Case Reports

THE CASE

A 71-year-old woman came to our clinic with a left subconjunctival hemorrhage. She had a history of atrial flutter and had received a liver transplant approximately 10 years ago. The patient reported having a procedure 2 weeks before her visit with us to remove a basal cell carcinoma on her lower left eyelid, but had no recent changes in vision or physical damage to the eye.

In the past year, she had been started on dabigatran 150 mg twice daily after developing symptomatic atrial fibrillation. Our patient had also been receiving tacrolimus 3 mg twice daily since her transplant. Other medications she was taking included hydroxychloroquine 200 mg/d for rheumatoid arthritis, propafenone 225 mg twice daily for atrial fibrillation, valsartan 80 mg/d for hypertension, and ranitidine 150 mg/d for reflux.

Venipuncture coagulation tests showed a partial thromboplastin time (PTT) of 75.1 seconds, a prothrombin time (PT) of 46.1 seconds, and an elevated international normalized ratio (INR) of 4.5 (normal range: 0.8-1.2). Point-of-care INR results were not obtained.

A complete blood count (CBC) was unremarkable with the exception of a low platelet count and high red blood cell distribution width (RDW). Our patient’s aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were both within normal limits.

Kidney function tests told another story. The patient’s serum creatinine (SCr) and blood urea nitrogen (BUN) levels were elevated (1.54 mg/dL and 29 mg/dL, respectively) and her creatinine clearance (CrCl; 30.2 mL/min) suggested moderate to severe renal dysfunction.

The patient’s CHADS₂ score was calculated as 1, suggesting she had a low-to-moderate risk of stroke.

THE DIAGNOSIS

Our patient had a left subconjunctival hemorrhage and an elevated venipuncture INR. Based on her renal dysfunction, we suspected that her elevated INR was likely due to an excessive dose of dabigatran, as well as an interaction between dabigatran and tacrolimus.

DISCUSSION

Dabigatran is an oral direct thrombin inhibitor approved for the prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation. An important advantage of dabigatran compared to warfarin is that the fixed-dose regimen does not require routine anticoagulation monitoring. In cases where anticoagulation monitoring is needed, PTT is the preferred method.¹

While PT and INR have generally not been shown to accurately reflect the degree of anticoagulation with dabigatran at therapeutic doses, there have been in vitro reports of elevated INRs with supratherapeutic dabigatran levels.^2,3 At a typical peak therapeutic dabigatran concentration of approximately 184 ng/mL, the INR generally ranged from 1.1 to 1.7.² However, at a dabigatran concentration of 1000 ng/mL, the INR was elevated to 4.5,^2,3 which is the same venipuncture INR recorded in our patient. While there have been published reports of falsely elevated point-of-care INR results compared to corresponding venipuncture INR results in patients taking dabigatran,^4,5 a literature review found only a case of an elevated venipuncture INR in an end-stage renal disease patient receiving hemodialysis.⁶

Concurrent use of any P-gp inhibitor (such as tacrolimus) and dabigatran is contraindicated in patients with severe renal dysfunction.

In the case noted above, as well as our patient, an accumulation of dabigatran due to the patient’s renal dysfunction likely resulted in high plasma concentrations and therefore an elevated venipuncture INR. The elimination half-life for dabigatran is approximately 14 hours in patients with normal renal function; in a patient with severe renal impairment, the half-life can be up to 28 hours.⁷ Our patient’s CrCl at the time of presentation was 30.2 mL/min, which indicated moderate to severe renal dysfunction. Based on dabigatran prescribing recommendations, a dose adjustment to 75 mg bid might be appropriate.¹ (Our patient was taking 150 mg bid.)

We do not believe our patient’s elevated INR was due to her liver transplant because there were no clinical signs of liver dysfunction. A more likely contributing factor was a drug interaction with tacrolimus. Dabigatran is a moderate affinity P-glycoprotein (P-gp) substrate and tacrolimus is both a P-gp substrate and inhibitor. While an interaction between tacrolimus and dabigatran has not been studied directly, concurrent use of any P-gp inhibitor and dabigatran is contraindicated in patients with severe renal dysfunction (CrCl: 15-30 mL/min).¹ For these theoretical interactions, the Drug Interaction Probability Scale (DIPS) has been developed.⁸ In our patient’s case, the calculated DIPS score of 5 suggests a probable interaction, likely due to P-gp inhibition. The other medications our patient was taking did not have this interaction and were unlikely to contribute to the elevated INR and subconjunctival hemorrhage.

Our patient was instructed to stop taking dabigatran and return in 3 days for additional lab tests. At her follow-up visit, the lab results were PTT, 34.3 seconds; PT, 11.6 seconds; and venipuncture INR, 1.1. Her CBC was unremarkable and unchanged. Shortly after the follow-up visit, our patient was assessed by her cardiologist. Due to her renal dysfunction, risk of bleeding, and relatively low CHADS₂ score, the cardiologist decided to discontinue dabigatran and start her on aspirin.

THE TAKEAWAY

Dabigatran may cause elevated INR levels in patients with renal dysfunction and/or those taking other medications that could interact with dabigatran. Concurrent use of any P-gp inhibitor (such as tacrolimus) and dabigatran is contraindicated in patients with severe renal dysfunction. Despite the lack of required routine laboratory monitoring, renal function and drug interactions associated with dabigatran therapy should be monitored closely.

THE CASE

A 71-year-old woman came to our clinic with a left subconjunctival hemorrhage. She had a history of atrial flutter and had received a liver transplant approximately 10 years ago. The patient reported having a procedure 2 weeks before her visit with us to remove a basal cell carcinoma on her lower left eyelid, but had no recent changes in vision or physical damage to the eye.

In the past year, she had been started on dabigatran 150 mg twice daily after developing symptomatic atrial fibrillation. Our patient had also been receiving tacrolimus 3 mg twice daily since her transplant. Other medications she was taking included hydroxychloroquine 200 mg/d for rheumatoid arthritis, propafenone 225 mg twice daily for atrial fibrillation, valsartan 80 mg/d for hypertension, and ranitidine 150 mg/d for reflux.

Venipuncture coagulation tests showed a partial thromboplastin time (PTT) of 75.1 seconds, a prothrombin time (PT) of 46.1 seconds, and an elevated international normalized ratio (INR) of 4.5 (normal range: 0.8-1.2). Point-of-care INR results were not obtained.

A complete blood count (CBC) was unremarkable with the exception of a low platelet count and high red blood cell distribution width (RDW). Our patient’s aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were both within normal limits.

Kidney function tests told another story. The patient’s serum creatinine (SCr) and blood urea nitrogen (BUN) levels were elevated (1.54 mg/dL and 29 mg/dL, respectively) and her creatinine clearance (CrCl; 30.2 mL/min) suggested moderate to severe renal dysfunction.

The patient’s CHADS₂ score was calculated as 1, suggesting she had a low-to-moderate risk of stroke.

THE DIAGNOSIS

Our patient had a left subconjunctival hemorrhage and an elevated venipuncture INR. Based on her renal dysfunction, we suspected that her elevated INR was likely due to an excessive dose of dabigatran, as well as an interaction between dabigatran and tacrolimus.

DISCUSSION

Dabigatran is an oral direct thrombin inhibitor approved for the prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation. An important advantage of dabigatran compared to warfarin is that the fixed-dose regimen does not require routine anticoagulation monitoring. In cases where anticoagulation monitoring is needed, PTT is the preferred method.¹

While PT and INR have generally not been shown to accurately reflect the degree of anticoagulation with dabigatran at therapeutic doses, there have been in vitro reports of elevated INRs with supratherapeutic dabigatran levels.^2,3 At a typical peak therapeutic dabigatran concentration of approximately 184 ng/mL, the INR generally ranged from 1.1 to 1.7.² However, at a dabigatran concentration of 1000 ng/mL, the INR was elevated to 4.5,^2,3 which is the same venipuncture INR recorded in our patient. While there have been published reports of falsely elevated point-of-care INR results compared to corresponding venipuncture INR results in patients taking dabigatran,^4,5 a literature review found only a case of an elevated venipuncture INR in an end-stage renal disease patient receiving hemodialysis.⁶

Concurrent use of any P-gp inhibitor (such as tacrolimus) and dabigatran is contraindicated in patients with severe renal dysfunction.

In the case noted above, as well as our patient, an accumulation of dabigatran due to the patient’s renal dysfunction likely resulted in high plasma concentrations and therefore an elevated venipuncture INR. The elimination half-life for dabigatran is approximately 14 hours in patients with normal renal function; in a patient with severe renal impairment, the half-life can be up to 28 hours.⁷ Our patient’s CrCl at the time of presentation was 30.2 mL/min, which indicated moderate to severe renal dysfunction. Based on dabigatran prescribing recommendations, a dose adjustment to 75 mg bid might be appropriate.¹ (Our patient was taking 150 mg bid.)

We do not believe our patient’s elevated INR was due to her liver transplant because there were no clinical signs of liver dysfunction. A more likely contributing factor was a drug interaction with tacrolimus. Dabigatran is a moderate affinity P-glycoprotein (P-gp) substrate and tacrolimus is both a P-gp substrate and inhibitor. While an interaction between tacrolimus and dabigatran has not been studied directly, concurrent use of any P-gp inhibitor and dabigatran is contraindicated in patients with severe renal dysfunction (CrCl: 15-30 mL/min).¹ For these theoretical interactions, the Drug Interaction Probability Scale (DIPS) has been developed.⁸ In our patient’s case, the calculated DIPS score of 5 suggests a probable interaction, likely due to P-gp inhibition. The other medications our patient was taking did not have this interaction and were unlikely to contribute to the elevated INR and subconjunctival hemorrhage.

Our patient was instructed to stop taking dabigatran and return in 3 days for additional lab tests. At her follow-up visit, the lab results were PTT, 34.3 seconds; PT, 11.6 seconds; and venipuncture INR, 1.1. Her CBC was unremarkable and unchanged. Shortly after the follow-up visit, our patient was assessed by her cardiologist. Due to her renal dysfunction, risk of bleeding, and relatively low CHADS₂ score, the cardiologist decided to discontinue dabigatran and start her on aspirin.

THE TAKEAWAY

Dabigatran may cause elevated INR levels in patients with renal dysfunction and/or those taking other medications that could interact with dabigatran. Concurrent use of any P-gp inhibitor (such as tacrolimus) and dabigatran is contraindicated in patients with severe renal dysfunction. Despite the lack of required routine laboratory monitoring, renal function and drug interactions associated with dabigatran therapy should be monitored closely.

THE CASE

A 71-year-old woman came to our clinic with a left subconjunctival hemorrhage. She had a history of atrial flutter and had received a liver transplant approximately 10 years ago. The patient reported having a procedure 2 weeks before her visit with us to remove a basal cell carcinoma on her lower left eyelid, but had no recent changes in vision or physical damage to the eye.

In the past year, she had been started on dabigatran 150 mg twice daily after developing symptomatic atrial fibrillation. Our patient had also been receiving tacrolimus 3 mg twice daily since her transplant. Other medications she was taking included hydroxychloroquine 200 mg/d for rheumatoid arthritis, propafenone 225 mg twice daily for atrial fibrillation, valsartan 80 mg/d for hypertension, and ranitidine 150 mg/d for reflux.

Venipuncture coagulation tests showed a partial thromboplastin time (PTT) of 75.1 seconds, a prothrombin time (PT) of 46.1 seconds, and an elevated international normalized ratio (INR) of 4.5 (normal range: 0.8-1.2). Point-of-care INR results were not obtained.

A complete blood count (CBC) was unremarkable with the exception of a low platelet count and high red blood cell distribution width (RDW). Our patient’s aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were both within normal limits.

Kidney function tests told another story. The patient’s serum creatinine (SCr) and blood urea nitrogen (BUN) levels were elevated (1.54 mg/dL and 29 mg/dL, respectively) and her creatinine clearance (CrCl; 30.2 mL/min) suggested moderate to severe renal dysfunction.

The patient’s CHADS₂ score was calculated as 1, suggesting she had a low-to-moderate risk of stroke.

THE DIAGNOSIS

Our patient had a left subconjunctival hemorrhage and an elevated venipuncture INR. Based on her renal dysfunction, we suspected that her elevated INR was likely due to an excessive dose of dabigatran, as well as an interaction between dabigatran and tacrolimus.

DISCUSSION

Dabigatran is an oral direct thrombin inhibitor approved for the prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation. An important advantage of dabigatran compared to warfarin is that the fixed-dose regimen does not require routine anticoagulation monitoring. In cases where anticoagulation monitoring is needed, PTT is the preferred method.¹

While PT and INR have generally not been shown to accurately reflect the degree of anticoagulation with dabigatran at therapeutic doses, there have been in vitro reports of elevated INRs with supratherapeutic dabigatran levels.^2,3 At a typical peak therapeutic dabigatran concentration of approximately 184 ng/mL, the INR generally ranged from 1.1 to 1.7.² However, at a dabigatran concentration of 1000 ng/mL, the INR was elevated to 4.5,^2,3 which is the same venipuncture INR recorded in our patient. While there have been published reports of falsely elevated point-of-care INR results compared to corresponding venipuncture INR results in patients taking dabigatran,^4,5 a literature review found only a case of an elevated venipuncture INR in an end-stage renal disease patient receiving hemodialysis.⁶

Concurrent use of any P-gp inhibitor (such as tacrolimus) and dabigatran is contraindicated in patients with severe renal dysfunction.

In the case noted above, as well as our patient, an accumulation of dabigatran due to the patient’s renal dysfunction likely resulted in high plasma concentrations and therefore an elevated venipuncture INR. The elimination half-life for dabigatran is approximately 14 hours in patients with normal renal function; in a patient with severe renal impairment, the half-life can be up to 28 hours.⁷ Our patient’s CrCl at the time of presentation was 30.2 mL/min, which indicated moderate to severe renal dysfunction. Based on dabigatran prescribing recommendations, a dose adjustment to 75 mg bid might be appropriate.¹ (Our patient was taking 150 mg bid.)

We do not believe our patient’s elevated INR was due to her liver transplant because there were no clinical signs of liver dysfunction. A more likely contributing factor was a drug interaction with tacrolimus. Dabigatran is a moderate affinity P-glycoprotein (P-gp) substrate and tacrolimus is both a P-gp substrate and inhibitor. While an interaction between tacrolimus and dabigatran has not been studied directly, concurrent use of any P-gp inhibitor and dabigatran is contraindicated in patients with severe renal dysfunction (CrCl: 15-30 mL/min).¹ For these theoretical interactions, the Drug Interaction Probability Scale (DIPS) has been developed.⁸ In our patient’s case, the calculated DIPS score of 5 suggests a probable interaction, likely due to P-gp inhibition. The other medications our patient was taking did not have this interaction and were unlikely to contribute to the elevated INR and subconjunctival hemorrhage.

Our patient was instructed to stop taking dabigatran and return in 3 days for additional lab tests. At her follow-up visit, the lab results were PTT, 34.3 seconds; PT, 11.6 seconds; and venipuncture INR, 1.1. Her CBC was unremarkable and unchanged. Shortly after the follow-up visit, our patient was assessed by her cardiologist. Due to her renal dysfunction, risk of bleeding, and relatively low CHADS₂ score, the cardiologist decided to discontinue dabigatran and start her on aspirin.

THE TAKEAWAY

Dabigatran may cause elevated INR levels in patients with renal dysfunction and/or those taking other medications that could interact with dabigatran. Concurrent use of any P-gp inhibitor (such as tacrolimus) and dabigatran is contraindicated in patients with severe renal dysfunction. Despite the lack of required routine laboratory monitoring, renal function and drug interactions associated with dabigatran therapy should be monitored closely.

Case

A 17-year-old adolescent girl with a history of depression and opioid dependence, for which she was taking buprenorphine until 2 weeks earlier, presented to the ED via emergency medical services (EMS) after her father found her lying on the couch unresponsive and with shallow respirations. Naloxone was administered by EMS and her mental status improved.

At presentation, the patient admitted to insufflation of an unknown amount of heroin and ingestion of 2 mg of alprazolam earlier in the day. She denied any past or current use of intravenous (IV) drugs. During monitoring, she began to complain of numbness in her legs and an inability to urinate. Examination revealed paralysis and decreased sensation of her bilateral lower extremities to the midthigh, with decreased rectal tone. Because of the patient’s history of drug use and temporal association with the heroin overdose, both neurosurgery and toxicology services were consulted.

What can cause lower extremity paralysis in a drug user?

The differential diagnosis for the patient at this point included toxin-induced myelopathy, Guillain-Barré syndrome, hypokalemic periodic paralysis, spinal compression, epidural abscess, cerebrovascular accident, spinal lesion, and spinal artery dissection or infarction.

