Case Reports

Subtrochanteric fractures and other complications related to hardware removal in children with slipped capital femoral epiphysis (SCFE) have been well documented.^1-3 Subtrochanteric fractures after cannulated screw fixation of femoral neck fractures in adults have also been well recognized,⁴ and there are several reports on the topic.^4,5 However, there are no reports on subtrochanteric fractures after removal of the screws for femoral neck fractures in children.

In this article, we report the case of a child who sustained a subtrochanteric fracture after the screw removal and healing that followed a femoral neck fracture. The patient’s parent provided written informed consent for print and electronic publication of this case report. In addition, our institutional review board approved this case report.

Case Report

A 10-year-old boy was brought to our emergency department with the chief complaint of left hip pain after a car accident. Anteroposterior and axial lateral radiographs showed a displaced cervicotrochanteric femoral neck fracture (Figures 1A, 1B). The patient was admitted to the hospital and underwent closed reduction and internal fixation with two 3.5-mm cannulated titanium screws within 12 hours of arrival. The screws did not cross the physis to avoid iatrogenic injury of the capital femoral epiphysis (Figures 2A, 2B). The entry point was located at the lower level of the lesser trochanter. The lateral cortex was penetrated only once by the guide wire for the placement of each screw.

The patient was discharged to home care with a crutch and an ischial weight-bearing long leg brace for protection from unexpected external force. Two months after surgery, we allowed the patient to walk with the brace and without the crutch. Full-weight-bearing ambulation was allowed 3 months after surgery.

About 9 months after initial surgery, we removed 2 titanium screws, which were completely covered with growing new bone. The lateral cortex surrounding the screw heads was chiseled from the lower level of the lesser trochanter to remove the completely immersed screw heads (Figures 3A, 3B).

After screw removal, we recommended non-weight-bearing crutch-walking for 2 weeks followed by partial weight-bearing with crutch for another month. However, the patient started full weight-bearing 2 weeks after screw removal. One month after screw removal, he was brought to the emergency department with severe left hip pain after missing a step on a path. Anteroposterior and lateral radiographs showed an oblique subtrochanteric fracture at the empty screw holes (Figures 4A, 4B). A plate and 4 screws were placed to stabilize the subtrochanteric fracture, and a hip spica cast was applied and was to be worn for 3 weeks (Figures 5A, 5B).

At final follow-up, 6 months after the second surgery, the fracture was healed, and there had been no complications, such as avascular necrosis of the femoral head and leg-length discrepancy (Figures 6A, 6B).

Discussion

Although in situ pinning of SCFE is a common procedure with good results, the rate of complications of hardware removal can be as high as 34%; these complications are well documented.⁵ Subtrochanteric fracture as a complication of proximal femoral neck pinning in adults is also well documented.^4,5 However, there are no reports on subtrochanteric fractures after screw removal in the treatment of femoral neck fractures in children.

Brooks and colleagues⁶ emphasized the point that multiple passes weakened the lateral cortex, decreased the energy-absorbing capacity by 55.2%, and increased local stress. Even if a screw is placed in a relatively safe zone above the lesser trochanter, pie-crusting of the cortex can weaken it enough to predispose it to failure under a relatively normal load.⁷ We inserted 2 cannulated titanium screws without repositioning or multiple drilling, and the femoral neck fracture was united.

The common denominator for subtrochanteric fractures after screw or pin fixation of femoral neck fractures in adults seemed to be the entry point of the lateral cortex below the level of the most inferior edge of the lesser trochanter.⁴ The pin should have its entry site proximal to the level of the lesser trochnater. Paloski and colleagues⁷ and Canale and colleagues⁸ hypothesized that this screw acted as a stress riser to the normal bone, which underwent abnormal loads caused by the patient’s habitus and later mechanism of injury. In our patient’s case, the appropriate starting point for perpendicular penetration of the femoral neck fracture line was on the lateral femoral cortex at the level of the lesser trochanter. We thought this entry on the lateral cortex might predispose the patient to a subtrochanteric fracture. The starting point of the screw is considered the most important factor in preventing fracture after screw removal.

 

As titanium pins cause very tight bone ingrowth,^9,10 the surface of titanium screws used for femoral neck fractures in children are smoothed to reduce turning force.¹ The hexagonal sockets wore off rapidly and proved to be too weak to overcome the necessary torque for loosening the pin from the bone.

Lee and colleagues¹⁰ found that significantly more operative time was needed to remove titanium pins (vs steel pins) after 12 months or longer. When Asnis III pins (Howmedica, Rutherford, New Jersey) were used in the treatment of femoral neck fractures in aged patients, similar problems did not occur. One possible explanation is that bone density is higher in adolescents than in adults. In addition, more bone ingrowth and higher bone compression might occur in adolescent bones.¹ Given the considerable disadvantages noted in their series, Ilchmann and Parsch¹ concluded that use of cannulated titanium screws should be suspended and that stainless steel pins are safe to use in SCFE.

In our patient’s case, we also struggled to remove titanium screws. Subtrochanteric fractures can be complications after removal of screws for femoral neck fractures in children. If there are no specific screw-related symptoms, one should consider leaving the screw in place and avoiding screw removal.

Subtrochanteric fractures and other complications related to hardware removal in children with slipped capital femoral epiphysis (SCFE) have been well documented.^1-3 Subtrochanteric fractures after cannulated screw fixation of femoral neck fractures in adults have also been well recognized,⁴ and there are several reports on the topic.^4,5 However, there are no reports on subtrochanteric fractures after removal of the screws for femoral neck fractures in children.

In this article, we report the case of a child who sustained a subtrochanteric fracture after the screw removal and healing that followed a femoral neck fracture. The patient’s parent provided written informed consent for print and electronic publication of this case report. In addition, our institutional review board approved this case report.

Case Report

A 10-year-old boy was brought to our emergency department with the chief complaint of left hip pain after a car accident. Anteroposterior and axial lateral radiographs showed a displaced cervicotrochanteric femoral neck fracture (Figures 1A, 1B). The patient was admitted to the hospital and underwent closed reduction and internal fixation with two 3.5-mm cannulated titanium screws within 12 hours of arrival. The screws did not cross the physis to avoid iatrogenic injury of the capital femoral epiphysis (Figures 2A, 2B). The entry point was located at the lower level of the lesser trochanter. The lateral cortex was penetrated only once by the guide wire for the placement of each screw.

The patient was discharged to home care with a crutch and an ischial weight-bearing long leg brace for protection from unexpected external force. Two months after surgery, we allowed the patient to walk with the brace and without the crutch. Full-weight-bearing ambulation was allowed 3 months after surgery.

About 9 months after initial surgery, we removed 2 titanium screws, which were completely covered with growing new bone. The lateral cortex surrounding the screw heads was chiseled from the lower level of the lesser trochanter to remove the completely immersed screw heads (Figures 3A, 3B).

After screw removal, we recommended non-weight-bearing crutch-walking for 2 weeks followed by partial weight-bearing with crutch for another month. However, the patient started full weight-bearing 2 weeks after screw removal. One month after screw removal, he was brought to the emergency department with severe left hip pain after missing a step on a path. Anteroposterior and lateral radiographs showed an oblique subtrochanteric fracture at the empty screw holes (Figures 4A, 4B). A plate and 4 screws were placed to stabilize the subtrochanteric fracture, and a hip spica cast was applied and was to be worn for 3 weeks (Figures 5A, 5B).

At final follow-up, 6 months after the second surgery, the fracture was healed, and there had been no complications, such as avascular necrosis of the femoral head and leg-length discrepancy (Figures 6A, 6B).

Discussion

Although in situ pinning of SCFE is a common procedure with good results, the rate of complications of hardware removal can be as high as 34%; these complications are well documented.⁵ Subtrochanteric fracture as a complication of proximal femoral neck pinning in adults is also well documented.^4,5 However, there are no reports on subtrochanteric fractures after screw removal in the treatment of femoral neck fractures in children.

Brooks and colleagues⁶ emphasized the point that multiple passes weakened the lateral cortex, decreased the energy-absorbing capacity by 55.2%, and increased local stress. Even if a screw is placed in a relatively safe zone above the lesser trochanter, pie-crusting of the cortex can weaken it enough to predispose it to failure under a relatively normal load.⁷ We inserted 2 cannulated titanium screws without repositioning or multiple drilling, and the femoral neck fracture was united.

The common denominator for subtrochanteric fractures after screw or pin fixation of femoral neck fractures in adults seemed to be the entry point of the lateral cortex below the level of the most inferior edge of the lesser trochanter.⁴ The pin should have its entry site proximal to the level of the lesser trochnater. Paloski and colleagues⁷ and Canale and colleagues⁸ hypothesized that this screw acted as a stress riser to the normal bone, which underwent abnormal loads caused by the patient’s habitus and later mechanism of injury. In our patient’s case, the appropriate starting point for perpendicular penetration of the femoral neck fracture line was on the lateral femoral cortex at the level of the lesser trochanter. We thought this entry on the lateral cortex might predispose the patient to a subtrochanteric fracture. The starting point of the screw is considered the most important factor in preventing fracture after screw removal.

