorectum.² With more severe drug reactions, patients may require a treatment break, which has been shown to reduce locoregional control and survival, particularly in patients with head and neck cancer.³ This has prompted clinicians to affect rapid therapy to reverse the drug eruption. Given the controversy around rapid and effective reversal of this drug reaction, this report aims to address the current status of clinical management using an actual patient vignette.

Case presentation and summary

The patient was a 57-year-old white man who had been diagnosed with stage 4 T4N0M1 grade 3 cutaneous squamous cell carcinoma (SCC) of the right postauricular soft tissues, with erosion into the right mastoid and biopsy-proven metastatic disease involving the contralateral left supraclavicular fossa and bilateral lungs. His disease became chemotherapy-refractory, and he was referred for palliative local therapy to the base of skull. Because of the size of the tumor (4 cm × 5 cm), he was considered for sensitizing chemotherapy, but cisplatin was not appropriate because of chronic hearing loss.⁴ The patient was recommended sensitizing doses of cetuximab. This EGFR antibody has been shown to offer similar benefits to those seen with cisplatin in the definitive management of head and neck SCC.⁵

The standard loading dose of cetuximab was given at 400 mg/m² intravenously (IV). The following week, the sensitizing dose of 250 mg/m² IV was given along with daily radiotherapy to the target volumes. The weekly dose of cetuximab continued at 250 mg/m². The radiotherapy prescription was for 6,000 cGy in 200 cGy daily fractions, encompassing the gross tumor volume as identified on a computed-tomographic scan with 3-mm cuts. We used a noncoplanar arc radiotherapy beam arrangement because it inherently spreads the dose over a larger volume of normal tissue while conformally delivering its largest dose to the gross tumor volume. As such, a volume of the patient’s oropharynx and oral cavity was included within the radiotherapy dose penumbra. After receiving 3 weekly doses of cetuximab (1 loading dose and 2 weekly sensitizing doses) and 2,000 cGy of radiotherapy, the patient developed a robust grade 2 cutaneous eruption delimited to the face, with few scattered lesions on the upper anterior chest. He was seen in the medical oncology department and was prescribed doxycycline 100 mg orally twice daily and topical clindamycin 2% ointment twice daily.

In the radiation oncology clinic, his drug therapy was manipulated. His cetuximab cutaneous reaction was a grade 2, manifested by moderate erythema with nonconfluent moist desquamation. Because of concern that the patient would develop oral candida, which would further delay his therapy, the oral and topical antibiotics were discontinued, as was the oral prednisone. He was prescribed triamcinolone cream 0.1% to be applied to the facial and few chest wall areas twice daily and an oncology mouth rinse to address early nonconfluent mucositis. The accompanying images show the extent of the patient’s cetuximab cutaneous reaction at baseline before treatment initiation (Figure 1), at 4 days after the intervention (Figure 2), and again at 6 days after the intervention (Figure 3). The patient consented to having his photographs taken and understood that they would be used for educational and research publication purposes.

Discussion

A cetuximab-induced rash is common. In a 2011 meta-analysis quantifying grades 1 to 4 in severity, about 75% of patients treated with an EGFR inhibitor experienced a rash. Most of the rashes were lower than grade 3, and the drug was either dose-reduced or temporarily held, but it was not generally discontinued.⁶ Of note is that in a nonselected survey of medical oncologists who were prescribing cetuximab, 76% reported holding the drug owing to rash severity, 60% reported dose reductions for a drug rash, and 32% reported changing the drug because of rash severity.⁷

In the initial pharmaceutical registration trial, 76% to 88% of patients who received cetuximab developed a rash, 17% of which were at least grade 3. The pharma recommendations for managing the drug rash include a drug delay for up to 2 weeks for a rash of grade 3 or less and to terminate use of the drug if there is no clinical improvement after 2 weeks.⁸ Biopsies of the rash confirm a suppurative inflammatory reaction separate from an infectious acne reaction,⁹ resulting in a recommendation to treat with topical steroid therapy. In some circumstances, the drug reaction can become infected or involve the paronychia, often related to Staphylococcus aureus.¹⁰ Despite what would otherwise be a problem addressed by anti-inflammatory medical therapy, the clinical appearance of the rash marked by pustules, coupled with the relative immunosuppressed state of a cancer patient, has prompted medical oncologists to prescribe antibiotic therapy.