Although Guillain-Barré syndrome presents with ascending paralysis, there is usually an antecedent respiratory or gastrointestinal infection. While epidural abscess with spinal compression is associated with IV drug use and can present similarly, the patient in this case denied IV use. In the absence of any risk factors, cerebrovascular accident and spinal artery dissection were also unlikely.

Case Continuation

A bladder catheter was placed due to the patient’s inability to urinate, and approximately 1 L of urine output was retrieved. Immediate magnetic resonance imaging (MRI) demonstrated increased T2 signal intensity and expansion of the distal thoracic cord and conus without mass lesion, consistent with transverse myelitis (TM).

What is transverse myelitis and why does it occur?

Transverse myelitis is an inflammatory demyelinating disorder that focally affects the spinal cord, resulting in a specific pattern of motor, sensory, and autonomic dysfunction.¹ Signs and symptoms include paresthesia, paralysis of the extremities, and loss of bladder and bowel control. The level of the spinal cord affected determines the clinical effects. Demyelination typically occurs at the thoracic segment, producing findings in the legs, as well as bladder and bowel dysfunction. 

The exact cause of TM is unknown, but the inflammation may result from a viral complication or an abnormal immune response. Infectious viral agents suspected of causing TM include varicella zoster, herpes simplex, cytomegalovirus, Epstein-Barr, influenza, human immunodeficiency virus, hepatitis A, and rubella. It has also been postulated that an autoimmune reaction is responsible for the condition.

In some individuals, TM represents the first manifestation of an underlying demyelinating disorder such as multiple sclerosis or neuromyelitis optica. A diagnosis of TM is made through patient history, physical examination, and characteristic findings on neuroimaging, specifically MRI.

Heroin use has long been associated with the development of TM, and is usually associated with IV administration of the drug after a period of abstinence.² This association strengthens the basis for an immunologic etiology—an initial sensitization and subsequent reexposure causing the effects of TM. There have also been cases of TM coexisting with rhabdomyolysis due to the patient being found in a contorted position.³ Another theory of the etiology of heroin-associated TM is a reaction to a possible adulterant or contaminant in the heroin.⁴

What is the treatment and prognosis of transverse myelitis?

Since there is no cure for TM, treatment is directed at reducing inflammation in the spinal cord. Initial therapy generally includes corticosteroids. In patients with a minimal response to corticosteroids, plasma exchange can be attempted. There are also limited data to suggest a beneficial role for the use of IV immunoglobulin.⁵ In addition to treatment, general supportive care must also be optimized, such as the use of prophylaxis for thrombophlebitis due to immobility and physical therapy, if possible.

The prognosis of patients with TM is variable, and up to two thirds of patients will have moderate-to-severe residual neurological disability.⁶ Recovery is slow, with most patients beginning to show improvement within the first 2 to 12 weeks from treatment and supportive care. The recovery process can continue for 2 years. However, if no improvement is made within the first 3 to 6 months, recovery is unlikely.⁷ Cases of heroin-associated TM may have a more favorable prognosis.⁸

A majority of individuals will only experience this clinical entity once, but there are rare causes of recurrent or relapsing TM.⁷ In these situations, a search for underlying demyelinating diseases should be performed.

Case Conclusion 

The patient was immediately started on IV corticosteroids, but as there was no improvement after 5 days, plasmapheresis was performed. She received 5 cycles of plasmapheresis and a 5-day course of IV immunoglobulin but still without any improvement. A repeat MRI of the thoracic spine was performed and raised the possibility of cord infarct, but infectious or inflammatory myelitis remained within differential consideration. The patient continued to make minimal improvement with physical therapy and, after a 3-week hospital course, she was transferred to inpatient rehabilitation for further care. Over the next 2 months, the loss of sensation and motor ability of her legs did not improve, but she did regain control of her bowels and bladder.

Dr Regina is a medical toxicology fellow in the department of emergency medicine at North Shore Long Island Jewish Health System, New York. Dr Nelson, editor of “Case Studies in Toxicology,” is a professor in the department of emergency medicine and director of the medical toxicology fellowship program at the New York University School of Medicine and the New York City Poison Control Center. He is also associate editor, toxicology, of the EMERGENCY MEDICINE editorial board.

Case

A 17-year-old adolescent girl with a history of depression and opioid dependence, for which she was taking buprenorphine until 2 weeks earlier, presented to the ED via emergency medical services (EMS) after her father found her lying on the couch unresponsive and with shallow respirations. Naloxone was administered by EMS and her mental status improved.

At presentation, the patient admitted to insufflation of an unknown amount of heroin and ingestion of 2 mg of alprazolam earlier in the day. She denied any past or current use of intravenous (IV) drugs. During monitoring, she began to complain of numbness in her legs and an inability to urinate. Examination revealed paralysis and decreased sensation of her bilateral lower extremities to the midthigh, with decreased rectal tone. Because of the patient’s history of drug use and temporal association with the heroin overdose, both neurosurgery and toxicology services were consulted.

What can cause lower extremity paralysis in a drug user?

The differential diagnosis for the patient at this point included toxin-induced myelopathy, Guillain-Barré syndrome, hypokalemic periodic paralysis, spinal compression, epidural abscess, cerebrovascular accident, spinal lesion, and spinal artery dissection or infarction.

Although Guillain-Barré syndrome presents with ascending paralysis, there is usually an antecedent respiratory or gastrointestinal infection. While epidural abscess with spinal compression is associated with IV drug use and can present similarly, the patient in this case denied IV use. In the absence of any risk factors, cerebrovascular accident and spinal artery dissection were also unlikely.

Case Continuation

A bladder catheter was placed due to the patient’s inability to urinate, and approximately 1 L of urine output was retrieved. Immediate magnetic resonance imaging (MRI) demonstrated increased T2 signal intensity and expansion of the distal thoracic cord and conus without mass lesion, consistent with transverse myelitis (TM).

What is transverse myelitis and why does it occur?

Transverse myelitis is an inflammatory demyelinating disorder that focally affects the spinal cord, resulting in a specific pattern of motor, sensory, and autonomic dysfunction.¹ Signs and symptoms include paresthesia, paralysis of the extremities, and loss of bladder and bowel control. The level of the spinal cord affected determines the clinical effects. Demyelination typically occurs at the thoracic segment, producing findings in the legs, as well as bladder and bowel dysfunction. 

The exact cause of TM is unknown, but the inflammation may result from a viral complication or an abnormal immune response. Infectious viral agents suspected of causing TM include varicella zoster, herpes simplex, cytomegalovirus, Epstein-Barr, influenza, human immunodeficiency virus, hepatitis A, and rubella. It has also been postulated that an autoimmune reaction is responsible for the condition.

In some individuals, TM represents the first manifestation of an underlying demyelinating disorder such as multiple sclerosis or neuromyelitis optica. A diagnosis of TM is made through patient history, physical examination, and characteristic findings on neuroimaging, specifically MRI.

Heroin use has long been associated with the development of TM, and is usually associated with IV administration of the drug after a period of abstinence.² This association strengthens the basis for an immunologic etiology—an initial sensitization and subsequent reexposure causing the effects of TM. There have also been cases of TM coexisting with rhabdomyolysis due to the patient being found in a contorted position.³ Another theory of the etiology of heroin-associated TM is a reaction to a possible adulterant or contaminant in the heroin.⁴

What is the treatment and prognosis of transverse myelitis?

Since there is no cure for TM, treatment is directed at reducing inflammation in the spinal cord. Initial therapy generally includes corticosteroids. In patients with a minimal response to corticosteroids, plasma exchange can be attempted. There are also limited data to suggest a beneficial role for the use of IV immunoglobulin.⁵ In addition to treatment, general supportive care must also be optimized, such as the use of prophylaxis for thrombophlebitis due to immobility and physical therapy, if possible.

The prognosis of patients with TM is variable, and up to two thirds of patients will have moderate-to-severe residual neurological disability.⁶ Recovery is slow, with most patients beginning to show improvement within the first 2 to 12 weeks from treatment and supportive care. The recovery process can continue for 2 years. However, if no improvement is made within the first 3 to 6 months, recovery is unlikely.⁷ Cases of heroin-associated TM may have a more favorable prognosis.⁸

A majority of individuals will only experience this clinical entity once, but there are rare causes of recurrent or relapsing TM.⁷ In these situations, a search for underlying demyelinating diseases should be performed.

Case Conclusion 

The patient was immediately started on IV corticosteroids, but as there was no improvement after 5 days, plasmapheresis was performed. She received 5 cycles of plasmapheresis and a 5-day course of IV immunoglobulin but still without any improvement. A repeat MRI of the thoracic spine was performed and raised the possibility of cord infarct, but infectious or inflammatory myelitis remained within differential consideration. The patient continued to make minimal improvement with physical therapy and, after a 3-week hospital course, she was transferred to inpatient rehabilitation for further care. Over the next 2 months, the loss of sensation and motor ability of her legs did not improve, but she did regain control of her bowels and bladder.

Dr Regina is a medical toxicology fellow in the department of emergency medicine at North Shore Long Island Jewish Health System, New York. Dr Nelson, editor of “Case Studies in Toxicology,” is a professor in the department of emergency medicine and director of the medical toxicology fellowship program at the New York University School of Medicine and the New York City Poison Control Center. He is also associate editor, toxicology, of the EMERGENCY MEDICINE editorial board.

Case

A 17-year-old adolescent girl with a history of depression and opioid dependence, for which she was taking buprenorphine until 2 weeks earlier, presented to the ED via emergency medical services (EMS) after her father found her lying on the couch unresponsive and with shallow respirations. Naloxone was administered by EMS and her mental status improved.

At presentation, the patient admitted to insufflation of an unknown amount of heroin and ingestion of 2 mg of alprazolam earlier in the day. She denied any past or current use of intravenous (IV) drugs. During monitoring, she began to complain of numbness in her legs and an inability to urinate. Examination revealed paralysis and decreased sensation of her bilateral lower extremities to the midthigh, with decreased rectal tone. Because of the patient’s history of drug use and temporal association with the heroin overdose, both neurosurgery and toxicology services were consulted.

What can cause lower extremity paralysis in a drug user?

The differential diagnosis for the patient at this point included toxin-induced myelopathy, Guillain-Barré syndrome, hypokalemic periodic paralysis, spinal compression, epidural abscess, cerebrovascular accident, spinal lesion, and spinal artery dissection or infarction.

Although Guillain-Barré syndrome presents with ascending paralysis, there is usually an antecedent respiratory or gastrointestinal infection. While epidural abscess with spinal compression is associated with IV drug use and can present similarly, the patient in this case denied IV use. In the absence of any risk factors, cerebrovascular accident and spinal artery dissection were also unlikely.

Case Continuation

A bladder catheter was placed due to the patient’s inability to urinate, and approximately 1 L of urine output was retrieved. Immediate magnetic resonance imaging (MRI) demonstrated increased T2 signal intensity and expansion of the distal thoracic cord and conus without mass lesion, consistent with transverse myelitis (TM).

What is transverse myelitis and why does it occur?

Transverse myelitis is an inflammatory demyelinating disorder that focally affects the spinal cord, resulting in a specific pattern of motor, sensory, and autonomic dysfunction.¹ Signs and symptoms include paresthesia, paralysis of the extremities, and loss of bladder and bowel control. The level of the spinal cord affected determines the clinical effects. Demyelination typically occurs at the thoracic segment, producing findings in the legs, as well as bladder and bowel dysfunction. 

The exact cause of TM is unknown, but the inflammation may result from a viral complication or an abnormal immune response. Infectious viral agents suspected of causing TM include varicella zoster, herpes simplex, cytomegalovirus, Epstein-Barr, influenza, human immunodeficiency virus, hepatitis A, and rubella. It has also been postulated that an autoimmune reaction is responsible for the condition.

In some individuals, TM represents the first manifestation of an underlying demyelinating disorder such as multiple sclerosis or neuromyelitis optica. A diagnosis of TM is made through patient history, physical examination, and characteristic findings on neuroimaging, specifically MRI.

Heroin use has long been associated with the development of TM, and is usually associated with IV administration of the drug after a period of abstinence.² This association strengthens the basis for an immunologic etiology—an initial sensitization and subsequent reexposure causing the effects of TM. There have also been cases of TM coexisting with rhabdomyolysis due to the patient being found in a contorted position.³ Another theory of the etiology of heroin-associated TM is a reaction to a possible adulterant or contaminant in the heroin.⁴

What is the treatment and prognosis of transverse myelitis?

Since there is no cure for TM, treatment is directed at reducing inflammation in the spinal cord. Initial therapy generally includes corticosteroids. In patients with a minimal response to corticosteroids, plasma exchange can be attempted. There are also limited data to suggest a beneficial role for the use of IV immunoglobulin.⁵ In addition to treatment, general supportive care must also be optimized, such as the use of prophylaxis for thrombophlebitis due to immobility and physical therapy, if possible.

The prognosis of patients with TM is variable, and up to two thirds of patients will have moderate-to-severe residual neurological disability.⁶ Recovery is slow, with most patients beginning to show improvement within the first 2 to 12 weeks from treatment and supportive care. The recovery process can continue for 2 years. However, if no improvement is made within the first 3 to 6 months, recovery is unlikely.⁷ Cases of heroin-associated TM may have a more favorable prognosis.⁸

A majority of individuals will only experience this clinical entity once, but there are rare causes of recurrent or relapsing TM.⁷ In these situations, a search for underlying demyelinating diseases should be performed.

Case Conclusion 

The patient was immediately started on IV corticosteroids, but as there was no improvement after 5 days, plasmapheresis was performed. She received 5 cycles of plasmapheresis and a 5-day course of IV immunoglobulin but still without any improvement. A repeat MRI of the thoracic spine was performed and raised the possibility of cord infarct, but infectious or inflammatory myelitis remained within differential consideration. The patient continued to make minimal improvement with physical therapy and, after a 3-week hospital course, she was transferred to inpatient rehabilitation for further care. Over the next 2 months, the loss of sensation and motor ability of her legs did not improve, but she did regain control of her bowels and bladder.

Dr Regina is a medical toxicology fellow in the department of emergency medicine at North Shore Long Island Jewish Health System, New York. Dr Nelson, editor of “Case Studies in Toxicology,” is a professor in the department of emergency medicine and director of the medical toxicology fellowship program at the New York University School of Medicine and the New York City Poison Control Center. He is also associate editor, toxicology, of the EMERGENCY MEDICINE editorial board.

Fractures of the clavicle, which account for 2.6% of all fractures, are displaced in 70% of cases and are mid-diaphyseal in 80% of cases.^1-3 Historically, both displaced and nondisplaced fractures were treated nonoperatively with excellent outcomes reported in the majority of patients.^1-3 Traditionally, the indications for surgical fixation of a clavicular fracture include open fractures, which occur infrequently, accounting for only 3.2% of clavicle fractures.⁴ Other indications include floating shoulder girdle or scapulothoracic dissociation, neurovascular injury, and skin “tenting” by the fracture fragments.^3,5 Recently, both meta-analyses and randomized clinical trials have reported reduced malunion rates and improved patient outcomes with open reduction and internal fixation (ORIF).^6-9 Consequently, operative fixation could be considered in patients with 100% displacement or greater than 1.5 cm shortening.^6-9 Open reduction and internal fixation of the clavicle has been demonstrated to have excellent outcomes in pediatric populations as well.¹⁰

The clavicle is subcutaneous for much of its length and, thus, displaced clavicular fractures often result in a visible deformity with a stretch of the soft-tissue envelope over the fracture. While this has been suggested as an operative indication, several recent sources indicate that this concern may only be theoretical. According to the fourth edition of Skeletal Trauma, “It is often stated that open reduction and internal fixation should be considered if the skin is threatened by pressure from a prominent clavicle fracture fragment; however, it is extremely rare of the skin to be perforated from within.”⁵ The most recent Journal of Bone and Joint Surgery Current Concepts Review on the subject stated that “open fractures or soft-tissue tenting sufficient to produce skin necrosis is uncommon.”³ To the best of our knowledge, there is no reported case of a displaced midshaft clavicle fracture with secondary skin necrosis and conversion into an open fracture, validating the conclusion that this complication may be only theoretical. Given that surgical fixation carries a risk of complications including wound complications, infection, nonunion, malunion, and damage to the nearby neurovascular structures and pleural apices,¹¹ some surgeons may be uncertain how to proceed in cases at risk for disturbance of the soft tissues.

We report 2 adolescent cases of displaced, comminuted clavicle fractures in which the skin was initially intact. Both were managed nonoperatively and both secondarily presented with open lesions at the fracture site requiring urgent irrigation and débridement (I&D) and ORIF. The patients and their guardians provided written informed consent for print and electronic publication of these case reports.