 

As titanium pins cause very tight bone ingrowth,^9,10 the surface of titanium screws used for femoral neck fractures in children are smoothed to reduce turning force.¹ The hexagonal sockets wore off rapidly and proved to be too weak to overcome the necessary torque for loosening the pin from the bone.

Lee and colleagues¹⁰ found that significantly more operative time was needed to remove titanium pins (vs steel pins) after 12 months or longer. When Asnis III pins (Howmedica, Rutherford, New Jersey) were used in the treatment of femoral neck fractures in aged patients, similar problems did not occur. One possible explanation is that bone density is higher in adolescents than in adults. In addition, more bone ingrowth and higher bone compression might occur in adolescent bones.¹ Given the considerable disadvantages noted in their series, Ilchmann and Parsch¹ concluded that use of cannulated titanium screws should be suspended and that stainless steel pins are safe to use in SCFE.

In our patient’s case, we also struggled to remove titanium screws. Subtrochanteric fractures can be complications after removal of screws for femoral neck fractures in children. If there are no specific screw-related symptoms, one should consider leaving the screw in place and avoiding screw removal.

Subtrochanteric fractures and other complications related to hardware removal in children with slipped capital femoral epiphysis (SCFE) have been well documented.^1-3 Subtrochanteric fractures after cannulated screw fixation of femoral neck fractures in adults have also been well recognized,⁴ and there are several reports on the topic.^4,5 However, there are no reports on subtrochanteric fractures after removal of the screws for femoral neck fractures in children.

In this article, we report the case of a child who sustained a subtrochanteric fracture after the screw removal and healing that followed a femoral neck fracture. The patient’s parent provided written informed consent for print and electronic publication of this case report. In addition, our institutional review board approved this case report.

Case Report

A 10-year-old boy was brought to our emergency department with the chief complaint of left hip pain after a car accident. Anteroposterior and axial lateral radiographs showed a displaced cervicotrochanteric femoral neck fracture (Figures 1A, 1B). The patient was admitted to the hospital and underwent closed reduction and internal fixation with two 3.5-mm cannulated titanium screws within 12 hours of arrival. The screws did not cross the physis to avoid iatrogenic injury of the capital femoral epiphysis (Figures 2A, 2B). The entry point was located at the lower level of the lesser trochanter. The lateral cortex was penetrated only once by the guide wire for the placement of each screw.

The patient was discharged to home care with a crutch and an ischial weight-bearing long leg brace for protection from unexpected external force. Two months after surgery, we allowed the patient to walk with the brace and without the crutch. Full-weight-bearing ambulation was allowed 3 months after surgery.

About 9 months after initial surgery, we removed 2 titanium screws, which were completely covered with growing new bone. The lateral cortex surrounding the screw heads was chiseled from the lower level of the lesser trochanter to remove the completely immersed screw heads (Figures 3A, 3B).

After screw removal, we recommended non-weight-bearing crutch-walking for 2 weeks followed by partial weight-bearing with crutch for another month. However, the patient started full weight-bearing 2 weeks after screw removal. One month after screw removal, he was brought to the emergency department with severe left hip pain after missing a step on a path. Anteroposterior and lateral radiographs showed an oblique subtrochanteric fracture at the empty screw holes (Figures 4A, 4B). A plate and 4 screws were placed to stabilize the subtrochanteric fracture, and a hip spica cast was applied and was to be worn for 3 weeks (Figures 5A, 5B).

At final follow-up, 6 months after the second surgery, the fracture was healed, and there had been no complications, such as avascular necrosis of the femoral head and leg-length discrepancy (Figures 6A, 6B).

Discussion

Although in situ pinning of SCFE is a common procedure with good results, the rate of complications of hardware removal can be as high as 34%; these complications are well documented.⁵ Subtrochanteric fracture as a complication of proximal femoral neck pinning in adults is also well documented.^4,5 However, there are no reports on subtrochanteric fractures after screw removal in the treatment of femoral neck fractures in children.

Brooks and colleagues⁶ emphasized the point that multiple passes weakened the lateral cortex, decreased the energy-absorbing capacity by 55.2%, and increased local stress. Even if a screw is placed in a relatively safe zone above the lesser trochanter, pie-crusting of the cortex can weaken it enough to predispose it to failure under a relatively normal load.⁷ We inserted 2 cannulated titanium screws without repositioning or multiple drilling, and the femoral neck fracture was united.

The common denominator for subtrochanteric fractures after screw or pin fixation of femoral neck fractures in adults seemed to be the entry point of the lateral cortex below the level of the most inferior edge of the lesser trochanter.⁴ The pin should have its entry site proximal to the level of the lesser trochnater. Paloski and colleagues⁷ and Canale and colleagues⁸ hypothesized that this screw acted as a stress riser to the normal bone, which underwent abnormal loads caused by the patient’s habitus and later mechanism of injury. In our patient’s case, the appropriate starting point for perpendicular penetration of the femoral neck fracture line was on the lateral femoral cortex at the level of the lesser trochanter. We thought this entry on the lateral cortex might predispose the patient to a subtrochanteric fracture. The starting point of the screw is considered the most important factor in preventing fracture after screw removal.

 

As titanium pins cause very tight bone ingrowth,^9,10 the surface of titanium screws used for femoral neck fractures in children are smoothed to reduce turning force.¹ The hexagonal sockets wore off rapidly and proved to be too weak to overcome the necessary torque for loosening the pin from the bone.

Lee and colleagues¹⁰ found that significantly more operative time was needed to remove titanium pins (vs steel pins) after 12 months or longer. When Asnis III pins (Howmedica, Rutherford, New Jersey) were used in the treatment of femoral neck fractures in aged patients, similar problems did not occur. One possible explanation is that bone density is higher in adolescents than in adults. In addition, more bone ingrowth and higher bone compression might occur in adolescent bones.¹ Given the considerable disadvantages noted in their series, Ilchmann and Parsch¹ concluded that use of cannulated titanium screws should be suspended and that stainless steel pins are safe to use in SCFE.

In our patient’s case, we also struggled to remove titanium screws. Subtrochanteric fractures can be complications after removal of screws for femoral neck fractures in children. If there are no specific screw-related symptoms, one should consider leaving the screw in place and avoiding screw removal.

A 32-year-old G1P0 woman at 7 4/7 weeks’ gestation (intrauterine pregnancy confirmed by ultrasound) with a history of hypertension and anxiety presented to the emergency department reporting right upper quadrant and right flank pain. The patient did not report taking any medications. Her symptoms had begun about 6 hours prior to presentation. She did not report fever, chills, nausea, vomiting, anorexia, or urinary or bowels symptoms. She was afebrile, and her initial blood pressure was 174/84 mm Hg, but within an hour of presentation had decreased to 137/82 mm Hg.

On exam she was in moderate-to-severe discomfort. She was tender to palpation in the right flank as well as the right upper quadrant and had a positive Murphy sign. Her white blood cell (WBC) count was 18,800/μL, hemoglobin and hematocrit were normal, liver function tests were within normal limits, and sodium was 133 mmol/L; otherwise, electrolytes were normal. The urinalysis showed a specific gravity of 1.022 with 3 red blood cells per high powered field, but was otherwise normal. A right upper quadrant ultrasound showed a normal gallbladder without cholelithiasis. Subsequently, a magnetic resonance imaging (MRI) of the abdomen was obtained (Figure 1).

 

• What is your diagnosis?
• How would you treat this patient? 

[Click through to the next page to see the answer.]

 

Our Treatment

An MRI of the abdomen revealed a right adrenal hematoma (Figure 2). This diagnosis was made after common causes of the right upper quadrant and right flank pain were excluded and was confirmed with an MRI of the abdomen.

Discussion

Common causes of right flank and right upper quadrant pain include urolithiasis, pyelonephritis, and acute cholecystitis (Table). Patients with urolithiasis will typically present with sudden onset of colicky flank pain, which may radiate to the inguinal region, with or without nausea and vomiting.^1,2 Microhematuria may be absent in as many as 10% to 20% of patients.² Computed tomography is nearly 97% sensitive and 96% specific for urolithiasis.² Although only 37% to 64% sensitive in detecting urolithiasis, abdominal ultrasound is 85% to 94% specific for detecting hydronephrosis.²

Symptoms of pyelonephritis range from mild to severe costovertebral angle pain with or without fever and/or lower urinary tract symptoms to sepsis.³ Urinalysis typically shows leukocyte esterase (72%-97% sensitive, 41%-86% specific for culture-confirmed urinary tract infection) and microscopic pyuria (90%-96% sensitive, 47%-50% specific for culture-confirmed urinary tract infection).^3,4 Urine culture will reveal 10⁵ colony-forming units in 95% of patients with acute uncomplicated pyelonephritis.³ Imaging is typically reserved for atypical presentations or in cases of nonresolution of symptoms despite treatment in order to rule out structural anomalies, an obstructive process, or abscess formation.³

Acute cholecystitis is suggested by right upper quadrant pain, often following a meal. Symptoms include fever, leukocytosis, and an elevated serum bilirubin level.⁵ Ultrasound findings of acute cholecystitis are gallbladder wall thickening, pericholecystic fluid, or sonographic Murphy sign.⁶ Ultrasound has a sensitivity of 81% and a specificity of 83% for detecting acute cholecystitis, whereas hepatobiliary iminodiacetic acid scan (cholescintigraphy) is 96% sensitive and 90% specific and is used in cases where ultrasound is indeterminate.⁶

Spontaneous adrenal hemorrhage (SAH) typically presents with symptoms of upper abdominal pain and/or flank pain. If hemorrhage is significant, hypotension and a drop in hematocrit may occur. Alternatively, if SAH is bilateral and results in a deficiency of adrenal hormones, hypotension may be secondary to adrenal insufficiency. Due to its ready availability, most SAHs are diagnosed by CT scan ordered for alternate reasons.⁷

In the pregnant patient, to avoid exposing the developing fetus to radiation, ultrasound is typically the first imaging modality to investigate right upper quadrant or flank pain. However, ultrasound is less sensitive for evaluating anomalies of the adrenal gland.⁸ An MRI is the most accurate imaging modality for diagnosing adrenal hemorrhage and avoids exposing the developing fetus to radiation.⁹ In this case, the absence of an etiology of the patient’s symptoms on right upper quadrant ultrasound led to further evaluation, with abdominal MRI revealing the diagnosis.