To address the many single-institutional reports on management of the EGFR rash, several guidelines have been published. The earliest guideline – after a report that concurrent cetuximab and radiotherapy was superior to radiotherapy alone in locally advanced head and neck cancer, which documented a 23% incidence of at least grade 3 cutaneous toxicity in the cetuximab arm¹ – attempted to score the severity of the rash according to the National Cancer Institute’s (NCI) Common Terminology Criteria for Adverse Events (CTCAE). Under those criteria, the authors defined grade 2 toxicity as moderate to brisk erythema with patchy moist desquamation, mostly confined to skin folds and creases. Grade 3 toxicity was described as moist desquamation other than skin folds and creases with bleeding induced by minor trauma, and grade 4 skin toxicity was defined as skin necrosis or ulceration of full thickness dermis with spontaneous bleeding from the involved site. The authors went on to describe a grade-related treatment algorithm that included gently washing the skin, keeping it dry, and using topical anti-inflammatory agents, including steroids. Antibiotics should be used in the presence of a suspected infection after culturing the area, and grade 4 toxicity should be referred to a wound care center.¹¹

In a consensus statement from the National Comprehensive Cancer Network, the authors noted that most management recommendations were anecdotal. They recommended against the use of astringents and other drying agents because they exacerbate pain. The ultimate choice of topical steroids or antibiotics was based entirely on subjective judgement given the absence of prospective data.¹²

A Spanish consensus conference report argued against any prophylaxis against a skin reaction, other than keeping the skin clean and dry.¹³ The authors of the report recommended against washing the affected skin more than twice a day to avoid excess drying, and they advocated for moisturizers and debridement of skin crusting with hydrogels to reduce superinfection and bleeding.¹³ The authors also noted that some guidelines have suggested that topical steroids might exacerbate a skin rash,¹⁴ but they concluded that topical steroids are beneficial as long as they are used for less than 2 weeks. Any use of antibiotics should be based on clear evidence of an infection.¹³

In the first modification of the NCI’s CTCAE rash grading scale, an international panel addressed the increasing number of reports in the literature suggesting that the previous toxicity scale was possibly inadequate in its recommendations for appropriate treatment. The initial scale had defined only the skin reaction and not what therapy should be administered; therefore, in the update, the descriptions for grades 1 and 2 toxicity remained unchanged, but oral antibiotics were recommended for grade 3 lesion, and parenteral antibiotics with skin grafting were required with grade 4 toxicity.¹⁵

An Asian expert panel suggested modifying the bioradiation dermatitis scale, defining a grade 3 dermatitis as >50% moist desquamation of the involved field with formation of confluent lesions because of treatment. They recommended both topical and oral therapy, wound care, and possible hospitalization in severe cases. The panel suggested topical and systemic steroids and antibiotics.¹⁶

Finally, in an Italian consensus report, the members again modified the skin toxicity grading and were notably more aggressive in terms of their management recommendations. They defined grade 2 toxicity as pustules or papules covering 10% to 30% of the body surface area, with potential pruritus or tenderness. They also noted the psychosocial impact of skin toxicities on patients and the limits to their activities of daily living. They recommended vitamin K1 (menadione) cream, topical antibiotics, topical intermediate potency steroids, and oral antibiotic therapy for up to 4 weeks for grade 2 toxicity. Despite this aggressive treatment course, the authors admitted that the utility of topical steroids and antibiotics was unknown. They defined grade 3 toxicity as pustules or papules covering more than 30% of the body surface area, with signs of possible pruritus and tenderness. Activities of daily living and self-care were affected, and there was evidence of a superinfection. The panel suggested use of antibiotics pending culture results, oral prednisone, antihistamines, and oral analgesics. Topical therapy was not included.¹⁷ It is noteworthy that only the Italian panel recommended the use of vitamin K1 cream. In a prospective randomized, double-blinded, placebo-controlled phase 2 trial of 30 patients, menadione exhibited no clinical benefit in terms of reducing the severity of cetuximab skin lesions.¹⁸

Figure 4 illustrates our institutional approach to treating cetuximab rash based on a combination of the Spanish and NCI approaches.