Case Reports

Case 1

A 15-year-old boy with no significant medical or surgical history flipped over the handlebars of his bicycle the day prior to presentation and sustained a clavicle fracture on his left nondominant upper extremity. This was an isolated injury. On examination, his skin was intact with an area of tender mild osseous protuberance at the midclavicle with associated surrounding edema. He was neurovascularly intact. Radiographs showed a displaced fracture of the midshaft of the clavicle with 20% shortening with a vertically angulated piece of comminution (Figure 1A). After a discussion of the treatment options with the family, the decision was made to pursue nonoperative treatment with sling immobilization as needed and restriction from gym and sports.

Two and a half weeks later, the patient presented at follow-up with significant reduction but persistence of his pain and a new complaint of drainage from the area of the fracture. On examination, he was found to have a puncture wound of the skin with exposed clavicle protruding through the wound with a 1-cm circumferential area of erythema without purulence present or expressible. The patient denied reinjury and endorsed compliance with sling immobilization. He was taken for urgent I&D and ORIF. After excision of the eschar surrounding the open lesion and full I&D of the soft tissues, the protruding spike was partially excised and the fracture site was débrided. The fracture was reduced and fixated with a lag screw and neutralization plate technique using an anatomically contoured locking clavicle plate (Synthes). Vancomycin powder was sprinkled into the wound at the completion of the procedure to reduce the chance of infection.¹²

Postoperatively, the patient was prescribed oral clindamycin but was subsequently switched to oral cephalexin because of mild signs of an allergic reaction, for a total course of antibiotics of 1 week. The patient was immobilized in a sling for comfort for the first 9 weeks postoperatively until radiographic union occurred. The patient’s wound healed uneventfully and with acceptable cosmesis. He was released to full activities at 10 weeks postoperatively. At final follow-up 6 months after surgery, the patient had returned to all of his regular activities without pain, and with full range of motion and no demonstrable deficits with radiographic union (Figure 1B).

Case 2

An 11-year-old boy with no significant medical or surgical history fell onto his right dominant upper extremity while doing a jump on his dirt bike 1 week prior to presentation, sustaining a clavicle fracture. This was an isolated injury. He was seen and evaluated by an outside orthopedist who noted that the soft-tissue envelope was intact and the patient was neurovascularly intact. Radiographs showed a displaced fracture of the midshaft of the clavicle with 15% shortening and with a vertically angulated piece of comminution (Figure 1C). Nonoperative treatment with a figure-of-8 brace was recommended. The patient’s discomfort completely resolved.

One week later, when he presented to the outside orthopedist for follow-up, the development of a wound overlying the fracture site was noted, and the patient was started on oral trimethoprim/sulfamethoxazole and referred to our office for treatment (Figure 1D). The patient denied reinjury and endorsed compliance with brace immobilization. On examination, the patient was afebrile and was noted to have a puncture wound at the fracture site with a protruding spike of bone and surrounding erythema but without present or expressible discharge (Figure 2). The patient was taken urgently for I&D and ORIF, using a similar technique to case 1, except that no lag screw was employed.

Postoperatively, the patient did well with no complications; he was prescribed oral cephalexin for 1 week. The patient was immobilized in a sling for the first 5 weeks after surgery until radiographic union had occurred, after which the sling was discontinued. The patient’s wound healed uneventfully and with acceptable cosmesis. The patient was released from activity restrictions at 6 weeks postoperatively. At final follow-up 5 weeks after surgery, the patient had full painless range of motion, no tenderness at the fracture site, no signs of infection on examination, and radiographic union (Figure 1D).

Discussion

Optimal treatment of displaced clavicle fractures is controversial. While nonoperative treatment has been recommended,^1-3 especially in skeletally immature populations with a capacity for remodeling,^7-9 2 recent randomized clinical trials have demonstrated improved patient outcomes with ORIF.^6,8,9 Traditionally, ORIF was recommended with tenting of the skin because of concern for an impending open fracture. However, recent review materials have implied that this complication may only be theoretical.^3,5 Indeed, in 2 randomized trials, sufficient displacement to cause concern for impending violation of the skin envelope was not listed as an exclusion criteria.^8,9 We report 2 cases of displaced comminuted clavicle fractures that were initially managed nonoperatively but developed open lesions at the fracture site. This complication, while rare, is possible, and surgeons must consider it as a possibility when assessing patients with displaced clavicle fractures. To the best of the authors’ knowledge, no guidelines exist to direct antibiotic choice and duration in secondarily open fractures.

These 2 cases have several features in common that may serve as risk factors for impending violation of the skin envelope. Both fractures had a vertically angulated segmental piece of comminution with a sharp spike. This feature has been identified as a potential risk factor for subsequent development of an open fracture in a case report of fragment excision without reduction or fixation to allow rapid return to play in a professional jockey.¹³ Both patients in these cases presented with high-velocity mechanisms of injury and significant displacement, both of which may serve as risk factors. In the only similar case the authors could identify, Strauss and colleagues¹⁴ described a distal clavicle fracture with significant displacement and with secondary ulceration of the skin complicated by infection presenting with purulent discharge, cultured positive for methicillin-sensitive Staphylococcus aureus, requiring management with an external fixator and 6 weeks of intravenous antibiotics. Because both cases presented here occurred in healthy adolescent patients who were taken urgently for I&D and ORIF as soon as the wound was discovered, deep infection was avoided in these cases. Finally, in 1 case, a figure-of-8 brace was employed, which may also have placed pressure on the skin overlying the fracture and may have predisposed this patient to this complication.

Conclusion

In displaced midshaft clavicle fractures, tenting of the skin sufficient to cause subsequent violation of the soft-tissue envelope is possible and is more than a theoretical risk. At-risk patients, ie, those with a vertically angulated sharp fragment of comminution, should be counseled appropriately and observed closely or considered for primary ORIF.

Fractures of the clavicle, which account for 2.6% of all fractures, are displaced in 70% of cases and are mid-diaphyseal in 80% of cases.^1-3 Historically, both displaced and nondisplaced fractures were treated nonoperatively with excellent outcomes reported in the majority of patients.^1-3 Traditionally, the indications for surgical fixation of a clavicular fracture include open fractures, which occur infrequently, accounting for only 3.2% of clavicle fractures.⁴ Other indications include floating shoulder girdle or scapulothoracic dissociation, neurovascular injury, and skin “tenting” by the fracture fragments.^3,5 Recently, both meta-analyses and randomized clinical trials have reported reduced malunion rates and improved patient outcomes with open reduction and internal fixation (ORIF).^6-9 Consequently, operative fixation could be considered in patients with 100% displacement or greater than 1.5 cm shortening.^6-9 Open reduction and internal fixation of the clavicle has been demonstrated to have excellent outcomes in pediatric populations as well.¹⁰

The clavicle is subcutaneous for much of its length and, thus, displaced clavicular fractures often result in a visible deformity with a stretch of the soft-tissue envelope over the fracture. While this has been suggested as an operative indication, several recent sources indicate that this concern may only be theoretical. According to the fourth edition of Skeletal Trauma, “It is often stated that open reduction and internal fixation should be considered if the skin is threatened by pressure from a prominent clavicle fracture fragment; however, it is extremely rare of the skin to be perforated from within.”⁵ The most recent Journal of Bone and Joint Surgery Current Concepts Review on the subject stated that “open fractures or soft-tissue tenting sufficient to produce skin necrosis is uncommon.”³ To the best of our knowledge, there is no reported case of a displaced midshaft clavicle fracture with secondary skin necrosis and conversion into an open fracture, validating the conclusion that this complication may be only theoretical. Given that surgical fixation carries a risk of complications including wound complications, infection, nonunion, malunion, and damage to the nearby neurovascular structures and pleural apices,¹¹ some surgeons may be uncertain how to proceed in cases at risk for disturbance of the soft tissues.

We report 2 adolescent cases of displaced, comminuted clavicle fractures in which the skin was initially intact. Both were managed nonoperatively and both secondarily presented with open lesions at the fracture site requiring urgent irrigation and débridement (I&D) and ORIF. The patients and their guardians provided written informed consent for print and electronic publication of these case reports.

Case Reports

Case 1

A 15-year-old boy with no significant medical or surgical history flipped over the handlebars of his bicycle the day prior to presentation and sustained a clavicle fracture on his left nondominant upper extremity. This was an isolated injury. On examination, his skin was intact with an area of tender mild osseous protuberance at the midclavicle with associated surrounding edema. He was neurovascularly intact. Radiographs showed a displaced fracture of the midshaft of the clavicle with 20% shortening with a vertically angulated piece of comminution (Figure 1A). After a discussion of the treatment options with the family, the decision was made to pursue nonoperative treatment with sling immobilization as needed and restriction from gym and sports.

Two and a half weeks later, the patient presented at follow-up with significant reduction but persistence of his pain and a new complaint of drainage from the area of the fracture. On examination, he was found to have a puncture wound of the skin with exposed clavicle protruding through the wound with a 1-cm circumferential area of erythema without purulence present or expressible. The patient denied reinjury and endorsed compliance with sling immobilization. He was taken for urgent I&D and ORIF. After excision of the eschar surrounding the open lesion and full I&D of the soft tissues, the protruding spike was partially excised and the fracture site was débrided. The fracture was reduced and fixated with a lag screw and neutralization plate technique using an anatomically contoured locking clavicle plate (Synthes). Vancomycin powder was sprinkled into the wound at the completion of the procedure to reduce the chance of infection.¹²

Postoperatively, the patient was prescribed oral clindamycin but was subsequently switched to oral cephalexin because of mild signs of an allergic reaction, for a total course of antibiotics of 1 week. The patient was immobilized in a sling for comfort for the first 9 weeks postoperatively until radiographic union occurred. The patient’s wound healed uneventfully and with acceptable cosmesis. He was released to full activities at 10 weeks postoperatively. At final follow-up 6 months after surgery, the patient had returned to all of his regular activities without pain, and with full range of motion and no demonstrable deficits with radiographic union (Figure 1B).

Case 2

An 11-year-old boy with no significant medical or surgical history fell onto his right dominant upper extremity while doing a jump on his dirt bike 1 week prior to presentation, sustaining a clavicle fracture. This was an isolated injury. He was seen and evaluated by an outside orthopedist who noted that the soft-tissue envelope was intact and the patient was neurovascularly intact. Radiographs showed a displaced fracture of the midshaft of the clavicle with 15% shortening and with a vertically angulated piece of comminution (Figure 1C). Nonoperative treatment with a figure-of-8 brace was recommended. The patient’s discomfort completely resolved.

One week later, when he presented to the outside orthopedist for follow-up, the development of a wound overlying the fracture site was noted, and the patient was started on oral trimethoprim/sulfamethoxazole and referred to our office for treatment (Figure 1D). The patient denied reinjury and endorsed compliance with brace immobilization. On examination, the patient was afebrile and was noted to have a puncture wound at the fracture site with a protruding spike of bone and surrounding erythema but without present or expressible discharge (Figure 2). The patient was taken urgently for I&D and ORIF, using a similar technique to case 1, except that no lag screw was employed.

Postoperatively, the patient did well with no complications; he was prescribed oral cephalexin for 1 week. The patient was immobilized in a sling for the first 5 weeks after surgery until radiographic union had occurred, after which the sling was discontinued. The patient’s wound healed uneventfully and with acceptable cosmesis. The patient was released from activity restrictions at 6 weeks postoperatively. At final follow-up 5 weeks after surgery, the patient had full painless range of motion, no tenderness at the fracture site, no signs of infection on examination, and radiographic union (Figure 1D).

Discussion

Optimal treatment of displaced clavicle fractures is controversial. While nonoperative treatment has been recommended,^1-3 especially in skeletally immature populations with a capacity for remodeling,^7-9 2 recent randomized clinical trials have demonstrated improved patient outcomes with ORIF.^6,8,9 Traditionally, ORIF was recommended with tenting of the skin because of concern for an impending open fracture. However, recent review materials have implied that this complication may only be theoretical.^3,5 Indeed, in 2 randomized trials, sufficient displacement to cause concern for impending violation of the skin envelope was not listed as an exclusion criteria.^8,9 We report 2 cases of displaced comminuted clavicle fractures that were initially managed nonoperatively but developed open lesions at the fracture site. This complication, while rare, is possible, and surgeons must consider it as a possibility when assessing patients with displaced clavicle fractures. To the best of the authors’ knowledge, no guidelines exist to direct antibiotic choice and duration in secondarily open fractures.

These 2 cases have several features in common that may serve as risk factors for impending violation of the skin envelope. Both fractures had a vertically angulated segmental piece of comminution with a sharp spike. This feature has been identified as a potential risk factor for subsequent development of an open fracture in a case report of fragment excision without reduction or fixation to allow rapid return to play in a professional jockey.¹³ Both patients in these cases presented with high-velocity mechanisms of injury and significant displacement, both of which may serve as risk factors. In the only similar case the authors could identify, Strauss and colleagues¹⁴ described a distal clavicle fracture with significant displacement and with secondary ulceration of the skin complicated by infection presenting with purulent discharge, cultured positive for methicillin-sensitive Staphylococcus aureus, requiring management with an external fixator and 6 weeks of intravenous antibiotics. Because both cases presented here occurred in healthy adolescent patients who were taken urgently for I&D and ORIF as soon as the wound was discovered, deep infection was avoided in these cases. Finally, in 1 case, a figure-of-8 brace was employed, which may also have placed pressure on the skin overlying the fracture and may have predisposed this patient to this complication.

Conclusion

In displaced midshaft clavicle fractures, tenting of the skin sufficient to cause subsequent violation of the soft-tissue envelope is possible and is more than a theoretical risk. At-risk patients, ie, those with a vertically angulated sharp fragment of comminution, should be counseled appropriately and observed closely or considered for primary ORIF.

Fractures of the clavicle, which account for 2.6% of all fractures, are displaced in 70% of cases and are mid-diaphyseal in 80% of cases.^1-3 Historically, both displaced and nondisplaced fractures were treated nonoperatively with excellent outcomes reported in the majority of patients.^1-3 Traditionally, the indications for surgical fixation of a clavicular fracture include open fractures, which occur infrequently, accounting for only 3.2% of clavicle fractures.⁴ Other indications include floating shoulder girdle or scapulothoracic dissociation, neurovascular injury, and skin “tenting” by the fracture fragments.^3,5 Recently, both meta-analyses and randomized clinical trials have reported reduced malunion rates and improved patient outcomes with open reduction and internal fixation (ORIF).^6-9 Consequently, operative fixation could be considered in patients with 100% displacement or greater than 1.5 cm shortening.^6-9 Open reduction and internal fixation of the clavicle has been demonstrated to have excellent outcomes in pediatric populations as well.¹⁰

The clavicle is subcutaneous for much of its length and, thus, displaced clavicular fractures often result in a visible deformity with a stretch of the soft-tissue envelope over the fracture. While this has been suggested as an operative indication, several recent sources indicate that this concern may only be theoretical. According to the fourth edition of Skeletal Trauma, “It is often stated that open reduction and internal fixation should be considered if the skin is threatened by pressure from a prominent clavicle fracture fragment; however, it is extremely rare of the skin to be perforated from within.”⁵ The most recent Journal of Bone and Joint Surgery Current Concepts Review on the subject stated that “open fractures or soft-tissue tenting sufficient to produce skin necrosis is uncommon.”³ To the best of our knowledge, there is no reported case of a displaced midshaft clavicle fracture with secondary skin necrosis and conversion into an open fracture, validating the conclusion that this complication may be only theoretical. Given that surgical fixation carries a risk of complications including wound complications, infection, nonunion, malunion, and damage to the nearby neurovascular structures and pleural apices,¹¹ some surgeons may be uncertain how to proceed in cases at risk for disturbance of the soft tissues.

We report 2 adolescent cases of displaced, comminuted clavicle fractures in which the skin was initially intact. Both were managed nonoperatively and both secondarily presented with open lesions at the fracture site requiring urgent irrigation and débridement (I&D) and ORIF. The patients and their guardians provided written informed consent for print and electronic publication of these case reports.

Case Reports

Case 1

A 15-year-old boy with no significant medical or surgical history flipped over the handlebars of his bicycle the day prior to presentation and sustained a clavicle fracture on his left nondominant upper extremity. This was an isolated injury. On examination, his skin was intact with an area of tender mild osseous protuberance at the midclavicle with associated surrounding edema. He was neurovascularly intact. Radiographs showed a displaced fracture of the midshaft of the clavicle with 20% shortening with a vertically angulated piece of comminution (Figure 1A). After a discussion of the treatment options with the family, the decision was made to pursue nonoperative treatment with sling immobilization as needed and restriction from gym and sports.