In a review of 141 cases of adrenal hemorrhage at the Mayo Clinic in Rochester, Minnesota, 16 patients had a spontaneous adrenal hemorrhage, which presented with severe abdominal and flank pain of sudden onset.⁷ Seven of these patients required surgery to control the bleeding.⁷ None of the 13 patients with a functioning contralateral adrenal gland required adrenal replacement therapy.⁷ Gavrilova-Jordan and colleagues reported a case of a spontaneous unilateral adrenal hemorrhage in the third trimester of pregnancy that was managed conservatively with favorable outcomes.¹⁰ Potential causes of spontaneous adrenal hemorrhage include:

   • Antiphospholipid antibody syndrome

   • Heparin-induced thrombocytopenia

   • Trauma

   • Physiologic stress (eg, sepsis)

   • Adrenal mass (pheochromocytoma, angiomyolipoma, or metastatic cancer).^6,11

Case Outcome

Initial right upper quadrant ultrasound in this patient revealed a normal gallbladder and a small amount of perinephric fluid surrounding the right kidney. On the night of admission she experienced an elevated temperature of 101.9°F. Due to the fever and leukocytosis, she was initiated on antibiotics (ampicillin-sulbactam 3 g IV every 6 hours) covering a urinary source. On hospital day 2, right lower quadrant and right upper quadrant ultrasounds  were performed, revealing decreased fluid surrounding her right kidney and without evidence of appendicitis. Due to persistent leukocytosis (WBC count peaked at 26,400/μL), essentially normal urinalysis, and negative urine culture, an abdominal MRI was obtained to evaluate for a perinephric abscess. The MRI revealed the spontaneous adrenal hemorrhage.

The patient was observed as an inpatient for the following 48 hours. Her electrolytes remained normal and hemoglobin decreased to 9.5 g/dL. She remained hemodynamically stable. Serum cortisol and angiotensin levels were normal, antibiotics were discontinued, and the patient was discharged to outpatient follow-up.

A repeat MRI of the abdomen obtained 6 weeks later revealed a resolving right adrenal hematoma. At 39 weeks’ gestation, she delivered a vigorous female infant weighing 3,200 g (7 lb 1 oz) with an Apgar score at birth of 8/9.

Author disclosures
The authors report no actual or potential conflicts of interest with regard to this article.

Disclaimer
The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the U.S. Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.

A 32-year-old G1P0 woman at 7 4/7 weeks’ gestation (intrauterine pregnancy confirmed by ultrasound) with a history of hypertension and anxiety presented to the emergency department reporting right upper quadrant and right flank pain. The patient did not report taking any medications. Her symptoms had begun about 6 hours prior to presentation. She did not report fever, chills, nausea, vomiting, anorexia, or urinary or bowels symptoms. She was afebrile, and her initial blood pressure was 174/84 mm Hg, but within an hour of presentation had decreased to 137/82 mm Hg.

On exam she was in moderate-to-severe discomfort. She was tender to palpation in the right flank as well as the right upper quadrant and had a positive Murphy sign. Her white blood cell (WBC) count was 18,800/μL, hemoglobin and hematocrit were normal, liver function tests were within normal limits, and sodium was 133 mmol/L; otherwise, electrolytes were normal. The urinalysis showed a specific gravity of 1.022 with 3 red blood cells per high powered field, but was otherwise normal. A right upper quadrant ultrasound showed a normal gallbladder without cholelithiasis. Subsequently, a magnetic resonance imaging (MRI) of the abdomen was obtained (Figure 1).

 

• What is your diagnosis?
• How would you treat this patient? 

[Click through to the next page to see the answer.]

 

Our Treatment

An MRI of the abdomen revealed a right adrenal hematoma (Figure 2). This diagnosis was made after common causes of the right upper quadrant and right flank pain were excluded and was confirmed with an MRI of the abdomen.

Discussion

Common causes of right flank and right upper quadrant pain include urolithiasis, pyelonephritis, and acute cholecystitis (Table). Patients with urolithiasis will typically present with sudden onset of colicky flank pain, which may radiate to the inguinal region, with or without nausea and vomiting.^1,2 Microhematuria may be absent in as many as 10% to 20% of patients.² Computed tomography is nearly 97% sensitive and 96% specific for urolithiasis.² Although only 37% to 64% sensitive in detecting urolithiasis, abdominal ultrasound is 85% to 94% specific for detecting hydronephrosis.²

Symptoms of pyelonephritis range from mild to severe costovertebral angle pain with or without fever and/or lower urinary tract symptoms to sepsis.³ Urinalysis typically shows leukocyte esterase (72%-97% sensitive, 41%-86% specific for culture-confirmed urinary tract infection) and microscopic pyuria (90%-96% sensitive, 47%-50% specific for culture-confirmed urinary tract infection).^3,4 Urine culture will reveal 10⁵ colony-forming units in 95% of patients with acute uncomplicated pyelonephritis.³ Imaging is typically reserved for atypical presentations or in cases of nonresolution of symptoms despite treatment in order to rule out structural anomalies, an obstructive process, or abscess formation.³

Acute cholecystitis is suggested by right upper quadrant pain, often following a meal. Symptoms include fever, leukocytosis, and an elevated serum bilirubin level.⁵ Ultrasound findings of acute cholecystitis are gallbladder wall thickening, pericholecystic fluid, or sonographic Murphy sign.⁶ Ultrasound has a sensitivity of 81% and a specificity of 83% for detecting acute cholecystitis, whereas hepatobiliary iminodiacetic acid scan (cholescintigraphy) is 96% sensitive and 90% specific and is used in cases where ultrasound is indeterminate.⁶

Spontaneous adrenal hemorrhage (SAH) typically presents with symptoms of upper abdominal pain and/or flank pain. If hemorrhage is significant, hypotension and a drop in hematocrit may occur. Alternatively, if SAH is bilateral and results in a deficiency of adrenal hormones, hypotension may be secondary to adrenal insufficiency. Due to its ready availability, most SAHs are diagnosed by CT scan ordered for alternate reasons.⁷

In the pregnant patient, to avoid exposing the developing fetus to radiation, ultrasound is typically the first imaging modality to investigate right upper quadrant or flank pain. However, ultrasound is less sensitive for evaluating anomalies of the adrenal gland.⁸ An MRI is the most accurate imaging modality for diagnosing adrenal hemorrhage and avoids exposing the developing fetus to radiation.⁹ In this case, the absence of an etiology of the patient’s symptoms on right upper quadrant ultrasound led to further evaluation, with abdominal MRI revealing the diagnosis.

In a review of 141 cases of adrenal hemorrhage at the Mayo Clinic in Rochester, Minnesota, 16 patients had a spontaneous adrenal hemorrhage, which presented with severe abdominal and flank pain of sudden onset.⁷ Seven of these patients required surgery to control the bleeding.⁷ None of the 13 patients with a functioning contralateral adrenal gland required adrenal replacement therapy.⁷ Gavrilova-Jordan and colleagues reported a case of a spontaneous unilateral adrenal hemorrhage in the third trimester of pregnancy that was managed conservatively with favorable outcomes.¹⁰ Potential causes of spontaneous adrenal hemorrhage include:

   • Antiphospholipid antibody syndrome

   • Heparin-induced thrombocytopenia

   • Trauma

   • Physiologic stress (eg, sepsis)

   • Adrenal mass (pheochromocytoma, angiomyolipoma, or metastatic cancer).^6,11

Case Outcome

Initial right upper quadrant ultrasound in this patient revealed a normal gallbladder and a small amount of perinephric fluid surrounding the right kidney. On the night of admission she experienced an elevated temperature of 101.9°F. Due to the fever and leukocytosis, she was initiated on antibiotics (ampicillin-sulbactam 3 g IV every 6 hours) covering a urinary source. On hospital day 2, right lower quadrant and right upper quadrant ultrasounds  were performed, revealing decreased fluid surrounding her right kidney and without evidence of appendicitis. Due to persistent leukocytosis (WBC count peaked at 26,400/μL), essentially normal urinalysis, and negative urine culture, an abdominal MRI was obtained to evaluate for a perinephric abscess. The MRI revealed the spontaneous adrenal hemorrhage.