Epidermal growth factor receptor antibodies (EGFR) such as cetuximab have been approved for use as first-line management as well as salvage therapy for head and neck and colorectal cancers. Among the most common expected toxicity is a cutaneous eruption described as acneiform. The presence of a rash has been postulated to predict a more favorable treatment outcome for cancers of the head and neck¹ but not for colorectum.² With more severe drug reactions, patients may require a treatment break, which has been shown to reduce locoregional control and survival, particularly in patients with head and neck cancer.³ This has prompted clinicians to affect rapid therapy to reverse the drug eruption. Given the controversy around rapid and effective reversal of this drug reaction, this report aims to address the current status of clinical management using an actual patient vignette.

Case presentation and summary

The patient was a 57-year-old white man who had been diagnosed with stage 4 T4N0M1 grade 3 cutaneous squamous cell carcinoma (SCC) of the right postauricular soft tissues, with erosion into the right mastoid and biopsy-proven metastatic disease involving the contralateral left supraclavicular fossa and bilateral lungs. His disease became chemotherapy-refractory, and he was referred for palliative local therapy to the base of skull. Because of the size of the tumor (4 cm × 5 cm), he was considered for sensitizing chemotherapy, but cisplatin was not appropriate because of chronic hearing loss.⁴ The patient was recommended sensitizing doses of cetuximab. This EGFR antibody has been shown to offer similar benefits to those seen with cisplatin in the definitive management of head and neck SCC.⁵

The standard loading dose of cetuximab was given at 400 mg/m² intravenously (IV). The following week, the sensitizing dose of 250 mg/m² IV was given along with daily radiotherapy to the target volumes. The weekly dose of cetuximab continued at 250 mg/m². The radiotherapy prescription was for 6,000 cGy in 200 cGy daily fractions, encompassing the gross tumor volume as identified on a computed-tomographic scan with 3-mm cuts. We used a noncoplanar arc radiotherapy beam arrangement because it inherently spreads the dose over a larger volume of normal tissue while conformally delivering its largest dose to the gross tumor volume. As such, a volume of the patient’s oropharynx and oral cavity was included within the radiotherapy dose penumbra. After receiving 3 weekly doses of cetuximab (1 loading dose and 2 weekly sensitizing doses) and 2,000 cGy of radiotherapy, the patient developed a robust grade 2 cutaneous eruption delimited to the face, with few scattered lesions on the upper anterior chest. He was seen in the medical oncology department and was prescribed doxycycline 100 mg orally twice daily and topical clindamycin 2% ointment twice daily.

In the radiation oncology clinic, his drug therapy was manipulated. His cetuximab cutaneous reaction was a grade 2, manifested by moderate erythema with nonconfluent moist desquamation. Because of concern that the patient would develop oral candida, which would further delay his therapy, the oral and topical antibiotics were discontinued, as was the oral prednisone. He was prescribed triamcinolone cream 0.1% to be applied to the facial and few chest wall areas twice daily and an oncology mouth rinse to address early nonconfluent mucositis. The accompanying images show the extent of the patient’s cetuximab cutaneous reaction at baseline before treatment initiation (Figure 1), at 4 days after the intervention (Figure 2), and again at 6 days after the intervention (Figure 3). The patient consented to having his photographs taken and understood that they would be used for educational and research publication purposes.

Discussion

A cetuximab-induced rash is common. In a 2011 meta-analysis quantifying grades 1 to 4 in severity, about 75% of patients treated with an EGFR inhibitor experienced a rash. Most of the rashes were lower than grade 3, and the drug was either dose-reduced or temporarily held, but it was not generally discontinued.⁶ Of note is that in a nonselected survey of medical oncologists who were prescribing cetuximab, 76% reported holding the drug owing to rash severity, 60% reported dose reductions for a drug rash, and 32% reported changing the drug because of rash severity.⁷

In the initial pharmaceutical registration trial, 76% to 88% of patients who received cetuximab developed a rash, 17% of which were at least grade 3. The pharma recommendations for managing the drug rash include a drug delay for up to 2 weeks for a rash of grade 3 or less and to terminate use of the drug if there is no clinical improvement after 2 weeks.⁸ Biopsies of the rash confirm a suppurative inflammatory reaction separate from an infectious acne reaction,⁹ resulting in a recommendation to treat with topical steroid therapy. In some circumstances, the drug reaction can become infected or involve the paronychia, often related to Staphylococcus aureus.¹⁰ Despite what would otherwise be a problem addressed by anti-inflammatory medical therapy, the clinical appearance of the rash marked by pustules, coupled with the relative immunosuppressed state of a cancer patient, has prompted medical oncologists to prescribe antibiotic therapy.