Two and a half weeks later, the patient presented at follow-up with significant reduction but persistence of his pain and a new complaint of drainage from the area of the fracture. On examination, he was found to have a puncture wound of the skin with exposed clavicle protruding through the wound with a 1-cm circumferential area of erythema without purulence present or expressible. The patient denied reinjury and endorsed compliance with sling immobilization. He was taken for urgent I&D and ORIF. After excision of the eschar surrounding the open lesion and full I&D of the soft tissues, the protruding spike was partially excised and the fracture site was débrided. The fracture was reduced and fixated with a lag screw and neutralization plate technique using an anatomically contoured locking clavicle plate (Synthes). Vancomycin powder was sprinkled into the wound at the completion of the procedure to reduce the chance of infection.¹²

Postoperatively, the patient was prescribed oral clindamycin but was subsequently switched to oral cephalexin because of mild signs of an allergic reaction, for a total course of antibiotics of 1 week. The patient was immobilized in a sling for comfort for the first 9 weeks postoperatively until radiographic union occurred. The patient’s wound healed uneventfully and with acceptable cosmesis. He was released to full activities at 10 weeks postoperatively. At final follow-up 6 months after surgery, the patient had returned to all of his regular activities without pain, and with full range of motion and no demonstrable deficits with radiographic union (Figure 1B).

Case 2

An 11-year-old boy with no significant medical or surgical history fell onto his right dominant upper extremity while doing a jump on his dirt bike 1 week prior to presentation, sustaining a clavicle fracture. This was an isolated injury. He was seen and evaluated by an outside orthopedist who noted that the soft-tissue envelope was intact and the patient was neurovascularly intact. Radiographs showed a displaced fracture of the midshaft of the clavicle with 15% shortening and with a vertically angulated piece of comminution (Figure 1C). Nonoperative treatment with a figure-of-8 brace was recommended. The patient’s discomfort completely resolved.

One week later, when he presented to the outside orthopedist for follow-up, the development of a wound overlying the fracture site was noted, and the patient was started on oral trimethoprim/sulfamethoxazole and referred to our office for treatment (Figure 1D). The patient denied reinjury and endorsed compliance with brace immobilization. On examination, the patient was afebrile and was noted to have a puncture wound at the fracture site with a protruding spike of bone and surrounding erythema but without present or expressible discharge (Figure 2). The patient was taken urgently for I&D and ORIF, using a similar technique to case 1, except that no lag screw was employed.

Postoperatively, the patient did well with no complications; he was prescribed oral cephalexin for 1 week. The patient was immobilized in a sling for the first 5 weeks after surgery until radiographic union had occurred, after which the sling was discontinued. The patient’s wound healed uneventfully and with acceptable cosmesis. The patient was released from activity restrictions at 6 weeks postoperatively. At final follow-up 5 weeks after surgery, the patient had full painless range of motion, no tenderness at the fracture site, no signs of infection on examination, and radiographic union (Figure 1D).

Discussion

Optimal treatment of displaced clavicle fractures is controversial. While nonoperative treatment has been recommended,^1-3 especially in skeletally immature populations with a capacity for remodeling,^7-9 2 recent randomized clinical trials have demonstrated improved patient outcomes with ORIF.^6,8,9 Traditionally, ORIF was recommended with tenting of the skin because of concern for an impending open fracture. However, recent review materials have implied that this complication may only be theoretical.^3,5 Indeed, in 2 randomized trials, sufficient displacement to cause concern for impending violation of the skin envelope was not listed as an exclusion criteria.^8,9 We report 2 cases of displaced comminuted clavicle fractures that were initially managed nonoperatively but developed open lesions at the fracture site. This complication, while rare, is possible, and surgeons must consider it as a possibility when assessing patients with displaced clavicle fractures. To the best of the authors’ knowledge, no guidelines exist to direct antibiotic choice and duration in secondarily open fractures.

These 2 cases have several features in common that may serve as risk factors for impending violation of the skin envelope. Both fractures had a vertically angulated segmental piece of comminution with a sharp spike. This feature has been identified as a potential risk factor for subsequent development of an open fracture in a case report of fragment excision without reduction or fixation to allow rapid return to play in a professional jockey.¹³ Both patients in these cases presented with high-velocity mechanisms of injury and significant displacement, both of which may serve as risk factors. In the only similar case the authors could identify, Strauss and colleagues¹⁴ described a distal clavicle fracture with significant displacement and with secondary ulceration of the skin complicated by infection presenting with purulent discharge, cultured positive for methicillin-sensitive Staphylococcus aureus, requiring management with an external fixator and 6 weeks of intravenous antibiotics. Because both cases presented here occurred in healthy adolescent patients who were taken urgently for I&D and ORIF as soon as the wound was discovered, deep infection was avoided in these cases. Finally, in 1 case, a figure-of-8 brace was employed, which may also have placed pressure on the skin overlying the fracture and may have predisposed this patient to this complication.

Conclusion

In displaced midshaft clavicle fractures, tenting of the skin sufficient to cause subsequent violation of the soft-tissue envelope is possible and is more than a theoretical risk. At-risk patients, ie, those with a vertically angulated sharp fragment of comminution, should be counseled appropriately and observed closely or considered for primary ORIF.

Adamantinomas are rare primary malignant bone tumors (less than 1% of all bone tumors) that arise most commonly in the tibia.¹ There is a predilection for adult men aged 20 to 50 years, with rare occurrences in children. These tumors are malignant, highly invasive, and have significant metastatic potential.² A rarely seen, benign variant, known as osteofibrous dysplasia–like adamantinoma, is described in the literature, with fewer than 135 cases reported.^3-5 This variant predominantly has benign characteristics of an osteofibrous dysplasia lesion but has the potential to transform into an adamantinoma.⁶ Osteofibrous dysplasia–like adamantinoma has been observed to regress with age and is also referred to as a regressing adamantinoma or differentiated adamantinoma.⁷ 

We report an uncommon case of an osteofibrous dysplasia–like adamantinoma of the tibia in a 15-year-old girl. We decided to observe the tumor with regular 3- to 6-month follow-ups. Osteofibrous dysplasia–like adamantinoma in our patient has remained stable for 2 years and has an excellent prognosis.⁸ We report this case for its rarity, its short-term stability, and significant treatment implications due to its potential to regress or develop into a malignant form. The patient and the patient’s guardian provided written informed consent for print and electronic publication of this case report. 

Case Report

A healthy 15-year-old girl was referred to our institution for evaluation of anterior left knee pain. She had sustained a fall while playing basketball 3 months earlier and had been having left knee pain since that time. She did not have any swelling, catching, or locking in her left knee. She denied any recent fever, chills, night sweats, weight loss, nausea, vomiting, or diarrhea. On physical examination, her gait was normal and no swelling, erythema, or tenderness was noticed around the left knee.

Plain radiographs revealed a heterogeneous lesion with sclerosis and thickening of the anteromedial cortex of the proximal left tibia (Figures 1A, 1B). A computed tomography (CT) scan of the abdomen, pelvis, and chest showed no osseous abnormalities. A whole-body bone scan showed activity in the anterior aspect of the left proximal tibia. No other areas of activity were noted. Magnetic resonance imaging of the left leg showed an elongated, multiloculated, enhancing mass arising from the anterolateral cortex and extending from the tibial tuberosity to the mid-diaphysis of the left tibia. Histologic examination of the CT-guided core needle biopsy specimen showed that the lesion was composed of dense fibrocollagenous tissue separating irregular bony trabeculae with osteoblastic and osteoclastic activity. There was no evidence of any atypical cells, necrosis, or significant mitotic activity. No epithelial cells were identified on hematoxylin-eosin (H&E) stain (Figure 2). However, immunohistochemical staining was positive for focal cytokeratin-positive epithelial cells (Figure 3). The lesion was diagnosed as an osteofibrous dysplasia–like adamantinoma on the basis of the radiographic and histologic findings. We elected nonoperative intervention given the benign nature of the lesion and its potential to regress. Given the possibility of sampling error and potential for progression, the patient was followed regularly at 3- to 6-month intervals over a 2-year period without disease progression. 

Discussion

Osteofibrous dysplasia, osteofibrous dysplasia–like adamantinoma, and adamantinoma are rare fibro-osseous lesions that largely involve the midshaft of the tibia. Osteofibrous dysplasia accounts for 0.2% of primary bone tumors, whereas adamantinoma accounts for 0.1% to 0.5% of malignant bone tumors.⁹ Osteofibrous dysplasia is a benign lesion composed primarily of fibro-osseous tissue. Adamantinoma, however, is a slow-growing, low-grade, malignant biphasic tumor with intermingled epithelial and fibro-osseous components. It is an aggressive tumor that is locally invasive and can metastasize.² Osteofibrous dysplasia–like adamantinoma (also known as differentiated or regressing adamantinoma) is a benign lesion like osteofibrous dysplasia but has features of both osteofibrous dysplasia and adamantinoma. Osteofibrous dysplasia–like adamantinoma may progress and become a malignant adamantinoma.^6,10

The radiologic features of the 3 lesions are quite similar. It is not possible to distinguish between osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma based on imaging alone.⁹ Adamantinoma, being highly invasive, can be distinguished from osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma according to the extent of involvement of the medullary cavity seen on magnetic resonance imaging.⁹ Complete involvement of the medullary cavity is almost always seen in an adamantinoma. Involvement of the medullary cavity is minimal or absent in osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma lesions. 

Tissue confirmation through biopsy is crucial for accurate diagnosis. A biopsy should always be performed on any suspicious lesion,^3,6 and the fibro-osseous lesion should be treated as an adamantinoma if findings are equivocal. A biopsy also distinctly distinguishes these lesions from benign fibrous cortical defects, which have a similar radiographic appearance. While open biopsy is the gold standard, minimally invasive techniques such as core needle biopsy and fine needle biopsy are increasingly used.⁶ Because of the higher risk of misdiagnosis with minimally invasive techniques, radiographic confirmation is highly recommended.⁵

Histologically, both osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma do not stain for cytokeratin on H&E stain. However, they can be differentiated based on immunohistochemical staining for cytokeratin. Osteofibrous dysplasia lesions exhibit diffuse staining whereas osteofibrous dysplasia–like adamantinoma lesions show focal staining of small nests of epithelial cells. Adamantinoma, in comparison, exhibits a biphasic pattern on H&E stain, representing areas of epithelial and osteofibrous cells. Immunohistochemical staining for cytokeratin of an adamantinoma reveals large nests of epithelial cells.  

The association between osteofibrous dysplasia, osteofibrous dysplasia–like adamantinoma, and adamantinoma is not clearly established. However, it is widely believed that these 3 lesions represent a spectrum of the same disease and are linearly related in disease progression, with osteofibrous dysplasia at the benign end of the spectrum, osteofibrous dysplasia–like adamantinoma the intermediate form, and adamantinoma at the malignant end of the spectrum.¹¹

Hazelbag and colleagues⁶ and Springfield and colleagues¹⁰ point out that osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma could be precursor lesions of adamantinoma. We found several studies in the literature that support and document progression from osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma to an adamantinoma.^4,6,10,12 Other studies, however, showed no progression from either a benign osteofibrous dysplasia or an osteofibrous dysplasia–like adamantinoma lesion to a malignant adamantinoma. Park and colleagues¹³ described 41 cases of osteofibrous dysplasia that did not progress to adamantinoma. Kuruvilla and Steiner⁸ described 5 cases of osteofibrous dysplasia–like adamantinoma that showed no progression to adamantinoma. Additionally, our case has not progressed and has remained radiographically stable over a 2-year follow-up. Czerniak and colleagues⁷ and Ueda and colleagues¹⁴ postulated, based on histologic and immunohistochemical studies, that osteofibrous dysplasia–like adamantinoma might be a regressing form of an adamantinoma that is undergoing reparative processes that could result in complete elimination of all tumor cells.

In general, any lesion with absent to low malignant potential could be managed nonoperatively with periodic observation and without the need for surgical intervention. Thus, identification of a stable or nonprogressing osteofibrous dysplasia–like adamantinoma lesion has significant treatment implications. Campanacci and Laus¹⁵ at the Rizzoli Institute in Milan, through long term follow-up of their patients with osteofibrous dysplasia, found that most lesions had a tendency to regress spontaneously by puberty. They recommended that nonextensive osteofibrous dysplasia lesions should be observed, and surgery should be delayed until puberty. Gleason and colleagues¹⁶ also recommended nonoperative management of osteofibrous dysplasia lesions, with surgery used only for large, deforming, and highly invasive lesions. We recommend a similar treatment approach for osteofibrous dysplasia–like adamantinoma lesions.

Adamantinomas, however, are usually symptomatic, are highly invasive, have a high recurrence rate, and can metastasize.9 In these patients, a wide en bloc resection or amputation should be performed as soon as possible.11 Our case highlights that osteofibrous dysplasia–like adamantinoma lesions can occur in children and can remain stable, especially over the short term. Such lesions can be observed without surgical intervention.

Adamantinomas are rare primary malignant bone tumors (less than 1% of all bone tumors) that arise most commonly in the tibia.¹ There is a predilection for adult men aged 20 to 50 years, with rare occurrences in children. These tumors are malignant, highly invasive, and have significant metastatic potential.² A rarely seen, benign variant, known as osteofibrous dysplasia–like adamantinoma, is described in the literature, with fewer than 135 cases reported.^3-5 This variant predominantly has benign characteristics of an osteofibrous dysplasia lesion but has the potential to transform into an adamantinoma.⁶ Osteofibrous dysplasia–like adamantinoma has been observed to regress with age and is also referred to as a regressing adamantinoma or differentiated adamantinoma.⁷ 

We report an uncommon case of an osteofibrous dysplasia–like adamantinoma of the tibia in a 15-year-old girl. We decided to observe the tumor with regular 3- to 6-month follow-ups. Osteofibrous dysplasia–like adamantinoma in our patient has remained stable for 2 years and has an excellent prognosis.⁸ We report this case for its rarity, its short-term stability, and significant treatment implications due to its potential to regress or develop into a malignant form. The patient and the patient’s guardian provided written informed consent for print and electronic publication of this case report. 

Case Report

A healthy 15-year-old girl was referred to our institution for evaluation of anterior left knee pain. She had sustained a fall while playing basketball 3 months earlier and had been having left knee pain since that time. She did not have any swelling, catching, or locking in her left knee. She denied any recent fever, chills, night sweats, weight loss, nausea, vomiting, or diarrhea. On physical examination, her gait was normal and no swelling, erythema, or tenderness was noticed around the left knee.

Plain radiographs revealed a heterogeneous lesion with sclerosis and thickening of the anteromedial cortex of the proximal left tibia (Figures 1A, 1B). A computed tomography (CT) scan of the abdomen, pelvis, and chest showed no osseous abnormalities. A whole-body bone scan showed activity in the anterior aspect of the left proximal tibia. No other areas of activity were noted. Magnetic resonance imaging of the left leg showed an elongated, multiloculated, enhancing mass arising from the anterolateral cortex and extending from the tibial tuberosity to the mid-diaphysis of the left tibia. Histologic examination of the CT-guided core needle biopsy specimen showed that the lesion was composed of dense fibrocollagenous tissue separating irregular bony trabeculae with osteoblastic and osteoclastic activity. There was no evidence of any atypical cells, necrosis, or significant mitotic activity. No epithelial cells were identified on hematoxylin-eosin (H&E) stain (Figure 2). However, immunohistochemical staining was positive for focal cytokeratin-positive epithelial cells (Figure 3). The lesion was diagnosed as an osteofibrous dysplasia–like adamantinoma on the basis of the radiographic and histologic findings. We elected nonoperative intervention given the benign nature of the lesion and its potential to regress. Given the possibility of sampling error and potential for progression, the patient was followed regularly at 3- to 6-month intervals over a 2-year period without disease progression. 

Discussion

Osteofibrous dysplasia, osteofibrous dysplasia–like adamantinoma, and adamantinoma are rare fibro-osseous lesions that largely involve the midshaft of the tibia. Osteofibrous dysplasia accounts for 0.2% of primary bone tumors, whereas adamantinoma accounts for 0.1% to 0.5% of malignant bone tumors.⁹ Osteofibrous dysplasia is a benign lesion composed primarily of fibro-osseous tissue. Adamantinoma, however, is a slow-growing, low-grade, malignant biphasic tumor with intermingled epithelial and fibro-osseous components. It is an aggressive tumor that is locally invasive and can metastasize.² Osteofibrous dysplasia–like adamantinoma (also known as differentiated or regressing adamantinoma) is a benign lesion like osteofibrous dysplasia but has features of both osteofibrous dysplasia and adamantinoma. Osteofibrous dysplasia–like adamantinoma may progress and become a malignant adamantinoma.^6,10

The radiologic features of the 3 lesions are quite similar. It is not possible to distinguish between osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma based on imaging alone.⁹ Adamantinoma, being highly invasive, can be distinguished from osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma according to the extent of involvement of the medullary cavity seen on magnetic resonance imaging.⁹ Complete involvement of the medullary cavity is almost always seen in an adamantinoma. Involvement of the medullary cavity is minimal or absent in osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma lesions. 