The patient was observed as an inpatient for the following 48 hours. Her electrolytes remained normal and hemoglobin decreased to 9.5 g/dL. She remained hemodynamically stable. Serum cortisol and angiotensin levels were normal, antibiotics were discontinued, and the patient was discharged to outpatient follow-up.

A repeat MRI of the abdomen obtained 6 weeks later revealed a resolving right adrenal hematoma. At 39 weeks’ gestation, she delivered a vigorous female infant weighing 3,200 g (7 lb 1 oz) with an Apgar score at birth of 8/9.

Author disclosures
The authors report no actual or potential conflicts of interest with regard to this article.

Disclaimer
The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the U.S. Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.

A 32-year-old G1P0 woman at 7 4/7 weeks’ gestation (intrauterine pregnancy confirmed by ultrasound) with a history of hypertension and anxiety presented to the emergency department reporting right upper quadrant and right flank pain. The patient did not report taking any medications. Her symptoms had begun about 6 hours prior to presentation. She did not report fever, chills, nausea, vomiting, anorexia, or urinary or bowels symptoms. She was afebrile, and her initial blood pressure was 174/84 mm Hg, but within an hour of presentation had decreased to 137/82 mm Hg.

On exam she was in moderate-to-severe discomfort. She was tender to palpation in the right flank as well as the right upper quadrant and had a positive Murphy sign. Her white blood cell (WBC) count was 18,800/μL, hemoglobin and hematocrit were normal, liver function tests were within normal limits, and sodium was 133 mmol/L; otherwise, electrolytes were normal. The urinalysis showed a specific gravity of 1.022 with 3 red blood cells per high powered field, but was otherwise normal. A right upper quadrant ultrasound showed a normal gallbladder without cholelithiasis. Subsequently, a magnetic resonance imaging (MRI) of the abdomen was obtained (Figure 1).

 

• What is your diagnosis?
• How would you treat this patient? 

[Click through to the next page to see the answer.]

 

Our Treatment

An MRI of the abdomen revealed a right adrenal hematoma (Figure 2). This diagnosis was made after common causes of the right upper quadrant and right flank pain were excluded and was confirmed with an MRI of the abdomen.

Discussion

Common causes of right flank and right upper quadrant pain include urolithiasis, pyelonephritis, and acute cholecystitis (Table). Patients with urolithiasis will typically present with sudden onset of colicky flank pain, which may radiate to the inguinal region, with or without nausea and vomiting.^1,2 Microhematuria may be absent in as many as 10% to 20% of patients.² Computed tomography is nearly 97% sensitive and 96% specific for urolithiasis.² Although only 37% to 64% sensitive in detecting urolithiasis, abdominal ultrasound is 85% to 94% specific for detecting hydronephrosis.²

Symptoms of pyelonephritis range from mild to severe costovertebral angle pain with or without fever and/or lower urinary tract symptoms to sepsis.³ Urinalysis typically shows leukocyte esterase (72%-97% sensitive, 41%-86% specific for culture-confirmed urinary tract infection) and microscopic pyuria (90%-96% sensitive, 47%-50% specific for culture-confirmed urinary tract infection).^3,4 Urine culture will reveal 10⁵ colony-forming units in 95% of patients with acute uncomplicated pyelonephritis.³ Imaging is typically reserved for atypical presentations or in cases of nonresolution of symptoms despite treatment in order to rule out structural anomalies, an obstructive process, or abscess formation.³

Acute cholecystitis is suggested by right upper quadrant pain, often following a meal. Symptoms include fever, leukocytosis, and an elevated serum bilirubin level.⁵ Ultrasound findings of acute cholecystitis are gallbladder wall thickening, pericholecystic fluid, or sonographic Murphy sign.⁶ Ultrasound has a sensitivity of 81% and a specificity of 83% for detecting acute cholecystitis, whereas hepatobiliary iminodiacetic acid scan (cholescintigraphy) is 96% sensitive and 90% specific and is used in cases where ultrasound is indeterminate.⁶

Spontaneous adrenal hemorrhage (SAH) typically presents with symptoms of upper abdominal pain and/or flank pain. If hemorrhage is significant, hypotension and a drop in hematocrit may occur. Alternatively, if SAH is bilateral and results in a deficiency of adrenal hormones, hypotension may be secondary to adrenal insufficiency. Due to its ready availability, most SAHs are diagnosed by CT scan ordered for alternate reasons.⁷

In the pregnant patient, to avoid exposing the developing fetus to radiation, ultrasound is typically the first imaging modality to investigate right upper quadrant or flank pain. However, ultrasound is less sensitive for evaluating anomalies of the adrenal gland.⁸ An MRI is the most accurate imaging modality for diagnosing adrenal hemorrhage and avoids exposing the developing fetus to radiation.⁹ In this case, the absence of an etiology of the patient’s symptoms on right upper quadrant ultrasound led to further evaluation, with abdominal MRI revealing the diagnosis.

In a review of 141 cases of adrenal hemorrhage at the Mayo Clinic in Rochester, Minnesota, 16 patients had a spontaneous adrenal hemorrhage, which presented with severe abdominal and flank pain of sudden onset.⁷ Seven of these patients required surgery to control the bleeding.⁷ None of the 13 patients with a functioning contralateral adrenal gland required adrenal replacement therapy.⁷ Gavrilova-Jordan and colleagues reported a case of a spontaneous unilateral adrenal hemorrhage in the third trimester of pregnancy that was managed conservatively with favorable outcomes.¹⁰ Potential causes of spontaneous adrenal hemorrhage include:

   • Antiphospholipid antibody syndrome

   • Heparin-induced thrombocytopenia

   • Trauma

   • Physiologic stress (eg, sepsis)

   • Adrenal mass (pheochromocytoma, angiomyolipoma, or metastatic cancer).^6,11

Case Outcome

Initial right upper quadrant ultrasound in this patient revealed a normal gallbladder and a small amount of perinephric fluid surrounding the right kidney. On the night of admission she experienced an elevated temperature of 101.9°F. Due to the fever and leukocytosis, she was initiated on antibiotics (ampicillin-sulbactam 3 g IV every 6 hours) covering a urinary source. On hospital day 2, right lower quadrant and right upper quadrant ultrasounds  were performed, revealing decreased fluid surrounding her right kidney and without evidence of appendicitis. Due to persistent leukocytosis (WBC count peaked at 26,400/μL), essentially normal urinalysis, and negative urine culture, an abdominal MRI was obtained to evaluate for a perinephric abscess. The MRI revealed the spontaneous adrenal hemorrhage.

The patient was observed as an inpatient for the following 48 hours. Her electrolytes remained normal and hemoglobin decreased to 9.5 g/dL. She remained hemodynamically stable. Serum cortisol and angiotensin levels were normal, antibiotics were discontinued, and the patient was discharged to outpatient follow-up.

A repeat MRI of the abdomen obtained 6 weeks later revealed a resolving right adrenal hematoma. At 39 weeks’ gestation, she delivered a vigorous female infant weighing 3,200 g (7 lb 1 oz) with an Apgar score at birth of 8/9.

Author disclosures
The authors report no actual or potential conflicts of interest with regard to this article.

Disclaimer
The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the U.S. Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.

Pulmonary vein thrombosis (PVT) is rare and underdiagnosed in clinical practice. It has been described following lobectomy, lung transplant, and in association with metastatic carcinoma.^1-4 Some cases have been described as idiopathic.^5-7 Its exact incidence is unknown, and treatment depends on etiology. On the other hand, pulmonary artery thromboembolism is a well-known entity with identified risk factors as well as clearly defined modalities of management. The following is a case of PVT, which occurred in the setting of small cell carcinoma of the lung (SCLC) and mantle cell lymphoma of the small intestine.

CASE PRESENTATION

A 66-year-old male veteran with a past medical history of type 2 diabetes mellitus, hypertension, and chronic obstructive pulmonary disease, who had a 40 pack-year history of cigarette smoking, was admitted to the hospital for severe, sudden onset abdominal pain. The pain was localized in the right lower quadrant and then became generalized. It was sharp, aggravated by movement, and relieved by rest. The patient reported being constipated for the past couple of days.

A review of systems revealed that he had been coughing for about 3 days prior to admission. A computed tomography (CT) scan of the abdomen showed pneumoperitoneum and a mass with mural thickening around the distal ileum/cecal area (Figure 1). There was also a partially visualized mass in the infrahilar area of the right lower lobe and bilateral adrenal masses seen on the scan. A chest CT with contrast was then performed, which showed a 7.5 x 6.6 x 6.6-cm mass in the right lower lobe posterior to the right hilum. The mass encased the right mainstem bronchus, and there was a low-density-filling defect in the inferior branch of the right pulmonary vein (Figure 2). An echocardiography did not show any thrombus within the atria or ventricles.