To address the many single-institutional reports on management of the EGFR rash, several guidelines have been published. The earliest guideline – after a report that concurrent cetuximab and radiotherapy was superior to radiotherapy alone in locally advanced head and neck cancer, which documented a 23% incidence of at least grade 3 cutaneous toxicity in the cetuximab arm¹ – attempted to score the severity of the rash according to the National Cancer Institute’s (NCI) Common Terminology Criteria for Adverse Events (CTCAE). Under those criteria, the authors defined grade 2 toxicity as moderate to brisk erythema with patchy moist desquamation, mostly confined to skin folds and creases. Grade 3 toxicity was described as moist desquamation other than skin folds and creases with bleeding induced by minor trauma, and grade 4 skin toxicity was defined as skin necrosis or ulceration of full thickness dermis with spontaneous bleeding from the involved site. The authors went on to describe a grade-related treatment algorithm that included gently washing the skin, keeping it dry, and using topical anti-inflammatory agents, including steroids. Antibiotics should be used in the presence of a suspected infection after culturing the area, and grade 4 toxicity should be referred to a wound care center.¹¹

In a consensus statement from the National Comprehensive Cancer Network, the authors noted that most management recommendations were anecdotal. They recommended against the use of astringents and other drying agents because they exacerbate pain. The ultimate choice of topical steroids or antibiotics was based entirely on subjective judgement given the absence of prospective data.¹²

A Spanish consensus conference report argued against any prophylaxis against a skin reaction, other than keeping the skin clean and dry.¹³ The authors of the report recommended against washing the affected skin more than twice a day to avoid excess drying, and they advocated for moisturizers and debridement of skin crusting with hydrogels to reduce superinfection and bleeding.¹³ The authors also noted that some guidelines have suggested that topical steroids might exacerbate a skin rash,¹⁴ but they concluded that topical steroids are beneficial as long as they are used for less than 2 weeks. Any use of antibiotics should be based on clear evidence of an infection.¹³

In the first modification of the NCI’s CTCAE rash grading scale, an international panel addressed the increasing number of reports in the literature suggesting that the previous toxicity scale was possibly inadequate in its recommendations for appropriate treatment. The initial scale had defined only the skin reaction and not what therapy should be administered; therefore, in the update, the descriptions for grades 1 and 2 toxicity remained unchanged, but oral antibiotics were recommended for grade 3 lesion, and parenteral antibiotics with skin grafting were required with grade 4 toxicity.¹⁵

An Asian expert panel suggested modifying the bioradiation dermatitis scale, defining a grade 3 dermatitis as >50% moist desquamation of the involved field with formation of confluent lesions because of treatment. They recommended both topical and oral therapy, wound care, and possible hospitalization in severe cases. The panel suggested topical and systemic steroids and antibiotics.¹⁶

Finally, in an Italian consensus report, the members again modified the skin toxicity grading and were notably more aggressive in terms of their management recommendations. They defined grade 2 toxicity as pustules or papules covering 10% to 30% of the body surface area, with potential pruritus or tenderness. They also noted the psychosocial impact of skin toxicities on patients and the limits to their activities of daily living. They recommended vitamin K1 (menadione) cream, topical antibiotics, topical intermediate potency steroids, and oral antibiotic therapy for up to 4 weeks for grade 2 toxicity. Despite this aggressive treatment course, the authors admitted that the utility of topical steroids and antibiotics was unknown. They defined grade 3 toxicity as pustules or papules covering more than 30% of the body surface area, with signs of possible pruritus and tenderness. Activities of daily living and self-care were affected, and there was evidence of a superinfection. The panel suggested use of antibiotics pending culture results, oral prednisone, antihistamines, and oral analgesics. Topical therapy was not included.¹⁷ It is noteworthy that only the Italian panel recommended the use of vitamin K1 cream. In a prospective randomized, double-blinded, placebo-controlled phase 2 trial of 30 patients, menadione exhibited no clinical benefit in terms of reducing the severity of cetuximab skin lesions.¹⁸

Figure 4 illustrates our institutional approach to treating cetuximab rash based on a combination of the Spanish and NCI approaches.