Tissue confirmation through biopsy is crucial for accurate diagnosis. A biopsy should always be performed on any suspicious lesion,^3,6 and the fibro-osseous lesion should be treated as an adamantinoma if findings are equivocal. A biopsy also distinctly distinguishes these lesions from benign fibrous cortical defects, which have a similar radiographic appearance. While open biopsy is the gold standard, minimally invasive techniques such as core needle biopsy and fine needle biopsy are increasingly used.⁶ Because of the higher risk of misdiagnosis with minimally invasive techniques, radiographic confirmation is highly recommended.⁵

Histologically, both osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma do not stain for cytokeratin on H&E stain. However, they can be differentiated based on immunohistochemical staining for cytokeratin. Osteofibrous dysplasia lesions exhibit diffuse staining whereas osteofibrous dysplasia–like adamantinoma lesions show focal staining of small nests of epithelial cells. Adamantinoma, in comparison, exhibits a biphasic pattern on H&E stain, representing areas of epithelial and osteofibrous cells. Immunohistochemical staining for cytokeratin of an adamantinoma reveals large nests of epithelial cells.  

The association between osteofibrous dysplasia, osteofibrous dysplasia–like adamantinoma, and adamantinoma is not clearly established. However, it is widely believed that these 3 lesions represent a spectrum of the same disease and are linearly related in disease progression, with osteofibrous dysplasia at the benign end of the spectrum, osteofibrous dysplasia–like adamantinoma the intermediate form, and adamantinoma at the malignant end of the spectrum.¹¹

Hazelbag and colleagues⁶ and Springfield and colleagues¹⁰ point out that osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma could be precursor lesions of adamantinoma. We found several studies in the literature that support and document progression from osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma to an adamantinoma.^4,6,10,12 Other studies, however, showed no progression from either a benign osteofibrous dysplasia or an osteofibrous dysplasia–like adamantinoma lesion to a malignant adamantinoma. Park and colleagues¹³ described 41 cases of osteofibrous dysplasia that did not progress to adamantinoma. Kuruvilla and Steiner⁸ described 5 cases of osteofibrous dysplasia–like adamantinoma that showed no progression to adamantinoma. Additionally, our case has not progressed and has remained radiographically stable over a 2-year follow-up. Czerniak and colleagues⁷ and Ueda and colleagues¹⁴ postulated, based on histologic and immunohistochemical studies, that osteofibrous dysplasia–like adamantinoma might be a regressing form of an adamantinoma that is undergoing reparative processes that could result in complete elimination of all tumor cells.

In general, any lesion with absent to low malignant potential could be managed nonoperatively with periodic observation and without the need for surgical intervention. Thus, identification of a stable or nonprogressing osteofibrous dysplasia–like adamantinoma lesion has significant treatment implications. Campanacci and Laus¹⁵ at the Rizzoli Institute in Milan, through long term follow-up of their patients with osteofibrous dysplasia, found that most lesions had a tendency to regress spontaneously by puberty. They recommended that nonextensive osteofibrous dysplasia lesions should be observed, and surgery should be delayed until puberty. Gleason and colleagues¹⁶ also recommended nonoperative management of osteofibrous dysplasia lesions, with surgery used only for large, deforming, and highly invasive lesions. We recommend a similar treatment approach for osteofibrous dysplasia–like adamantinoma lesions.

Adamantinomas, however, are usually symptomatic, are highly invasive, have a high recurrence rate, and can metastasize.9 In these patients, a wide en bloc resection or amputation should be performed as soon as possible.11 Our case highlights that osteofibrous dysplasia–like adamantinoma lesions can occur in children and can remain stable, especially over the short term. Such lesions can be observed without surgical intervention.

Adamantinomas are rare primary malignant bone tumors (less than 1% of all bone tumors) that arise most commonly in the tibia.¹ There is a predilection for adult men aged 20 to 50 years, with rare occurrences in children. These tumors are malignant, highly invasive, and have significant metastatic potential.² A rarely seen, benign variant, known as osteofibrous dysplasia–like adamantinoma, is described in the literature, with fewer than 135 cases reported.^3-5 This variant predominantly has benign characteristics of an osteofibrous dysplasia lesion but has the potential to transform into an adamantinoma.⁶ Osteofibrous dysplasia–like adamantinoma has been observed to regress with age and is also referred to as a regressing adamantinoma or differentiated adamantinoma.⁷ 

We report an uncommon case of an osteofibrous dysplasia–like adamantinoma of the tibia in a 15-year-old girl. We decided to observe the tumor with regular 3- to 6-month follow-ups. Osteofibrous dysplasia–like adamantinoma in our patient has remained stable for 2 years and has an excellent prognosis.⁸ We report this case for its rarity, its short-term stability, and significant treatment implications due to its potential to regress or develop into a malignant form. The patient and the patient’s guardian provided written informed consent for print and electronic publication of this case report. 

Case Report

A healthy 15-year-old girl was referred to our institution for evaluation of anterior left knee pain. She had sustained a fall while playing basketball 3 months earlier and had been having left knee pain since that time. She did not have any swelling, catching, or locking in her left knee. She denied any recent fever, chills, night sweats, weight loss, nausea, vomiting, or diarrhea. On physical examination, her gait was normal and no swelling, erythema, or tenderness was noticed around the left knee.

Plain radiographs revealed a heterogeneous lesion with sclerosis and thickening of the anteromedial cortex of the proximal left tibia (Figures 1A, 1B). A computed tomography (CT) scan of the abdomen, pelvis, and chest showed no osseous abnormalities. A whole-body bone scan showed activity in the anterior aspect of the left proximal tibia. No other areas of activity were noted. Magnetic resonance imaging of the left leg showed an elongated, multiloculated, enhancing mass arising from the anterolateral cortex and extending from the tibial tuberosity to the mid-diaphysis of the left tibia. Histologic examination of the CT-guided core needle biopsy specimen showed that the lesion was composed of dense fibrocollagenous tissue separating irregular bony trabeculae with osteoblastic and osteoclastic activity. There was no evidence of any atypical cells, necrosis, or significant mitotic activity. No epithelial cells were identified on hematoxylin-eosin (H&E) stain (Figure 2). However, immunohistochemical staining was positive for focal cytokeratin-positive epithelial cells (Figure 3). The lesion was diagnosed as an osteofibrous dysplasia–like adamantinoma on the basis of the radiographic and histologic findings. We elected nonoperative intervention given the benign nature of the lesion and its potential to regress. Given the possibility of sampling error and potential for progression, the patient was followed regularly at 3- to 6-month intervals over a 2-year period without disease progression. 

Discussion

Osteofibrous dysplasia, osteofibrous dysplasia–like adamantinoma, and adamantinoma are rare fibro-osseous lesions that largely involve the midshaft of the tibia. Osteofibrous dysplasia accounts for 0.2% of primary bone tumors, whereas adamantinoma accounts for 0.1% to 0.5% of malignant bone tumors.⁹ Osteofibrous dysplasia is a benign lesion composed primarily of fibro-osseous tissue. Adamantinoma, however, is a slow-growing, low-grade, malignant biphasic tumor with intermingled epithelial and fibro-osseous components. It is an aggressive tumor that is locally invasive and can metastasize.² Osteofibrous dysplasia–like adamantinoma (also known as differentiated or regressing adamantinoma) is a benign lesion like osteofibrous dysplasia but has features of both osteofibrous dysplasia and adamantinoma. Osteofibrous dysplasia–like adamantinoma may progress and become a malignant adamantinoma.^6,10

The radiologic features of the 3 lesions are quite similar. It is not possible to distinguish between osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma based on imaging alone.⁹ Adamantinoma, being highly invasive, can be distinguished from osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma according to the extent of involvement of the medullary cavity seen on magnetic resonance imaging.⁹ Complete involvement of the medullary cavity is almost always seen in an adamantinoma. Involvement of the medullary cavity is minimal or absent in osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma lesions. 

Tissue confirmation through biopsy is crucial for accurate diagnosis. A biopsy should always be performed on any suspicious lesion,^3,6 and the fibro-osseous lesion should be treated as an adamantinoma if findings are equivocal. A biopsy also distinctly distinguishes these lesions from benign fibrous cortical defects, which have a similar radiographic appearance. While open biopsy is the gold standard, minimally invasive techniques such as core needle biopsy and fine needle biopsy are increasingly used.⁶ Because of the higher risk of misdiagnosis with minimally invasive techniques, radiographic confirmation is highly recommended.⁵

Histologically, both osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma do not stain for cytokeratin on H&E stain. However, they can be differentiated based on immunohistochemical staining for cytokeratin. Osteofibrous dysplasia lesions exhibit diffuse staining whereas osteofibrous dysplasia–like adamantinoma lesions show focal staining of small nests of epithelial cells. Adamantinoma, in comparison, exhibits a biphasic pattern on H&E stain, representing areas of epithelial and osteofibrous cells. Immunohistochemical staining for cytokeratin of an adamantinoma reveals large nests of epithelial cells.  

The association between osteofibrous dysplasia, osteofibrous dysplasia–like adamantinoma, and adamantinoma is not clearly established. However, it is widely believed that these 3 lesions represent a spectrum of the same disease and are linearly related in disease progression, with osteofibrous dysplasia at the benign end of the spectrum, osteofibrous dysplasia–like adamantinoma the intermediate form, and adamantinoma at the malignant end of the spectrum.¹¹

Hazelbag and colleagues⁶ and Springfield and colleagues¹⁰ point out that osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma could be precursor lesions of adamantinoma. We found several studies in the literature that support and document progression from osteofibrous dysplasia and osteofibrous dysplasia–like adamantinoma to an adamantinoma.^4,6,10,12 Other studies, however, showed no progression from either a benign osteofibrous dysplasia or an osteofibrous dysplasia–like adamantinoma lesion to a malignant adamantinoma. Park and colleagues¹³ described 41 cases of osteofibrous dysplasia that did not progress to adamantinoma. Kuruvilla and Steiner⁸ described 5 cases of osteofibrous dysplasia–like adamantinoma that showed no progression to adamantinoma. Additionally, our case has not progressed and has remained radiographically stable over a 2-year follow-up. Czerniak and colleagues⁷ and Ueda and colleagues¹⁴ postulated, based on histologic and immunohistochemical studies, that osteofibrous dysplasia–like adamantinoma might be a regressing form of an adamantinoma that is undergoing reparative processes that could result in complete elimination of all tumor cells.

In general, any lesion with absent to low malignant potential could be managed nonoperatively with periodic observation and without the need for surgical intervention. Thus, identification of a stable or nonprogressing osteofibrous dysplasia–like adamantinoma lesion has significant treatment implications. Campanacci and Laus¹⁵ at the Rizzoli Institute in Milan, through long term follow-up of their patients with osteofibrous dysplasia, found that most lesions had a tendency to regress spontaneously by puberty. They recommended that nonextensive osteofibrous dysplasia lesions should be observed, and surgery should be delayed until puberty. Gleason and colleagues¹⁶ also recommended nonoperative management of osteofibrous dysplasia lesions, with surgery used only for large, deforming, and highly invasive lesions. We recommend a similar treatment approach for osteofibrous dysplasia–like adamantinoma lesions.

Adamantinomas, however, are usually symptomatic, are highly invasive, have a high recurrence rate, and can metastasize.9 In these patients, a wide en bloc resection or amputation should be performed as soon as possible.11 Our case highlights that osteofibrous dysplasia–like adamantinoma lesions can occur in children and can remain stable, especially over the short term. Such lesions can be observed without surgical intervention.

Interscalene brachial plexus anesthesia is commonly used for arthroscopic and open procedures of the shoulder. This regional anesthetic targets the trunks of the brachial plexus and anesthetizes the area about the shoulder and proximal arm. Its use may obviate the need for concomitant general anesthesia, potentially reducing the use of postoperative intravenous and oral pain medication. Furthermore, patients often bypass the acute postoperative anesthesia care unit and proceed directly to the ambulatory unit, permitting earlier hospital discharge. Previous reports in the literature have demonstrated higher rates of neurologic, cardiac, and pulmonary complications from this procedure; in particular, the incidence of pneumothorax was reported as high as 3%.¹ Techniques to localize the nerves, such as electrical nerve stimulation and, more recently, ultrasound guidance, have reduced these complication rates.^2,3 Successful administration of the block has been shown to result in satisfactory postoperative pain relief.² However, ultrasound-guided interscalene nerve blocks remain operator-dependent and complications may still occur.

We report a case of tension pneumothorax after arthroscopic rotator cuff repair and subacromial decompression with an ultrasound-guided interscalene block. Immediate recognition and treatment of this complication resulted in a good clinical outcome. The patient provided written informed consent for print and electronic publication of this case report. 

Case Report

A 56-year-old woman presented with 3 months of right shoulder pain after a fall. Examination was pertinent for weakness in forward elevation and positive rotator cuff impingement signs. She remained symptomatic despite a course of nonsurgical management that included cortisone injections and physical therapy. Magnetic resonance imaging of the shoulder showed a full-thickness supraspinatus tear with minimal fatty atrophy. After a discussion of her treatment options, she elected to undergo an arthroscopic rotator cuff repair with subacromial decompression. An evaluation by her internist revealed no pertinent medical history apart from obesity (body mass index, 36). Specifically, there was no reported history of chronic obstructive pulmonary disease or asthma. She denied any prior cigarette smoking.

The patient was evaluated by the regional anesthesia team and was classified as a class 2 airway. An interscalene brachial plexus block was performed using a 2-inch, 22-gauge needle inserted into the interscalene groove. Using an out-of-plane technique under direct ultrasound guidance, 30 mL of 0.52% ropivacaine was injected. The block was considered successful, and no complications, such as resistance, paresthesias, pain, or blood on aspiration, were noted during injection. The patient had no complaints of chest pain or shortness of breath immediately afterward, and all vital signs were stable throughout the procedure.

The patient was brought to the operating room and placed in the beach-chair position. Induction for general anesthesia was started 15 minutes after the regional anesthetic, with 2 intubation attempts necessary because of poor airway visualization. After placement of the endotracheal tube, breath sounds were noted to be equal bilaterally. The arthroscopic procedure consisted of double-row rotator cuff repair, subacromial decompression, and débridement of the glenohumeral joint for synovitis, using standard arthroscopic portals. There were no difficulties with trocar placement, and bleeding was minimal throughout the case. The total surgical time was 150 minutes and a pump pressure of 30 mm Hg was maintained during the arthroscopy.

Within the first 60 minutes of the start of the arthroscopic procedure, the patient was noted to be intermittently hypotensive with mean arterial pressure (MAP) ranging from the 30s to 130s mm Hg and pulse in the 70 to 80 beats/min range. FiO₂ in the 85% to 95% range was maintained throughout the procedure. During that time, 50 μg phenylephrine was administered on 4 separate occasions to maintain her blood pressure. The labile blood pressure was attributed by the anesthesiologist to the beach-chair position. During an attempted extubation upon conclusion of the surgery, the patient became hypotensive with MAP that ranged from the 40s to 60s mm Hg and tachycardic to 90 beats/min. The oxygen saturation was in the low 90s and tidal volume was poor. Absent lung sounds were noted on the right chest. An urgent portable chest radiograph showed a large right-sided tension pneumothorax with mediastinal shift (Figure 1). After an immediate general surgery consultation, a chest tube was placed in the operating room. The patient’s vital signs improved and a repeat chest radiograph revealed successful re-expansion of the lung (Figure 2). She was transferred to the acute postoperative anesthesia care unit and extubated in the intensive care unit later that day.

The patient’s chest tube was removed 2 days later and she was discharged home on hospital day 5 with a completely resolved pneumothorax. She was seen 1 week later in the office for a postoperative visit and reported feeling well without chest pain or shortness of breath.