The patient underwent emergent exploratory laparotomy for bowel perforation. The operative finding was a small perforation of the small intestine with an associated mass. There were metastatic lesions throughout the abdomen. A partial small bowel resection was performed. Post exploratory laparotomy, a fiberoptic bronchoscopy was performed, which revealed a 1-cm fungating lesion at the takeoff of the superior segment of the right lower lobe. Brushings were obtained from the mass. The pathology of the lung mass was small cell carcinoma, whereas that of the bowel mass was mantle cell lymphoma. Brain magnetic resonance imaging revealed that he had metastasis to the brain with a 4-cm mass in the cerebellum. He was anticoagulated with heparin for the PVT. Based on his poor functional status and his overall clinical condition, his prognosis was poor. He received hospice care and died 3 months later.

DISCUSSION

Pulmonary vein thrombosis is a rare condition. The incidence is unclear, as most of the literature includes case reports. The majority of PVT cases are reported following lobectomy for malignancy and lung transplantation.^1-3 The incidence following lung transplant was reported in the early postoperative period to be 15% in a center during the first 2 years of the study.³ Pulmonary vein thrombosis has also been described following metastatic cancer, such as liposarcoma.⁴

This patient’s case was discovered in the setting of SCLC and mantle cell lymphoma of the small intestine. Small cell carcinoma of the lung was reported to invade the pulmonary vein into the left atrium.⁸ In this patient, the left atrium was not invaded. There have been cases of spontaneous or idiopathic PVT described, presenting as abdominal pain, hemoptysis, and chest pain.^5-7 No precipitating causes were detected in these patients.

The pathogenesis of PVT from a tumor is unclear, although several theories have been postulated: It could result from direct extension of the tumor into the vein, from compression of the vein by the tumor, or from epithelial damage as a result of tumor invasion. The tumor thrombus has been described to extend into the right atrium.^6,8 The mechanism of thrombosis remains unclear in the patient postlobectomy or postlung transplantation, although intraoperative torsion and injury of vessels are implicated. Similar to deep vein thrombosis, PVT could also result from intimal damage or sluggish flow in the pulmonary stump in the postoperative patient.^2,9,10

The presentation of PVT is usually nonspecific, including dyspnea, cough, pleuritic chest pain, and hemoptysis. It has been reported as causing massive hemoptysis due to acute pulmonary infarction.⁷ Acute PVT occurring postoperatively in the lung transplant patient may be disastrous and lead to early postoperative allograft failure.¹¹ Pulmonary vein thrombosis may also present more insidiously with recurrent pulmonary edema and pulmonary fibrosis.¹² This patient presented with abdominal pain; further workup led to the finding of a lung mass. Pulmonary vein thrombosis has been reported to result in systemic emboli, resulting in cerebrovascular accidents, or it can manifest as aseptic and tumor emboli.^2,5,10,13,14

 

Newer CT techniques have made identifying PVT possible in a similar manner to which pulmonary arterial emboli are detected by using the pulmonary venous phase of a contrast CT of the chest.⁵ Echocardiography may demonstrate the extension of the thrombus into the atrium; a transesophageal echocardiogram would be preferable over a transthoracic echocardiogram. Magnetic resonance imaging of the chest is another useful modality for diagnosis, because it is able to distinguish between a bland thrombus and a tumor thrombus in the pulmonary vein.¹⁵

Treatment of PVT depends on the overall clinical condition of the patient. Irrespective of the etiology, a review of the literature does not indicate the preferred duration of anticoagulation or preference for modality of anticoagulation between oral vitamin K antagonists or heparin—low molecular or unfractionated.^1,3-6 Patients who develop PVT following malignancy are usually anticoagulated with therapy for the cancer. The treatment of PVT in the setting of lung transplant is more challenging and includes systemic heparinization, thrombolytics, and surgical thrombectomy.^3,11,16 The majority of the literature includes case reports with varying morbidity and mortality, depending on the etiology. Ninety-day mortality of 38% was reported following lung transplant.³

CONCLUSION

Pulmonary vein thrombosis presents in a nonspecific manner. The diagnosis is now more readily made with the advent of a variety of diagnostic modalities, especially with transesophageal echocardiography, which may be performed at the bedside in the intensive care unit. The treatment remains challenging with mortality dependent on the etiology. A diagnosis of PVT needs to be considered in patients with appropriate risk factors. A high index of suspicion will enable the diagnosis in the proper clinical scenario.

Author disclosures
The authors report no actual or potential conflicts of interest with regard to this article.

Disclaimer
The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the U.S. Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.

Pulmonary vein thrombosis (PVT) is rare and underdiagnosed in clinical practice. It has been described following lobectomy, lung transplant, and in association with metastatic carcinoma.^1-4 Some cases have been described as idiopathic.^5-7 Its exact incidence is unknown, and treatment depends on etiology. On the other hand, pulmonary artery thromboembolism is a well-known entity with identified risk factors as well as clearly defined modalities of management. The following is a case of PVT, which occurred in the setting of small cell carcinoma of the lung (SCLC) and mantle cell lymphoma of the small intestine.

CASE PRESENTATION

A 66-year-old male veteran with a past medical history of type 2 diabetes mellitus, hypertension, and chronic obstructive pulmonary disease, who had a 40 pack-year history of cigarette smoking, was admitted to the hospital for severe, sudden onset abdominal pain. The pain was localized in the right lower quadrant and then became generalized. It was sharp, aggravated by movement, and relieved by rest. The patient reported being constipated for the past couple of days.

A review of systems revealed that he had been coughing for about 3 days prior to admission. A computed tomography (CT) scan of the abdomen showed pneumoperitoneum and a mass with mural thickening around the distal ileum/cecal area (Figure 1). There was also a partially visualized mass in the infrahilar area of the right lower lobe and bilateral adrenal masses seen on the scan. A chest CT with contrast was then performed, which showed a 7.5 x 6.6 x 6.6-cm mass in the right lower lobe posterior to the right hilum. The mass encased the right mainstem bronchus, and there was a low-density-filling defect in the inferior branch of the right pulmonary vein (Figure 2). An echocardiography did not show any thrombus within the atria or ventricles.

The patient underwent emergent exploratory laparotomy for bowel perforation. The operative finding was a small perforation of the small intestine with an associated mass. There were metastatic lesions throughout the abdomen. A partial small bowel resection was performed. Post exploratory laparotomy, a fiberoptic bronchoscopy was performed, which revealed a 1-cm fungating lesion at the takeoff of the superior segment of the right lower lobe. Brushings were obtained from the mass. The pathology of the lung mass was small cell carcinoma, whereas that of the bowel mass was mantle cell lymphoma. Brain magnetic resonance imaging revealed that he had metastasis to the brain with a 4-cm mass in the cerebellum. He was anticoagulated with heparin for the PVT. Based on his poor functional status and his overall clinical condition, his prognosis was poor. He received hospice care and died 3 months later.

DISCUSSION

Pulmonary vein thrombosis is a rare condition. The incidence is unclear, as most of the literature includes case reports. The majority of PVT cases are reported following lobectomy for malignancy and lung transplantation.^1-3 The incidence following lung transplant was reported in the early postoperative period to be 15% in a center during the first 2 years of the study.³ Pulmonary vein thrombosis has also been described following metastatic cancer, such as liposarcoma.⁴

This patient’s case was discovered in the setting of SCLC and mantle cell lymphoma of the small intestine. Small cell carcinoma of the lung was reported to invade the pulmonary vein into the left atrium.⁸ In this patient, the left atrium was not invaded. There have been cases of spontaneous or idiopathic PVT described, presenting as abdominal pain, hemoptysis, and chest pain.^5-7 No precipitating causes were detected in these patients.

The pathogenesis of PVT from a tumor is unclear, although several theories have been postulated: It could result from direct extension of the tumor into the vein, from compression of the vein by the tumor, or from epithelial damage as a result of tumor invasion. The tumor thrombus has been described to extend into the right atrium.^6,8 The mechanism of thrombosis remains unclear in the patient postlobectomy or postlung transplantation, although intraoperative torsion and injury of vessels are implicated. Similar to deep vein thrombosis, PVT could also result from intimal damage or sluggish flow in the pulmonary stump in the postoperative patient.^2,9,10

The presentation of PVT is usually nonspecific, including dyspnea, cough, pleuritic chest pain, and hemoptysis. It has been reported as causing massive hemoptysis due to acute pulmonary infarction.⁷ Acute PVT occurring postoperatively in the lung transplant patient may be disastrous and lead to early postoperative allograft failure.¹¹ Pulmonary vein thrombosis may also present more insidiously with recurrent pulmonary edema and pulmonary fibrosis.¹² This patient presented with abdominal pain; further workup led to the finding of a lung mass. Pulmonary vein thrombosis has been reported to result in systemic emboli, resulting in cerebrovascular accidents, or it can manifest as aseptic and tumor emboli.^2,5,10,13,14

 

Newer CT techniques have made identifying PVT possible in a similar manner to which pulmonary arterial emboli are detected by using the pulmonary venous phase of a contrast CT of the chest.⁵ Echocardiography may demonstrate the extension of the thrombus into the atrium; a transesophageal echocardiogram would be preferable over a transthoracic echocardiogram. Magnetic resonance imaging of the chest is another useful modality for diagnosis, because it is able to distinguish between a bland thrombus and a tumor thrombus in the pulmonary vein.¹⁵

Treatment of PVT depends on the overall clinical condition of the patient. Irrespective of the etiology, a review of the literature does not indicate the preferred duration of anticoagulation or preference for modality of anticoagulation between oral vitamin K antagonists or heparin—low molecular or unfractionated.^1,3-6 Patients who develop PVT following malignancy are usually anticoagulated with therapy for the cancer. The treatment of PVT in the setting of lung transplant is more challenging and includes systemic heparinization, thrombolytics, and surgical thrombectomy.^3,11,16 The majority of the literature includes case reports with varying morbidity and mortality, depending on the etiology. Ninety-day mortality of 38% was reported following lung transplant.³

CONCLUSION

Pulmonary vein thrombosis presents in a nonspecific manner. The diagnosis is now more readily made with the advent of a variety of diagnostic modalities, especially with transesophageal echocardiography, which may be performed at the bedside in the intensive care unit. The treatment remains challenging with mortality dependent on the etiology. A diagnosis of PVT needs to be considered in patients with appropriate risk factors. A high index of suspicion will enable the diagnosis in the proper clinical scenario.

Author disclosures
The authors report no actual or potential conflicts of interest with regard to this article.

Disclaimer
The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the U.S. Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.

Pulmonary vein thrombosis (PVT) is rare and underdiagnosed in clinical practice. It has been described following lobectomy, lung transplant, and in association with metastatic carcinoma.^1-4 Some cases have been described as idiopathic.^5-7 Its exact incidence is unknown, and treatment depends on etiology. On the other hand, pulmonary artery thromboembolism is a well-known entity with identified risk factors as well as clearly defined modalities of management. The following is a case of PVT, which occurred in the setting of small cell carcinoma of the lung (SCLC) and mantle cell lymphoma of the small intestine.

CASE PRESENTATION

A 66-year-old male veteran with a past medical history of type 2 diabetes mellitus, hypertension, and chronic obstructive pulmonary disease, who had a 40 pack-year history of cigarette smoking, was admitted to the hospital for severe, sudden onset abdominal pain. The pain was localized in the right lower quadrant and then became generalized. It was sharp, aggravated by movement, and relieved by rest. The patient reported being constipated for the past couple of days.

A review of systems revealed that he had been coughing for about 3 days prior to admission. A computed tomography (CT) scan of the abdomen showed pneumoperitoneum and a mass with mural thickening around the distal ileum/cecal area (Figure 1). There was also a partially visualized mass in the infrahilar area of the right lower lobe and bilateral adrenal masses seen on the scan. A chest CT with contrast was then performed, which showed a 7.5 x 6.6 x 6.6-cm mass in the right lower lobe posterior to the right hilum. The mass encased the right mainstem bronchus, and there was a low-density-filling defect in the inferior branch of the right pulmonary vein (Figure 2). An echocardiography did not show any thrombus within the atria or ventricles.

The patient underwent emergent exploratory laparotomy for bowel perforation. The operative finding was a small perforation of the small intestine with an associated mass. There were metastatic lesions throughout the abdomen. A partial small bowel resection was performed. Post exploratory laparotomy, a fiberoptic bronchoscopy was performed, which revealed a 1-cm fungating lesion at the takeoff of the superior segment of the right lower lobe. Brushings were obtained from the mass. The pathology of the lung mass was small cell carcinoma, whereas that of the bowel mass was mantle cell lymphoma. Brain magnetic resonance imaging revealed that he had metastasis to the brain with a 4-cm mass in the cerebellum. He was anticoagulated with heparin for the PVT. Based on his poor functional status and his overall clinical condition, his prognosis was poor. He received hospice care and died 3 months later.

DISCUSSION

Pulmonary vein thrombosis is a rare condition. The incidence is unclear, as most of the literature includes case reports. The majority of PVT cases are reported following lobectomy for malignancy and lung transplantation.^1-3 The incidence following lung transplant was reported in the early postoperative period to be 15% in a center during the first 2 years of the study.³ Pulmonary vein thrombosis has also been described following metastatic cancer, such as liposarcoma.⁴

This patient’s case was discovered in the setting of SCLC and mantle cell lymphoma of the small intestine. Small cell carcinoma of the lung was reported to invade the pulmonary vein into the left atrium.⁸ In this patient, the left atrium was not invaded. There have been cases of spontaneous or idiopathic PVT described, presenting as abdominal pain, hemoptysis, and chest pain.^5-7 No precipitating causes were detected in these patients.

The pathogenesis of PVT from a tumor is unclear, although several theories have been postulated: It could result from direct extension of the tumor into the vein, from compression of the vein by the tumor, or from epithelial damage as a result of tumor invasion. The tumor thrombus has been described to extend into the right atrium.^6,8 The mechanism of thrombosis remains unclear in the patient postlobectomy or postlung transplantation, although intraoperative torsion and injury of vessels are implicated. Similar to deep vein thrombosis, PVT could also result from intimal damage or sluggish flow in the pulmonary stump in the postoperative patient.^2,9,10

The presentation of PVT is usually nonspecific, including dyspnea, cough, pleuritic chest pain, and hemoptysis. It has been reported as causing massive hemoptysis due to acute pulmonary infarction.⁷ Acute PVT occurring postoperatively in the lung transplant patient may be disastrous and lead to early postoperative allograft failure.¹¹ Pulmonary vein thrombosis may also present more insidiously with recurrent pulmonary edema and pulmonary fibrosis.¹² This patient presented with abdominal pain; further workup led to the finding of a lung mass. Pulmonary vein thrombosis has been reported to result in systemic emboli, resulting in cerebrovascular accidents, or it can manifest as aseptic and tumor emboli.^2,5,10,13,14

 

Newer CT techniques have made identifying PVT possible in a similar manner to which pulmonary arterial emboli are detected by using the pulmonary venous phase of a contrast CT of the chest.⁵ Echocardiography may demonstrate the extension of the thrombus into the atrium; a transesophageal echocardiogram would be preferable over a transthoracic echocardiogram. Magnetic resonance imaging of the chest is another useful modality for diagnosis, because it is able to distinguish between a bland thrombus and a tumor thrombus in the pulmonary vein.¹⁵

Treatment of PVT depends on the overall clinical condition of the patient. Irrespective of the etiology, a review of the literature does not indicate the preferred duration of anticoagulation or preference for modality of anticoagulation between oral vitamin K antagonists or heparin—low molecular or unfractionated.^1,3-6 Patients who develop PVT following malignancy are usually anticoagulated with therapy for the cancer. The treatment of PVT in the setting of lung transplant is more challenging and includes systemic heparinization, thrombolytics, and surgical thrombectomy.^3,11,16 The majority of the literature includes case reports with varying morbidity and mortality, depending on the etiology. Ninety-day mortality of 38% was reported following lung transplant.³

CONCLUSION

Pulmonary vein thrombosis presents in a nonspecific manner. The diagnosis is now more readily made with the advent of a variety of diagnostic modalities, especially with transesophageal echocardiography, which may be performed at the bedside in the intensive care unit. The treatment remains challenging with mortality dependent on the etiology. A diagnosis of PVT needs to be considered in patients with appropriate risk factors. A high index of suspicion will enable the diagnosis in the proper clinical scenario.

Author disclosures
The authors report no actual or potential conflicts of interest with regard to this article.

Disclaimer
The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the U.S. Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.

Case

A 100-year-old woman with a history of hypertension, hypothyroidism, and moderate Alzheimer dementia was brought to the ED by emergency medical services (EMS) for altered mental status after her home health aide (HHA) noted a change in the patient’s behavior. For the past few days, the patient’s appetite waned, and she became progressively more lethargic, not eating for over 24 hours. The aide activated 911 on the direction of the patient’s primary care physician. There were no reported changes to the patient’s medications which included aspirin, levothyroxine, and hydrochlorothiazide. She was unable to provide any meaningful history. 

On arrival to the ED, the patient appeared comfortable in bed. She was sleepy, but easily aroused. Initial vital signs were: heart rate, 110 beats/minute; respiratory rate, 12 breaths/minute; blood pressure, 163/103 mm Hg; oral temperature, 98.2˚F. Oxygen (O₂) saturation was 96% on room air. She was oriented to person only and responded appropriately to simple questions, intermittently following one-step commands. She was unable to attend and required redirection throughout the interview. (According to the aide, this behavior was different than her baseline.)

The patient’s head and neck examination were notable for some mild, boggy, periorbital edema and dry mucous membranes. Her thyroid examination was normal; her lungs were clear; and her cardiac examination revealed a 2/6 systolic ejection murmur over the second right intercostal space. Examination of the abdomen, extremities, and skin was unrevealing, and there were no gross focal neurological deficits. Her reflexes were normal throughout. 

Initial assessment of this patient suggested a diagnosis of dementia and hypoactive delirium—the latter due to one or more of several possible etiologies.