Discussion

Interscalene brachial plexus anesthesia was first described by Winnie⁴ in 1970. This block targets the trunks of the brachial plexus, which are enclosed in a fascial sheath between the anterior and middle scalene muscles. In this region lie several structures at risk: the phrenic nerve superficially and inferiorly; the carotid sheath located superficially and medially; the subclavian artery parallel to the trunks; and the cupula of the lung that lies deep and inferior to the anterior scalene muscle. Recognized complications of the block include vocal hoarseness, Horner syndrome, and hemidiaphragmatic paresis caused by the temporary blockade of the ipsilateral recurrent laryngeal nerve, stellate ganglion, and phrenic nerve, in that order.⁵ Use of the interscalene block has been associated with minimal risk for pneumothorax, because the needle entry point is superior and directed away from the lung pleura.⁶ This is in contrast to the more inferiorly placed supraclavicular block, located in closer proximity to the lung cupula.⁵

Two different approaches are commonly used during ultrasound-guided nerve blocks. The in-plane approach generates a long-axis view of the needle by advancing the needle parallel with the long axis of the ultrasound probe. While this allows direct visualization of the needle tip, it requires deeper needle insertion from lateral to medial, causing puncture of the middle scalene muscle that may increase patient discomfort and risk nerve injury within the muscle.⁷ The out-of-plane approach used on our patient involves needle insertion parallel to the brachial plexus, but along the short axis of the ultrasound probe. Although this permits the operator to assess the periphery of the nerve, it may lead to poor needle-tip visualization during the procedure. As a result, operators often use a combination of tissue disturbance and “hydrolocation,” in which fluid is injected to indicate the needle-tip location.^8,9

Tension pneumothorax represents the accumulation of air in the pleural space that leads to impaired pulmonary and cardiac function. It is often caused by disruption or puncture of the parietal or visceral pleura, creating a connection between the alveoli and pleural cavity. The gradual buildup of air in the pleural cavity results in increased intrapleural pressure, which compresses and ultimately collapses the ipsilateral lung. Venous compression restricts blood return to the heart and reduces cardiac output. Clinical manifestations include dyspnea, hypoxemia, tachycardia, and hypotension.¹⁰ Multiple techniques were developed to better localize the brachial plexus while reducing injury to nearby structures, including the lung. These include eliciting needle paresthesias, electrical nerve stimulation, and ultrasound guidance. While nerve stimulation was once the gold standard for brachial plexus localization, ultrasound guidance has gained in popularity because of its noninvasive nature and dynamic capability to identify nerves and surrounding structures.¹¹ Perlas and colleagues¹² determined the sensitivity of needle paresthesias and nerve stimulation to be 38% and 75%, respectively, in cases in which plexus localization had been confirmed by ultrasound.

Several studies have reported on the efficacy of interscalene nerve block with either nerve stimulation or ultrasound guidance in the setting of shoulder surgery.^2,3 Bishop and colleagues³ reviewed 547 patients who underwent interscalene regional anesthesia with nerve stimulation for both arthroscopic and open-shoulder procedures. They reported a 97% success rate and 12 (2.3%) minor complications, including sensory neuropathy and complex regional pain syndrome. There were no cases of pneumothorax, cardiac events, or other major complications.³ In a prospective study of 1319 patients, Singh and colleagues² reported a 99.6% success rate using ultrasound-guided interscalene blocks for their shoulder surgeries. A total of 38 adverse events (2.88%) were identified: 14 transient neurologic events, including ear numbness, digital numbness, and brachial plexitis; 1 case of intraoperative bradycardia, and 2 cancellations after the block for chest pain and flank pain, which yielded negative cardiac workups. Other complications included postoperative emergency room visits and hospital admissions for reasons unrelated to the block.² Interscalene regional anesthesia, therefore, provides effective anesthesia for shoulder surgery with low complication rates.

Pneumothorax after ultrasound-guided interscalene block has rarely been reported.^13,14 In a review of 144 ultrasound-guided indwelling interscalene catheter placements, a 98% successful block rate with a single complication of small pneumothorax after total shoulder arthroplasty was reported.¹³ Mandim and colleagues¹⁴ reported a case of pneumothorax in a smoker who underwent an ultrasound-guided brachial plexus block prior to open reduction and internal fixation of an ulnar fracture. While the patient was asymptomatic and vital signs remained stable during the procedure, the patient complained postoperatively of chest pain with hypoxia, tachycardia, and hypotension. A chest radiograph confirmed an ipsilateral pneumothorax, and the patient was treated successfully with chest-tube placement. The authors attributed this complication to a higher pleural dome resulting from a hyperinflated lung caused by chronic smoking. Our patient reported no history of smoking and her preoperative chest radiograph had no evidence of lung disease.

In contrast, several cases of pneumothorax after shoulder surgery have been reported in the absence of nerve block. Oldman and Peng¹ reported a 41-year-old nonsmoker who underwent arthroscopic labral repair and subacromial decompression. The preoperative nerve block was cancelled, and the patient received general endotracheal anesthesia alone. Fifty minutes after the case, the patient developed chest pain and hypoxia. A chest radiograph showed a small pneumothorax that was managed conservatively. The pneumothorax was attributed to spontaneous rupture of a preexisting lung bulla, suggesting that blocks are not always the cause of this complication. Furthermore, Dietzel and Ciullo¹⁵ reported 4 cases of spontaneous pneumothorax within 24 hours of uncomplicated arthroscopic shoulder procedures under general anesthesia in the lateral decubitus position. The patient ages ranged from 22 to 38 years, and medical histories were all significant for preexisting lung disease, remote history of pneumonia, and heavy smoking. Three of the patients experienced symptoms at home the day after surgery. The authors concluded that these cases were likely caused by rupture of blebs or bullae from underlying lung disease; these ruptured blebs or bullae are difficult to detect and usually located in the upper lung. The pressure gradient from the positive pressure of anesthesia and the ipsilateral upper lung is thought to be highest in the lateral decubitus position, increasing their chance of rupture.¹⁵

Finally, Lee and colleagues¹⁶ described 3 patients aged 40 to 45 years who underwent uncomplicated subacromial decompression in the beach-chair position under general anesthesia. Significant shoulder, neck, and axillary swelling were noted after surgery, and a chest radiograph showed tension pneumothorax, subcutaneous emphysema, and pneumomediastinum. The authors speculated that pressure in the subacromial space may become negative relative to atmospheric pressure when the shaver and suction are running, drawing in air through other portals. When the suction is discontinued, fluid infusion may push air into the surrounding tissue, leading to subcutaneous emphysema, which may spread to the mediastinum.¹⁶ 

Conclusion 

Ultrasound-guided interscalene nerve blocks have successfully provided anesthesia for shoulder surgeries with low complication rates. Although the incidence of pneumothorax has decreased significantly with ultrasound guidance, the success of this procedure is highly operator-dependent. We present the case of an otherwise healthy patient without known pulmonary disease who developed a tension pneumothorax after the administration of ultrasound-guided regional and general anesthesia for arthroscopic shoulder surgery. Orthopedic surgeons and anesthesiologists must remain vigilant for pneumothorax during the perioperative period after shoulder surgery performed under interscalene regional aesthesia, particularly in the setting of hypotension, hypoxia, and/or tachycardia. Risk factors, such as history of smoking and preexisting lung disease, may predispose patients to the development of pneumothorax. Timely recognition and placement of a chest tube result in satisfactory clinical outcomes.

Interscalene brachial plexus anesthesia is commonly used for arthroscopic and open procedures of the shoulder. This regional anesthetic targets the trunks of the brachial plexus and anesthetizes the area about the shoulder and proximal arm. Its use may obviate the need for concomitant general anesthesia, potentially reducing the use of postoperative intravenous and oral pain medication. Furthermore, patients often bypass the acute postoperative anesthesia care unit and proceed directly to the ambulatory unit, permitting earlier hospital discharge. Previous reports in the literature have demonstrated higher rates of neurologic, cardiac, and pulmonary complications from this procedure; in particular, the incidence of pneumothorax was reported as high as 3%.¹ Techniques to localize the nerves, such as electrical nerve stimulation and, more recently, ultrasound guidance, have reduced these complication rates.^2,3 Successful administration of the block has been shown to result in satisfactory postoperative pain relief.² However, ultrasound-guided interscalene nerve blocks remain operator-dependent and complications may still occur.

We report a case of tension pneumothorax after arthroscopic rotator cuff repair and subacromial decompression with an ultrasound-guided interscalene block. Immediate recognition and treatment of this complication resulted in a good clinical outcome. The patient provided written informed consent for print and electronic publication of this case report. 

Case Report

A 56-year-old woman presented with 3 months of right shoulder pain after a fall. Examination was pertinent for weakness in forward elevation and positive rotator cuff impingement signs. She remained symptomatic despite a course of nonsurgical management that included cortisone injections and physical therapy. Magnetic resonance imaging of the shoulder showed a full-thickness supraspinatus tear with minimal fatty atrophy. After a discussion of her treatment options, she elected to undergo an arthroscopic rotator cuff repair with subacromial decompression. An evaluation by her internist revealed no pertinent medical history apart from obesity (body mass index, 36). Specifically, there was no reported history of chronic obstructive pulmonary disease or asthma. She denied any prior cigarette smoking.

The patient was evaluated by the regional anesthesia team and was classified as a class 2 airway. An interscalene brachial plexus block was performed using a 2-inch, 22-gauge needle inserted into the interscalene groove. Using an out-of-plane technique under direct ultrasound guidance, 30 mL of 0.52% ropivacaine was injected. The block was considered successful, and no complications, such as resistance, paresthesias, pain, or blood on aspiration, were noted during injection. The patient had no complaints of chest pain or shortness of breath immediately afterward, and all vital signs were stable throughout the procedure.

The patient was brought to the operating room and placed in the beach-chair position. Induction for general anesthesia was started 15 minutes after the regional anesthetic, with 2 intubation attempts necessary because of poor airway visualization. After placement of the endotracheal tube, breath sounds were noted to be equal bilaterally. The arthroscopic procedure consisted of double-row rotator cuff repair, subacromial decompression, and débridement of the glenohumeral joint for synovitis, using standard arthroscopic portals. There were no difficulties with trocar placement, and bleeding was minimal throughout the case. The total surgical time was 150 minutes and a pump pressure of 30 mm Hg was maintained during the arthroscopy.

Within the first 60 minutes of the start of the arthroscopic procedure, the patient was noted to be intermittently hypotensive with mean arterial pressure (MAP) ranging from the 30s to 130s mm Hg and pulse in the 70 to 80 beats/min range. FiO₂ in the 85% to 95% range was maintained throughout the procedure. During that time, 50 μg phenylephrine was administered on 4 separate occasions to maintain her blood pressure. The labile blood pressure was attributed by the anesthesiologist to the beach-chair position. During an attempted extubation upon conclusion of the surgery, the patient became hypotensive with MAP that ranged from the 40s to 60s mm Hg and tachycardic to 90 beats/min. The oxygen saturation was in the low 90s and tidal volume was poor. Absent lung sounds were noted on the right chest. An urgent portable chest radiograph showed a large right-sided tension pneumothorax with mediastinal shift (Figure 1). After an immediate general surgery consultation, a chest tube was placed in the operating room. The patient’s vital signs improved and a repeat chest radiograph revealed successful re-expansion of the lung (Figure 2). She was transferred to the acute postoperative anesthesia care unit and extubated in the intensive care unit later that day.

The patient’s chest tube was removed 2 days later and she was discharged home on hospital day 5 with a completely resolved pneumothorax. She was seen 1 week later in the office for a postoperative visit and reported feeling well without chest pain or shortness of breath.

Discussion

Interscalene brachial plexus anesthesia was first described by Winnie⁴ in 1970. This block targets the trunks of the brachial plexus, which are enclosed in a fascial sheath between the anterior and middle scalene muscles. In this region lie several structures at risk: the phrenic nerve superficially and inferiorly; the carotid sheath located superficially and medially; the subclavian artery parallel to the trunks; and the cupula of the lung that lies deep and inferior to the anterior scalene muscle. Recognized complications of the block include vocal hoarseness, Horner syndrome, and hemidiaphragmatic paresis caused by the temporary blockade of the ipsilateral recurrent laryngeal nerve, stellate ganglion, and phrenic nerve, in that order.⁵ Use of the interscalene block has been associated with minimal risk for pneumothorax, because the needle entry point is superior and directed away from the lung pleura.⁶ This is in contrast to the more inferiorly placed supraclavicular block, located in closer proximity to the lung cupula.⁵

Two different approaches are commonly used during ultrasound-guided nerve blocks. The in-plane approach generates a long-axis view of the needle by advancing the needle parallel with the long axis of the ultrasound probe. While this allows direct visualization of the needle tip, it requires deeper needle insertion from lateral to medial, causing puncture of the middle scalene muscle that may increase patient discomfort and risk nerve injury within the muscle.⁷ The out-of-plane approach used on our patient involves needle insertion parallel to the brachial plexus, but along the short axis of the ultrasound probe. Although this permits the operator to assess the periphery of the nerve, it may lead to poor needle-tip visualization during the procedure. As a result, operators often use a combination of tissue disturbance and “hydrolocation,” in which fluid is injected to indicate the needle-tip location.^8,9

Tension pneumothorax represents the accumulation of air in the pleural space that leads to impaired pulmonary and cardiac function. It is often caused by disruption or puncture of the parietal or visceral pleura, creating a connection between the alveoli and pleural cavity. The gradual buildup of air in the pleural cavity results in increased intrapleural pressure, which compresses and ultimately collapses the ipsilateral lung. Venous compression restricts blood return to the heart and reduces cardiac output. Clinical manifestations include dyspnea, hypoxemia, tachycardia, and hypotension.¹⁰ Multiple techniques were developed to better localize the brachial plexus while reducing injury to nearby structures, including the lung. These include eliciting needle paresthesias, electrical nerve stimulation, and ultrasound guidance. While nerve stimulation was once the gold standard for brachial plexus localization, ultrasound guidance has gained in popularity because of its noninvasive nature and dynamic capability to identify nerves and surrounding structures.¹¹ Perlas and colleagues¹² determined the sensitivity of needle paresthesias and nerve stimulation to be 38% and 75%, respectively, in cases in which plexus localization had been confirmed by ultrasound.

Several studies have reported on the efficacy of interscalene nerve block with either nerve stimulation or ultrasound guidance in the setting of shoulder surgery.^2,3 Bishop and colleagues³ reviewed 547 patients who underwent interscalene regional anesthesia with nerve stimulation for both arthroscopic and open-shoulder procedures. They reported a 97% success rate and 12 (2.3%) minor complications, including sensory neuropathy and complex regional pain syndrome. There were no cases of pneumothorax, cardiac events, or other major complications.³ In a prospective study of 1319 patients, Singh and colleagues² reported a 99.6% success rate using ultrasound-guided interscalene blocks for their shoulder surgeries. A total of 38 adverse events (2.88%) were identified: 14 transient neurologic events, including ear numbness, digital numbness, and brachial plexitis; 1 case of intraoperative bradycardia, and 2 cancellations after the block for chest pain and flank pain, which yielded negative cardiac workups. Other complications included postoperative emergency room visits and hospital admissions for reasons unrelated to the block.² Interscalene regional anesthesia, therefore, provides effective anesthesia for shoulder surgery with low complication rates.

Pneumothorax after ultrasound-guided interscalene block has rarely been reported.^13,14 In a review of 144 ultrasound-guided indwelling interscalene catheter placements, a 98% successful block rate with a single complication of small pneumothorax after total shoulder arthroplasty was reported.¹³ Mandim and colleagues¹⁴ reported a case of pneumothorax in a smoker who underwent an ultrasound-guided brachial plexus block prior to open reduction and internal fixation of an ulnar fracture. While the patient was asymptomatic and vital signs remained stable during the procedure, the patient complained postoperatively of chest pain with hypoxia, tachycardia, and hypotension. A chest radiograph confirmed an ipsilateral pneumothorax, and the patient was treated successfully with chest-tube placement. The authors attributed this complication to a higher pleural dome resulting from a hyperinflated lung caused by chronic smoking. Our patient reported no history of smoking and her preoperative chest radiograph had no evidence of lung disease.

In contrast, several cases of pneumothorax after shoulder surgery have been reported in the absence of nerve block. Oldman and Peng¹ reported a 41-year-old nonsmoker who underwent arthroscopic labral repair and subacromial decompression. The preoperative nerve block was cancelled, and the patient received general endotracheal anesthesia alone. Fifty minutes after the case, the patient developed chest pain and hypoxia. A chest radiograph showed a small pneumothorax that was managed conservatively. The pneumothorax was attributed to spontaneous rupture of a preexisting lung bulla, suggesting that blocks are not always the cause of this complication. Furthermore, Dietzel and Ciullo¹⁵ reported 4 cases of spontaneous pneumothorax within 24 hours of uncomplicated arthroscopic shoulder procedures under general anesthesia in the lateral decubitus position. The patient ages ranged from 22 to 38 years, and medical histories were all significant for preexisting lung disease, remote history of pneumonia, and heavy smoking. Three of the patients experienced symptoms at home the day after surgery. The authors concluded that these cases were likely caused by rupture of blebs or bullae from underlying lung disease; these ruptured blebs or bullae are difficult to detect and usually located in the upper lung. The pressure gradient from the positive pressure of anesthesia and the ipsilateral upper lung is thought to be highest in the lateral decubitus position, increasing their chance of rupture.¹⁵

Finally, Lee and colleagues¹⁶ described 3 patients aged 40 to 45 years who underwent uncomplicated subacromial decompression in the beach-chair position under general anesthesia. Significant shoulder, neck, and axillary swelling were noted after surgery, and a chest radiograph showed tension pneumothorax, subcutaneous emphysema, and pneumomediastinum. The authors speculated that pressure in the subacromial space may become negative relative to atmospheric pressure when the shaver and suction are running, drawing in air through other portals. When the suction is discontinued, fluid infusion may push air into the surrounding tissue, leading to subcutaneous emphysema, which may spread to the mediastinum.¹⁶ 

Conclusion 

Ultrasound-guided interscalene nerve blocks have successfully provided anesthesia for shoulder surgeries with low complication rates. Although the incidence of pneumothorax has decreased significantly with ultrasound guidance, the success of this procedure is highly operator-dependent. We present the case of an otherwise healthy patient without known pulmonary disease who developed a tension pneumothorax after the administration of ultrasound-guided regional and general anesthesia for arthroscopic shoulder surgery. Orthopedic surgeons and anesthesiologists must remain vigilant for pneumothorax during the perioperative period after shoulder surgery performed under interscalene regional aesthesia, particularly in the setting of hypotension, hypoxia, and/or tachycardia. Risk factors, such as history of smoking and preexisting lung disease, may predispose patients to the development of pneumothorax. Timely recognition and placement of a chest tube result in satisfactory clinical outcomes.

Interscalene brachial plexus anesthesia is commonly used for arthroscopic and open procedures of the shoulder. This regional anesthetic targets the trunks of the brachial plexus and anesthetizes the area about the shoulder and proximal arm. Its use may obviate the need for concomitant general anesthesia, potentially reducing the use of postoperative intravenous and oral pain medication. Furthermore, patients often bypass the acute postoperative anesthesia care unit and proceed directly to the ambulatory unit, permitting earlier hospital discharge. Previous reports in the literature have demonstrated higher rates of neurologic, cardiac, and pulmonary complications from this procedure; in particular, the incidence of pneumothorax was reported as high as 3%.¹ Techniques to localize the nerves, such as electrical nerve stimulation and, more recently, ultrasound guidance, have reduced these complication rates.^2,3 Successful administration of the block has been shown to result in satisfactory postoperative pain relief.² However, ultrasound-guided interscalene nerve blocks remain operator-dependent and complications may still occur.

We report a case of tension pneumothorax after arthroscopic rotator cuff repair and subacromial decompression with an ultrasound-guided interscalene block. Immediate recognition and treatment of this complication resulted in a good clinical outcome. The patient provided written informed consent for print and electronic publication of this case report. 

Case Report

A 56-year-old woman presented with 3 months of right shoulder pain after a fall. Examination was pertinent for weakness in forward elevation and positive rotator cuff impingement signs. She remained symptomatic despite a course of nonsurgical management that included cortisone injections and physical therapy. Magnetic resonance imaging of the shoulder showed a full-thickness supraspinatus tear with minimal fatty atrophy. After a discussion of her treatment options, she elected to undergo an arthroscopic rotator cuff repair with subacromial decompression. An evaluation by her internist revealed no pertinent medical history apart from obesity (body mass index, 36). Specifically, there was no reported history of chronic obstructive pulmonary disease or asthma. She denied any prior cigarette smoking.

The patient was evaluated by the regional anesthesia team and was classified as a class 2 airway. An interscalene brachial plexus block was performed using a 2-inch, 22-gauge needle inserted into the interscalene groove. Using an out-of-plane technique under direct ultrasound guidance, 30 mL of 0.52% ropivacaine was injected. The block was considered successful, and no complications, such as resistance, paresthesias, pain, or blood on aspiration, were noted during injection. The patient had no complaints of chest pain or shortness of breath immediately afterward, and all vital signs were stable throughout the procedure.

The patient was brought to the operating room and placed in the beach-chair position. Induction for general anesthesia was started 15 minutes after the regional anesthetic, with 2 intubation attempts necessary because of poor airway visualization. After placement of the endotracheal tube, breath sounds were noted to be equal bilaterally. The arthroscopic procedure consisted of double-row rotator cuff repair, subacromial decompression, and débridement of the glenohumeral joint for synovitis, using standard arthroscopic portals. There were no difficulties with trocar placement, and bleeding was minimal throughout the case. The total surgical time was 150 minutes and a pump pressure of 30 mm Hg was maintained during the arthroscopy.

Within the first 60 minutes of the start of the arthroscopic procedure, the patient was noted to be intermittently hypotensive with mean arterial pressure (MAP) ranging from the 30s to 130s mm Hg and pulse in the 70 to 80 beats/min range. FiO₂ in the 85% to 95% range was maintained throughout the procedure. During that time, 50 μg phenylephrine was administered on 4 separate occasions to maintain her blood pressure. The labile blood pressure was attributed by the anesthesiologist to the beach-chair position. During an attempted extubation upon conclusion of the surgery, the patient became hypotensive with MAP that ranged from the 40s to 60s mm Hg and tachycardic to 90 beats/min. The oxygen saturation was in the low 90s and tidal volume was poor. Absent lung sounds were noted on the right chest. An urgent portable chest radiograph showed a large right-sided tension pneumothorax with mediastinal shift (Figure 1). After an immediate general surgery consultation, a chest tube was placed in the operating room. The patient’s vital signs improved and a repeat chest radiograph revealed successful re-expansion of the lung (Figure 2). She was transferred to the acute postoperative anesthesia care unit and extubated in the intensive care unit later that day.

The patient’s chest tube was removed 2 days later and she was discharged home on hospital day 5 with a completely resolved pneumothorax. She was seen 1 week later in the office for a postoperative visit and reported feeling well without chest pain or shortness of breath.

Discussion

Interscalene brachial plexus anesthesia was first described by Winnie⁴ in 1970. This block targets the trunks of the brachial plexus, which are enclosed in a fascial sheath between the anterior and middle scalene muscles. In this region lie several structures at risk: the phrenic nerve superficially and inferiorly; the carotid sheath located superficially and medially; the subclavian artery parallel to the trunks; and the cupula of the lung that lies deep and inferior to the anterior scalene muscle. Recognized complications of the block include vocal hoarseness, Horner syndrome, and hemidiaphragmatic paresis caused by the temporary blockade of the ipsilateral recurrent laryngeal nerve, stellate ganglion, and phrenic nerve, in that order.⁵ Use of the interscalene block has been associated with minimal risk for pneumothorax, because the needle entry point is superior and directed away from the lung pleura.⁶ This is in contrast to the more inferiorly placed supraclavicular block, located in closer proximity to the lung cupula.⁵

Two different approaches are commonly used during ultrasound-guided nerve blocks. The in-plane approach generates a long-axis view of the needle by advancing the needle parallel with the long axis of the ultrasound probe. While this allows direct visualization of the needle tip, it requires deeper needle insertion from lateral to medial, causing puncture of the middle scalene muscle that may increase patient discomfort and risk nerve injury within the muscle.⁷ The out-of-plane approach used on our patient involves needle insertion parallel to the brachial plexus, but along the short axis of the ultrasound probe. Although this permits the operator to assess the periphery of the nerve, it may lead to poor needle-tip visualization during the procedure. As a result, operators often use a combination of tissue disturbance and “hydrolocation,” in which fluid is injected to indicate the needle-tip location.^8,9

Tension pneumothorax represents the accumulation of air in the pleural space that leads to impaired pulmonary and cardiac function. It is often caused by disruption or puncture of the parietal or visceral pleura, creating a connection between the alveoli and pleural cavity. The gradual buildup of air in the pleural cavity results in increased intrapleural pressure, which compresses and ultimately collapses the ipsilateral lung. Venous compression restricts blood return to the heart and reduces cardiac output. Clinical manifestations include dyspnea, hypoxemia, tachycardia, and hypotension.¹⁰ Multiple techniques were developed to better localize the brachial plexus while reducing injury to nearby structures, including the lung. These include eliciting needle paresthesias, electrical nerve stimulation, and ultrasound guidance. While nerve stimulation was once the gold standard for brachial plexus localization, ultrasound guidance has gained in popularity because of its noninvasive nature and dynamic capability to identify nerves and surrounding structures.¹¹ Perlas and colleagues¹² determined the sensitivity of needle paresthesias and nerve stimulation to be 38% and 75%, respectively, in cases in which plexus localization had been confirmed by ultrasound.

Several studies have reported on the efficacy of interscalene nerve block with either nerve stimulation or ultrasound guidance in the setting of shoulder surgery.^2,3 Bishop and colleagues³ reviewed 547 patients who underwent interscalene regional anesthesia with nerve stimulation for both arthroscopic and open-shoulder procedures. They reported a 97% success rate and 12 (2.3%) minor complications, including sensory neuropathy and complex regional pain syndrome. There were no cases of pneumothorax, cardiac events, or other major complications.³ In a prospective study of 1319 patients, Singh and colleagues² reported a 99.6% success rate using ultrasound-guided interscalene blocks for their shoulder surgeries. A total of 38 adverse events (2.88%) were identified: 14 transient neurologic events, including ear numbness, digital numbness, and brachial plexitis; 1 case of intraoperative bradycardia, and 2 cancellations after the block for chest pain and flank pain, which yielded negative cardiac workups. Other complications included postoperative emergency room visits and hospital admissions for reasons unrelated to the block.² Interscalene regional anesthesia, therefore, provides effective anesthesia for shoulder surgery with low complication rates.

Pneumothorax after ultrasound-guided interscalene block has rarely been reported.^13,14 In a review of 144 ultrasound-guided indwelling interscalene catheter placements, a 98% successful block rate with a single complication of small pneumothorax after total shoulder arthroplasty was reported.¹³ Mandim and colleagues¹⁴ reported a case of pneumothorax in a smoker who underwent an ultrasound-guided brachial plexus block prior to open reduction and internal fixation of an ulnar fracture. While the patient was asymptomatic and vital signs remained stable during the procedure, the patient complained postoperatively of chest pain with hypoxia, tachycardia, and hypotension. A chest radiograph confirmed an ipsilateral pneumothorax, and the patient was treated successfully with chest-tube placement. The authors attributed this complication to a higher pleural dome resulting from a hyperinflated lung caused by chronic smoking. Our patient reported no history of smoking and her preoperative chest radiograph had no evidence of lung disease.

In contrast, several cases of pneumothorax after shoulder surgery have been reported in the absence of nerve block. Oldman and Peng¹ reported a 41-year-old nonsmoker who underwent arthroscopic labral repair and subacromial decompression. The preoperative nerve block was cancelled, and the patient received general endotracheal anesthesia alone. Fifty minutes after the case, the patient developed chest pain and hypoxia. A chest radiograph showed a small pneumothorax that was managed conservatively. The pneumothorax was attributed to spontaneous rupture of a preexisting lung bulla, suggesting that blocks are not always the cause of this complication. Furthermore, Dietzel and Ciullo¹⁵ reported 4 cases of spontaneous pneumothorax within 24 hours of uncomplicated arthroscopic shoulder procedures under general anesthesia in the lateral decubitus position. The patient ages ranged from 22 to 38 years, and medical histories were all significant for preexisting lung disease, remote history of pneumonia, and heavy smoking. Three of the patients experienced symptoms at home the day after surgery. The authors concluded that these cases were likely caused by rupture of blebs or bullae from underlying lung disease; these ruptured blebs or bullae are difficult to detect and usually located in the upper lung. The pressure gradient from the positive pressure of anesthesia and the ipsilateral upper lung is thought to be highest in the lateral decubitus position, increasing their chance of rupture.¹⁵

Finally, Lee and colleagues¹⁶ described 3 patients aged 40 to 45 years who underwent uncomplicated subacromial decompression in the beach-chair position under general anesthesia. Significant shoulder, neck, and axillary swelling were noted after surgery, and a chest radiograph showed tension pneumothorax, subcutaneous emphysema, and pneumomediastinum. The authors speculated that pressure in the subacromial space may become negative relative to atmospheric pressure when the shaver and suction are running, drawing in air through other portals. When the suction is discontinued, fluid infusion may push air into the surrounding tissue, leading to subcutaneous emphysema, which may spread to the mediastinum.¹⁶ 

Conclusion 

Ultrasound-guided interscalene nerve blocks have successfully provided anesthesia for shoulder surgeries with low complication rates. Although the incidence of pneumothorax has decreased significantly with ultrasound guidance, the success of this procedure is highly operator-dependent. We present the case of an otherwise healthy patient without known pulmonary disease who developed a tension pneumothorax after the administration of ultrasound-guided regional and general anesthesia for arthroscopic shoulder surgery. Orthopedic surgeons and anesthesiologists must remain vigilant for pneumothorax during the perioperative period after shoulder surgery performed under interscalene regional aesthesia, particularly in the setting of hypotension, hypoxia, and/or tachycardia. Risk factors, such as history of smoking and preexisting lung disease, may predispose patients to the development of pneumothorax. Timely recognition and placement of a chest tube result in satisfactory clinical outcomes.

Osteoid osteoma of the talar neck is an unusual clinical condition that is often overlooked on initial assessment of patients with ankle pain. Here, we present a case report of an adolescent male with talar neck osteoid osteoma who reported persistent pain after an injury. We discuss the differential diagnosis of persistent anterior ankle pain and assess the treatment options for osteoid osteoma of the talar neck. The patient’s guardian provided written informed consent for print and electronic publication of this case report. 

Case Report

A 13-year-old boy presented to our clinic 3 months after a right ankle sprain. He had visited the emergency department at the time of injury; radiographs of the ankle were reported negative for fractures, dislocations, or bone pathologies. He was treated conservatively with elastic support, icing, rest, elevation, and weight-bearing as tolerated. Upon presentation to our office, his pain involved the entire ankle joint. He had not put weight on it since the injury. On examination, he had a significant limp, anteromedial swelling, and tenderness over the ankle joint anteromedially. His neurologic and vascular examinations were normal.

His plain radiographs showed a cystic mass, located at the dorsal aspect of the talar neck (Figures 1A, 1B). Computed tomography (CT) showed a round lucent lesion involving the superior aspect of the talar neck, measuring 9 mm by 6 mm. A sclerotic radiodense focus was evident in the center (Figures 2A, 2B). Noncontrast multiplanar, multisequence magnetic resonance imaging (MRI) showed abnormal edema throughout the talus and a 9-mm rounded ossicle overlying the superior margin of the neck of the talus (Figures 3A, 3B).

Differential Diagnosis

The differential diagnosis for anterior ankle pain includes ankle sprain, monoarticular arthritis, anterior ankle impingement, and talar neck fractures. Other related findings include the presence of a talar ridge and a talar beak.

Ankle sprains are very common injuries. The mainstay treatment consists of ice, resting, elevation, and elastic or semirigid support, and patients usually recover over the course of a few weeks. These sprains are typically injuries of the lateral or medial ligaments of the ankle. Extension of a ligament tear across the anterior capsule can explain persistent anterior ankle pain. The presence of a bony lesion on plain radiographs, however, makes the diagnosis of an ankle sprain, with or without extension into the anterior capsule, less likely.

Monoarticular arthritis, which may present in the ankle and has a wide differential diagnosis, usually involves the whole joint.

Anterior ankle impingement typically occurs in athletes who participate in sports that involve kicking. It can be a bony or soft-tissue impingent. Clinically, patients present with pain and loss of motion, specifically dorsiflexion.

Talar neck fractures are usually the result of high-energy trauma. Stress fractures of the neck of the talus are uncommon and are associated with a recent sudden increase in physical activity, such as running, dancing, or military training. Radiographs, CT scans, and MRI help define the fracture line.

The talar ridge is the site of capsular and ligamentous attachment on the superior aspect of the talar neck and may become hypertrophic in athletes. A hypertrophic talar ridge is asymptomatic and is not considered a pathologic finding on radiographs.

The talar beak, a flaring of the anterosuperior aspect of the talar head, is an indirect sign of tarsal coalition. When symptomatic, patients complain of subtalar symptoms, typically pain and limitation of motion. It usually does not present acutely.