Altered Mental Status

While altered mental status is a billable medical ICD-9-CM code¹ used to specify a diagnosis on a reimbursement claim, it is not a disease state itself. Instead, it is a catchall phrase that incorporates any change in mental status, encompassing symptomatology that may have the largest differential diagnosis encountered in emergency medicine.

Delirium
An important category of altered mental status is delirium. The diagnostic criteria² for delirium in DSM-V have remained essentially unchanged from DSM-IV; however, the prevalence of delirium as one of the key geriatric syndromes has grown as a result of increased research and education, particularly in the emergency medical setting. Distilled down, delirium can be defined as an acute change in mental status not caused by underlying dementia. Its cause is often multifactorial, and it is frequently an underappreciated consequence of both critical illness and the hospital environment.³

Delirium is an emergency unto itself, with an in-hospital mortality rate mirroring that of sepsis or acute myocardial infarction.⁴ The older-adult population is especially at risk of delirium and can present with one of three clinical subtypes: hyperactive (ie, agitated, etc), hypoactive (ie, somnolent, lethargic, stuporous, etc), and mixed type.⁵

Delirium is a form of acute brain dysfunction involving a complex interaction between patient vulnerability factors and precipitating factors,^6,7 resulting in either impaired cerebral metabolism and/or neurotransmitter disequilibrium. Both sets of factors must be considered. For example, delirium may develop in a particularly vulnerable 92-year-old man with moderate dementia and a mild lower leg cellulitis without signs of sepsis.

The composition of factors precipitating the onset of delirium includes age, dementia, alcohol use, depression, illness severity, and drug exposure—notably benzodiazepines, opiates, and medications with anticholinergic properties.⁸ While major precipitants or causes of delirium, traditionally considered as potentially life-threatening acute events, are well known (Table 1), others are often overlooked (Table 2). Yet, ironically, these more frequently bypassed causes are more readily reversible, once discovered, such as inadequate pain control, urinary retention, constipation, dehydration, polypharmacy, and negative environmental conditions in the patient’s immediate surroundings. These findings have recently been corroborated.⁹

Hypoactive or “quiet” delirium is particularly difficult to detect because of the lack of agitation, and can therefore make the evaluation of the underlying precipitant especially challenging.

The Confusion Assessment Method
Identifying delirium can be a particular challenge for emergency physicians (EPs), especially when the patient has an underlying diagnosis of dementia and the specific degree of cognitive impairment is not known. The Confusion Assessment Method (CAM)¹⁰ is the most commonly used tool in the critical care setting11 and is the only validated tool for the ED, with an 86% sensitivity and 100% specificity.12 It evaluates four elements: (1) acute onset and fluctuating course; (2) inattention; (3) disorganized thinking; and (4) altered level of consciousness. A patient must demonstrate the first two elements in addition to either the third or fourth element to be considered to have delirium.¹⁰

The CAM intensive care unit scale has the potential to be even more applicable in the ED. Recent findings support its validation.⁹

 

Case Continued

With respect to the elderly patient in this case with dementia and multiple potential causes for hypoactive delirium, the life-saving measure was the New York City (NYC) mandate that all 911 responding EMS workers wear ambient carbon monoxide (CO) detectors. Though the value of these detectors is controversial due to their low sensitivity, the emergency medical technicians (EMTs) detected an elevated CO level of 300 ppm when they arrived at the patient’s home.

Without this information, the patient’s age and clinical presentation would almost certainly have prompted an extensive evaluation to determine the etiology of her change in mental status and most likely would have missed the true cause of CO toxicity, which was confirmed by venous co-oximetry showing the patient to have a carboxyhemoglobin (HbCO) level of 19.5%.

Carbon Monoxide Toxicity

Carbon monoxide affects multiple cell types. It binds to myoglobin and in high concentrations depresses myocardial contractility. In platelets, CO displaces nitric oxide potentially resulting in vasodilation. Life-threatening CO poisoning causes hypotension, syncope, tachycardia, and an altered mental status. Delayed neuropsychiatric sequelae also may occur as the result of free radical injury to the brain.¹³

Symptoms
Patients with chronic CO poisoning who can adequately communicate may report nausea, headache, lightheadedness, and lethargy mimicking other seasonal illnesses. In debilitated or cognitively impaired patients who are unable to communicate, findings may include tachycardia, a mild change in mental status, and little else. Prolonged exposure and physiologic accumulation of CO may cause depressed mental status, coma, or death.

Although HbCO levels are confirmatory of exposure, venous levels do not necessarily reflect tissue concentrations or outcomes. Patients with a similar level to that of this patient (19.5%) may present with no symptoms, mild headache, or a deep coma depending on the duration of exposure to CO. 

Definitive treatment is removal from the toxic environment and prompt administration of O₂. In some cases, hyperbaric therapy may be beneficial.¹⁴

Diagnosis
Although CO exposure is the most common cause of poisoning death worldwide, its detection requires a high index of suspicion, especially in areas where public-health protection measures are absent.

Although CO exposure is the most common cause of poisoning death worldwide, its detection requires a high index of suspicion, especially in areas where public-health protection measures are absent.

It is rarely easy to diagnose the first case of an illness of which one is unfamiliar or not accustomed to treating. Likewise, it is very difficult to consider, diagnose and, as a result, effectively manage the first presentation of a known condition that is typically seasonal or linked to a different geographic location. Acute presentations of environmental exposures, illicit drug poisonings, and communicable infectious diseases are increasingly the purview of emergency medicine. Whether it is the first case of Ebola, of severe acute respiratory syndrome, the influenza virus, a new lethal street drug overdose, or CO poisoning prior to the onset of winter, maintaining a high index of suspicion for the “index case” is of paramount importance. The patient presented here, the first CO poisoning of the season at the authors’ institution, illustrates the responsibility the EP to consider, diagnose, and prevent a wide-range of deadly consequences—injury prevention as the result of vigilance. Moreover, the consequences of missing the diagnosis would have placed others at risk for continued poisoning and possibly death.

Portable and Ambient Carboxyhemoglobin Monitors

The NYC Department of Health (NYCDOH) requires that all EMTs and paramedics wear CO detectors and all residential housing contain CO monitors. The NYCDOH also mandates that all identified cases of CO poisoning be reported to the NYC Poison Control Center. This centralization of data on any and all patients exposed to CO can result in an investigation of the source of CO by the fire department and capture symptomatic patients who present for care outside of the 911 response system. The source of CO in this patient was ultimately traced to a faulty furnace that was repaired to prevent others in the building from becoming victims of CO poisoning.

It should be noted that portable noninvasive HbCO monitors may be inadequate to rule out CO poisoning as the sensitivity of such devices can be as low as 48%.¹⁵ Carbon monoxide poisoning can result from brief exposure to a high ambient concentration, such as a fire in which environmental concentrations may exceed 500 ppm or more insidiously, in a setting of a chronic exposure. Faulty furnaces—a common seasonal cause of CO poisoning—may continue to produce adequate heat and fail to prompt any concerns.

Since CO is colorless and odorless, ambient CO detectors stationed in the home are the best means of alerting one to exposure. In this case, though mandated by NYCDOH, a CO detector was not present in the patient’s home.

 

Case Conclusion

Through the rapid identification of CO poisoning in this elderly patient with altered mental status, EMS was able to evacuate the building while bringing the elderly tenant and her home attendant to the ED.

Based on the elderly patient’s elevated HbCO level, she was treated with O₂ and discharged from the hospital the following day feeling well. In addition to the patient’s symptoms, when the aide was interviewed, she reported that she had been experiencing daily headaches, which she said soon resolved on departure from her client’s house. Her symptoms had been bothersome, but not so severe as to prompt her to seek medical attention. The aide was found to have an HbCO level of 12.5% and was discharged from the ED after 6 hours of observation and O₂ therapy. The third occupant of the building, a tenant, was also brought to the ED and found to have an HbCO level of 12%. The tenant was treated with O₂ therapy and discharged to home.

Dr Caldwell is an assistant professor of medicine in the department of emergency medicine, New York Presbyterian Hospital/Weill Cornell Medical Center, New York. Dr Rao is an assistant professor of emergency medicine; and the chief in the division of medical Toxicology, New York Presbyterian Hospital/Weill Cornell Medical Center, New York. Dr Stern is an assistant professor of medicine, department of emergency medicine; chief of geriatric emergency medicine; and codirector of geriatric emergency medicine fellowship at New York Presbyterian Hospital/Weill Cornell Medical Center, New York.

Case

A 100-year-old woman with a history of hypertension, hypothyroidism, and moderate Alzheimer dementia was brought to the ED by emergency medical services (EMS) for altered mental status after her home health aide (HHA) noted a change in the patient’s behavior. For the past few days, the patient’s appetite waned, and she became progressively more lethargic, not eating for over 24 hours. The aide activated 911 on the direction of the patient’s primary care physician. There were no reported changes to the patient’s medications which included aspirin, levothyroxine, and hydrochlorothiazide. She was unable to provide any meaningful history. 