Treatment

We offered the patient surgical excision, and his guardian consented to left ankle arthroscopy. We performed synovectomy using a combination of 3.5-mm shaver and radiofrequency probe. We identified the mass: round, soft, and located at the superior-medial aspect of the talar neck. We removed it in piecemeal fashion using manual arthroscopic instruments, and cauterized its base using the radiofrequency probe. We allowed the patient weight-bearing as tolerated starting the day after surgery.

We submitted the specimen for pathologic evaluation (Figure 4). It consisted of multiple pieces of tan/brown tissue. Histologic examination showed benign osteoblastic proliferation composed of anastomosing bony trabeculae with variable mineralization, lined by plump osteoblasts, within vascularized connective tissue; benign giant cells were present, consistent with a nidus of an osteoid osteoma.

On the first postoperative visit, the patient was pain-free and bearing weight with crutches. He was gradually weaned from his crutches and returned to full weight-bearing over the next 4 weeks. At 12-month follow-up, he was symptom-free with good range of motion and full return to previous level of activity.

Discussion

Osteoid osteoma is a small, benign, well-circumscribed osteoblastic cortical lesion, typically identified in long bones or, less frequently, in the subperiosteal region.¹ It often affects adolescents. Osteoid osteoma has been described in the talus in a few case series^2-7 and is associated with a typical nidus that can be identified on CT scans. It does not present acutely, however. The typical presentation for osteoid osteoma is bone pain at night that responds to nonsteroidal anti-inflammatory drugs. However, this presentation is not universal and is frequently missed.²

Juxta-articular osteoid osteomas in the ankle and foot can be difficult to diagnose. The most common site is the talus.³ The majority of patients link their pain to a remote ankle injury. The time delay to diagnosis is on average 2.5 years, but it can be as long as 10 years.^4-6 A CT scan is the best method to identify the nidus; MRI can be misleading if it shows only marrow edema but not a nidus.^4,5,7 In our patient, an injury was documented, and the patient denied prior symptoms. We cannot explain how an injury would trigger the formation of an osteoid osteoma or cause a previously asymptomatic osteoid osteoma to become symptomatic.

Medical treatment with nonsteroidal anti-inflammatory drugs has been used but is reported to take 2 to 4 years for resolution of symptoms; many patients may consider the treatment time frame too long when other alternatives are available.⁸ These include open resection, arthroscopic resection, and image-guided ablation. Open surgical techniques include en bloc resection and curettage. Bone grafting or internal fixation may be performed as needed. Arthroscopic excision of juxta-articular osteoid osteomas offers the advantages of good visualization and avoidance of soft-tissue dissection, and allows for complete excision of the lesion as well as synovectomy.^6,9,10 Arthroscopic excision also allows for quicker rehabilitation. Image-guided ablation, such as radionuclide-guided excision, CT-guided thermal ablation, and laser photocoagulation, may be even less invasive but do not allow for direct visualization, complete resection, and biopsy.¹¹

Conclusion

Osteoid osteoma is a small, benign, well-circumscribed osteoblastic cortical lesion, typically identified in long bones or, less frequently, in the subperiosteal region.¹ It often affects adolescents. Osteoid osteoma has been described in the talus in multiple case series and is associated with a typical nidus that can be identified on CT scans. Usually, it does not present acutely. The typical presentation for osteoid osteoma is bone pain at night that responds to nonsteroidal anti-inflammatory drugs. This presentation is not universal, however, and is frequently missed, especially when the pain is associated with a prior injury.² Arthroscopic exploration of the ankle with resection of subperiosteal osteoid osteoma and the associated synovitis using thermal ablation of the base with radiofrequency offers lasting cure with minimal morbidity.

Osteoid osteoma of the talar neck is an unusual clinical condition that is often overlooked on initial assessment of patients with ankle pain. Here, we present a case report of an adolescent male with talar neck osteoid osteoma who reported persistent pain after an injury. We discuss the differential diagnosis of persistent anterior ankle pain and assess the treatment options for osteoid osteoma of the talar neck. The patient’s guardian provided written informed consent for print and electronic publication of this case report. 

Case Report

A 13-year-old boy presented to our clinic 3 months after a right ankle sprain. He had visited the emergency department at the time of injury; radiographs of the ankle were reported negative for fractures, dislocations, or bone pathologies. He was treated conservatively with elastic support, icing, rest, elevation, and weight-bearing as tolerated. Upon presentation to our office, his pain involved the entire ankle joint. He had not put weight on it since the injury. On examination, he had a significant limp, anteromedial swelling, and tenderness over the ankle joint anteromedially. His neurologic and vascular examinations were normal.

His plain radiographs showed a cystic mass, located at the dorsal aspect of the talar neck (Figures 1A, 1B). Computed tomography (CT) showed a round lucent lesion involving the superior aspect of the talar neck, measuring 9 mm by 6 mm. A sclerotic radiodense focus was evident in the center (Figures 2A, 2B). Noncontrast multiplanar, multisequence magnetic resonance imaging (MRI) showed abnormal edema throughout the talus and a 9-mm rounded ossicle overlying the superior margin of the neck of the talus (Figures 3A, 3B).

Differential Diagnosis

The differential diagnosis for anterior ankle pain includes ankle sprain, monoarticular arthritis, anterior ankle impingement, and talar neck fractures. Other related findings include the presence of a talar ridge and a talar beak.

Ankle sprains are very common injuries. The mainstay treatment consists of ice, resting, elevation, and elastic or semirigid support, and patients usually recover over the course of a few weeks. These sprains are typically injuries of the lateral or medial ligaments of the ankle. Extension of a ligament tear across the anterior capsule can explain persistent anterior ankle pain. The presence of a bony lesion on plain radiographs, however, makes the diagnosis of an ankle sprain, with or without extension into the anterior capsule, less likely.

Monoarticular arthritis, which may present in the ankle and has a wide differential diagnosis, usually involves the whole joint.

Anterior ankle impingement typically occurs in athletes who participate in sports that involve kicking. It can be a bony or soft-tissue impingent. Clinically, patients present with pain and loss of motion, specifically dorsiflexion.

Talar neck fractures are usually the result of high-energy trauma. Stress fractures of the neck of the talus are uncommon and are associated with a recent sudden increase in physical activity, such as running, dancing, or military training. Radiographs, CT scans, and MRI help define the fracture line.

The talar ridge is the site of capsular and ligamentous attachment on the superior aspect of the talar neck and may become hypertrophic in athletes. A hypertrophic talar ridge is asymptomatic and is not considered a pathologic finding on radiographs.

The talar beak, a flaring of the anterosuperior aspect of the talar head, is an indirect sign of tarsal coalition. When symptomatic, patients complain of subtalar symptoms, typically pain and limitation of motion. It usually does not present acutely.

Treatment

We offered the patient surgical excision, and his guardian consented to left ankle arthroscopy. We performed synovectomy using a combination of 3.5-mm shaver and radiofrequency probe. We identified the mass: round, soft, and located at the superior-medial aspect of the talar neck. We removed it in piecemeal fashion using manual arthroscopic instruments, and cauterized its base using the radiofrequency probe. We allowed the patient weight-bearing as tolerated starting the day after surgery.

We submitted the specimen for pathologic evaluation (Figure 4). It consisted of multiple pieces of tan/brown tissue. Histologic examination showed benign osteoblastic proliferation composed of anastomosing bony trabeculae with variable mineralization, lined by plump osteoblasts, within vascularized connective tissue; benign giant cells were present, consistent with a nidus of an osteoid osteoma.

On the first postoperative visit, the patient was pain-free and bearing weight with crutches. He was gradually weaned from his crutches and returned to full weight-bearing over the next 4 weeks. At 12-month follow-up, he was symptom-free with good range of motion and full return to previous level of activity.

Discussion

Osteoid osteoma is a small, benign, well-circumscribed osteoblastic cortical lesion, typically identified in long bones or, less frequently, in the subperiosteal region.¹ It often affects adolescents. Osteoid osteoma has been described in the talus in a few case series^2-7 and is associated with a typical nidus that can be identified on CT scans. It does not present acutely, however. The typical presentation for osteoid osteoma is bone pain at night that responds to nonsteroidal anti-inflammatory drugs. However, this presentation is not universal and is frequently missed.²

Juxta-articular osteoid osteomas in the ankle and foot can be difficult to diagnose. The most common site is the talus.³ The majority of patients link their pain to a remote ankle injury. The time delay to diagnosis is on average 2.5 years, but it can be as long as 10 years.^4-6 A CT scan is the best method to identify the nidus; MRI can be misleading if it shows only marrow edema but not a nidus.^4,5,7 In our patient, an injury was documented, and the patient denied prior symptoms. We cannot explain how an injury would trigger the formation of an osteoid osteoma or cause a previously asymptomatic osteoid osteoma to become symptomatic.

Medical treatment with nonsteroidal anti-inflammatory drugs has been used but is reported to take 2 to 4 years for resolution of symptoms; many patients may consider the treatment time frame too long when other alternatives are available.⁸ These include open resection, arthroscopic resection, and image-guided ablation. Open surgical techniques include en bloc resection and curettage. Bone grafting or internal fixation may be performed as needed. Arthroscopic excision of juxta-articular osteoid osteomas offers the advantages of good visualization and avoidance of soft-tissue dissection, and allows for complete excision of the lesion as well as synovectomy.^6,9,10 Arthroscopic excision also allows for quicker rehabilitation. Image-guided ablation, such as radionuclide-guided excision, CT-guided thermal ablation, and laser photocoagulation, may be even less invasive but do not allow for direct visualization, complete resection, and biopsy.¹¹

Conclusion

Osteoid osteoma is a small, benign, well-circumscribed osteoblastic cortical lesion, typically identified in long bones or, less frequently, in the subperiosteal region.¹ It often affects adolescents. Osteoid osteoma has been described in the talus in multiple case series and is associated with a typical nidus that can be identified on CT scans. Usually, it does not present acutely. The typical presentation for osteoid osteoma is bone pain at night that responds to nonsteroidal anti-inflammatory drugs. This presentation is not universal, however, and is frequently missed, especially when the pain is associated with a prior injury.² Arthroscopic exploration of the ankle with resection of subperiosteal osteoid osteoma and the associated synovitis using thermal ablation of the base with radiofrequency offers lasting cure with minimal morbidity.

Osteoid osteoma of the talar neck is an unusual clinical condition that is often overlooked on initial assessment of patients with ankle pain. Here, we present a case report of an adolescent male with talar neck osteoid osteoma who reported persistent pain after an injury. We discuss the differential diagnosis of persistent anterior ankle pain and assess the treatment options for osteoid osteoma of the talar neck. The patient’s guardian provided written informed consent for print and electronic publication of this case report. 

Case Report

A 13-year-old boy presented to our clinic 3 months after a right ankle sprain. He had visited the emergency department at the time of injury; radiographs of the ankle were reported negative for fractures, dislocations, or bone pathologies. He was treated conservatively with elastic support, icing, rest, elevation, and weight-bearing as tolerated. Upon presentation to our office, his pain involved the entire ankle joint. He had not put weight on it since the injury. On examination, he had a significant limp, anteromedial swelling, and tenderness over the ankle joint anteromedially. His neurologic and vascular examinations were normal.

His plain radiographs showed a cystic mass, located at the dorsal aspect of the talar neck (Figures 1A, 1B). Computed tomography (CT) showed a round lucent lesion involving the superior aspect of the talar neck, measuring 9 mm by 6 mm. A sclerotic radiodense focus was evident in the center (Figures 2A, 2B). Noncontrast multiplanar, multisequence magnetic resonance imaging (MRI) showed abnormal edema throughout the talus and a 9-mm rounded ossicle overlying the superior margin of the neck of the talus (Figures 3A, 3B).

Differential Diagnosis

The differential diagnosis for anterior ankle pain includes ankle sprain, monoarticular arthritis, anterior ankle impingement, and talar neck fractures. Other related findings include the presence of a talar ridge and a talar beak.

Ankle sprains are very common injuries. The mainstay treatment consists of ice, resting, elevation, and elastic or semirigid support, and patients usually recover over the course of a few weeks. These sprains are typically injuries of the lateral or medial ligaments of the ankle. Extension of a ligament tear across the anterior capsule can explain persistent anterior ankle pain. The presence of a bony lesion on plain radiographs, however, makes the diagnosis of an ankle sprain, with or without extension into the anterior capsule, less likely.

Monoarticular arthritis, which may present in the ankle and has a wide differential diagnosis, usually involves the whole joint.

Anterior ankle impingement typically occurs in athletes who participate in sports that involve kicking. It can be a bony or soft-tissue impingent. Clinically, patients present with pain and loss of motion, specifically dorsiflexion.

Talar neck fractures are usually the result of high-energy trauma. Stress fractures of the neck of the talus are uncommon and are associated with a recent sudden increase in physical activity, such as running, dancing, or military training. Radiographs, CT scans, and MRI help define the fracture line.

The talar ridge is the site of capsular and ligamentous attachment on the superior aspect of the talar neck and may become hypertrophic in athletes. A hypertrophic talar ridge is asymptomatic and is not considered a pathologic finding on radiographs.

The talar beak, a flaring of the anterosuperior aspect of the talar head, is an indirect sign of tarsal coalition. When symptomatic, patients complain of subtalar symptoms, typically pain and limitation of motion. It usually does not present acutely.

Treatment

We offered the patient surgical excision, and his guardian consented to left ankle arthroscopy. We performed synovectomy using a combination of 3.5-mm shaver and radiofrequency probe. We identified the mass: round, soft, and located at the superior-medial aspect of the talar neck. We removed it in piecemeal fashion using manual arthroscopic instruments, and cauterized its base using the radiofrequency probe. We allowed the patient weight-bearing as tolerated starting the day after surgery.

We submitted the specimen for pathologic evaluation (Figure 4). It consisted of multiple pieces of tan/brown tissue. Histologic examination showed benign osteoblastic proliferation composed of anastomosing bony trabeculae with variable mineralization, lined by plump osteoblasts, within vascularized connective tissue; benign giant cells were present, consistent with a nidus of an osteoid osteoma.

On the first postoperative visit, the patient was pain-free and bearing weight with crutches. He was gradually weaned from his crutches and returned to full weight-bearing over the next 4 weeks. At 12-month follow-up, he was symptom-free with good range of motion and full return to previous level of activity.

Discussion

Osteoid osteoma is a small, benign, well-circumscribed osteoblastic cortical lesion, typically identified in long bones or, less frequently, in the subperiosteal region.¹ It often affects adolescents. Osteoid osteoma has been described in the talus in a few case series^2-7 and is associated with a typical nidus that can be identified on CT scans. It does not present acutely, however. The typical presentation for osteoid osteoma is bone pain at night that responds to nonsteroidal anti-inflammatory drugs. However, this presentation is not universal and is frequently missed.²

Juxta-articular osteoid osteomas in the ankle and foot can be difficult to diagnose. The most common site is the talus.³ The majority of patients link their pain to a remote ankle injury. The time delay to diagnosis is on average 2.5 years, but it can be as long as 10 years.^4-6 A CT scan is the best method to identify the nidus; MRI can be misleading if it shows only marrow edema but not a nidus.^4,5,7 In our patient, an injury was documented, and the patient denied prior symptoms. We cannot explain how an injury would trigger the formation of an osteoid osteoma or cause a previously asymptomatic osteoid osteoma to become symptomatic.

Medical treatment with nonsteroidal anti-inflammatory drugs has been used but is reported to take 2 to 4 years for resolution of symptoms; many patients may consider the treatment time frame too long when other alternatives are available.⁸ These include open resection, arthroscopic resection, and image-guided ablation. Open surgical techniques include en bloc resection and curettage. Bone grafting or internal fixation may be performed as needed. Arthroscopic excision of juxta-articular osteoid osteomas offers the advantages of good visualization and avoidance of soft-tissue dissection, and allows for complete excision of the lesion as well as synovectomy.^6,9,10 Arthroscopic excision also allows for quicker rehabilitation. Image-guided ablation, such as radionuclide-guided excision, CT-guided thermal ablation, and laser photocoagulation, may be even less invasive but do not allow for direct visualization, complete resection, and biopsy.¹¹

Conclusion

Osteoid osteoma is a small, benign, well-circumscribed osteoblastic cortical lesion, typically identified in long bones or, less frequently, in the subperiosteal region.¹ It often affects adolescents. Osteoid osteoma has been described in the talus in multiple case series and is associated with a typical nidus that can be identified on CT scans. Usually, it does not present acutely. The typical presentation for osteoid osteoma is bone pain at night that responds to nonsteroidal anti-inflammatory drugs. This presentation is not universal, however, and is frequently missed, especially when the pain is associated with a prior injury.² Arthroscopic exploration of the ankle with resection of subperiosteal osteoid osteoma and the associated synovitis using thermal ablation of the base with radiofrequency offers lasting cure with minimal morbidity.