On arrival to the ED, the patient appeared comfortable in bed. She was sleepy, but easily aroused. Initial vital signs were: heart rate, 110 beats/minute; respiratory rate, 12 breaths/minute; blood pressure, 163/103 mm Hg; oral temperature, 98.2˚F. Oxygen (O₂) saturation was 96% on room air. She was oriented to person only and responded appropriately to simple questions, intermittently following one-step commands. She was unable to attend and required redirection throughout the interview. (According to the aide, this behavior was different than her baseline.)

The patient’s head and neck examination were notable for some mild, boggy, periorbital edema and dry mucous membranes. Her thyroid examination was normal; her lungs were clear; and her cardiac examination revealed a 2/6 systolic ejection murmur over the second right intercostal space. Examination of the abdomen, extremities, and skin was unrevealing, and there were no gross focal neurological deficits. Her reflexes were normal throughout. 

Initial assessment of this patient suggested a diagnosis of dementia and hypoactive delirium—the latter due to one or more of several possible etiologies.

Altered Mental Status

While altered mental status is a billable medical ICD-9-CM code¹ used to specify a diagnosis on a reimbursement claim, it is not a disease state itself. Instead, it is a catchall phrase that incorporates any change in mental status, encompassing symptomatology that may have the largest differential diagnosis encountered in emergency medicine.

Delirium
An important category of altered mental status is delirium. The diagnostic criteria² for delirium in DSM-V have remained essentially unchanged from DSM-IV; however, the prevalence of delirium as one of the key geriatric syndromes has grown as a result of increased research and education, particularly in the emergency medical setting. Distilled down, delirium can be defined as an acute change in mental status not caused by underlying dementia. Its cause is often multifactorial, and it is frequently an underappreciated consequence of both critical illness and the hospital environment.³

Delirium is an emergency unto itself, with an in-hospital mortality rate mirroring that of sepsis or acute myocardial infarction.⁴ The older-adult population is especially at risk of delirium and can present with one of three clinical subtypes: hyperactive (ie, agitated, etc), hypoactive (ie, somnolent, lethargic, stuporous, etc), and mixed type.⁵

Delirium is a form of acute brain dysfunction involving a complex interaction between patient vulnerability factors and precipitating factors,^6,7 resulting in either impaired cerebral metabolism and/or neurotransmitter disequilibrium. Both sets of factors must be considered. For example, delirium may develop in a particularly vulnerable 92-year-old man with moderate dementia and a mild lower leg cellulitis without signs of sepsis.

The composition of factors precipitating the onset of delirium includes age, dementia, alcohol use, depression, illness severity, and drug exposure—notably benzodiazepines, opiates, and medications with anticholinergic properties.⁸ While major precipitants or causes of delirium, traditionally considered as potentially life-threatening acute events, are well known (Table 1), others are often overlooked (Table 2). Yet, ironically, these more frequently bypassed causes are more readily reversible, once discovered, such as inadequate pain control, urinary retention, constipation, dehydration, polypharmacy, and negative environmental conditions in the patient’s immediate surroundings. These findings have recently been corroborated.⁹

Hypoactive or “quiet” delirium is particularly difficult to detect because of the lack of agitation, and can therefore make the evaluation of the underlying precipitant especially challenging.

The Confusion Assessment Method
Identifying delirium can be a particular challenge for emergency physicians (EPs), especially when the patient has an underlying diagnosis of dementia and the specific degree of cognitive impairment is not known. The Confusion Assessment Method (CAM)¹⁰ is the most commonly used tool in the critical care setting11 and is the only validated tool for the ED, with an 86% sensitivity and 100% specificity.12 It evaluates four elements: (1) acute onset and fluctuating course; (2) inattention; (3) disorganized thinking; and (4) altered level of consciousness. A patient must demonstrate the first two elements in addition to either the third or fourth element to be considered to have delirium.¹⁰

The CAM intensive care unit scale has the potential to be even more applicable in the ED. Recent findings support its validation.⁹

 

Case Continued

With respect to the elderly patient in this case with dementia and multiple potential causes for hypoactive delirium, the life-saving measure was the New York City (NYC) mandate that all 911 responding EMS workers wear ambient carbon monoxide (CO) detectors. Though the value of these detectors is controversial due to their low sensitivity, the emergency medical technicians (EMTs) detected an elevated CO level of 300 ppm when they arrived at the patient’s home.

Without this information, the patient’s age and clinical presentation would almost certainly have prompted an extensive evaluation to determine the etiology of her change in mental status and most likely would have missed the true cause of CO toxicity, which was confirmed by venous co-oximetry showing the patient to have a carboxyhemoglobin (HbCO) level of 19.5%.

Carbon Monoxide Toxicity

Carbon monoxide affects multiple cell types. It binds to myoglobin and in high concentrations depresses myocardial contractility. In platelets, CO displaces nitric oxide potentially resulting in vasodilation. Life-threatening CO poisoning causes hypotension, syncope, tachycardia, and an altered mental status. Delayed neuropsychiatric sequelae also may occur as the result of free radical injury to the brain.¹³

Symptoms
Patients with chronic CO poisoning who can adequately communicate may report nausea, headache, lightheadedness, and lethargy mimicking other seasonal illnesses. In debilitated or cognitively impaired patients who are unable to communicate, findings may include tachycardia, a mild change in mental status, and little else. Prolonged exposure and physiologic accumulation of CO may cause depressed mental status, coma, or death.

Although HbCO levels are confirmatory of exposure, venous levels do not necessarily reflect tissue concentrations or outcomes. Patients with a similar level to that of this patient (19.5%) may present with no symptoms, mild headache, or a deep coma depending on the duration of exposure to CO. 

Definitive treatment is removal from the toxic environment and prompt administration of O₂. In some cases, hyperbaric therapy may be beneficial.¹⁴

Diagnosis
Although CO exposure is the most common cause of poisoning death worldwide, its detection requires a high index of suspicion, especially in areas where public-health protection measures are absent.

Although CO exposure is the most common cause of poisoning death worldwide, its detection requires a high index of suspicion, especially in areas where public-health protection measures are absent.

It is rarely easy to diagnose the first case of an illness of which one is unfamiliar or not accustomed to treating. Likewise, it is very difficult to consider, diagnose and, as a result, effectively manage the first presentation of a known condition that is typically seasonal or linked to a different geographic location. Acute presentations of environmental exposures, illicit drug poisonings, and communicable infectious diseases are increasingly the purview of emergency medicine. Whether it is the first case of Ebola, of severe acute respiratory syndrome, the influenza virus, a new lethal street drug overdose, or CO poisoning prior to the onset of winter, maintaining a high index of suspicion for the “index case” is of paramount importance. The patient presented here, the first CO poisoning of the season at the authors’ institution, illustrates the responsibility the EP to consider, diagnose, and prevent a wide-range of deadly consequences—injury prevention as the result of vigilance. Moreover, the consequences of missing the diagnosis would have placed others at risk for continued poisoning and possibly death.

Portable and Ambient Carboxyhemoglobin Monitors

The NYC Department of Health (NYCDOH) requires that all EMTs and paramedics wear CO detectors and all residential housing contain CO monitors. The NYCDOH also mandates that all identified cases of CO poisoning be reported to the NYC Poison Control Center. This centralization of data on any and all patients exposed to CO can result in an investigation of the source of CO by the fire department and capture symptomatic patients who present for care outside of the 911 response system. The source of CO in this patient was ultimately traced to a faulty furnace that was repaired to prevent others in the building from becoming victims of CO poisoning.

It should be noted that portable noninvasive HbCO monitors may be inadequate to rule out CO poisoning as the sensitivity of such devices can be as low as 48%.¹⁵ Carbon monoxide poisoning can result from brief exposure to a high ambient concentration, such as a fire in which environmental concentrations may exceed 500 ppm or more insidiously, in a setting of a chronic exposure. Faulty furnaces—a common seasonal cause of CO poisoning—may continue to produce adequate heat and fail to prompt any concerns.

Since CO is colorless and odorless, ambient CO detectors stationed in the home are the best means of alerting one to exposure. In this case, though mandated by NYCDOH, a CO detector was not present in the patient’s home.

 

Case Conclusion

Through the rapid identification of CO poisoning in this elderly patient with altered mental status, EMS was able to evacuate the building while bringing the elderly tenant and her home attendant to the ED.

Based on the elderly patient’s elevated HbCO level, she was treated with O₂ and discharged from the hospital the following day feeling well. In addition to the patient’s symptoms, when the aide was interviewed, she reported that she had been experiencing daily headaches, which she said soon resolved on departure from her client’s house. Her symptoms had been bothersome, but not so severe as to prompt her to seek medical attention. The aide was found to have an HbCO level of 12.5% and was discharged from the ED after 6 hours of observation and O₂ therapy. The third occupant of the building, a tenant, was also brought to the ED and found to have an HbCO level of 12%. The tenant was treated with O₂ therapy and discharged to home.

Dr Caldwell is an assistant professor of medicine in the department of emergency medicine, New York Presbyterian Hospital/Weill Cornell Medical Center, New York. Dr Rao is an assistant professor of emergency medicine; and the chief in the division of medical Toxicology, New York Presbyterian Hospital/Weill Cornell Medical Center, New York. Dr Stern is an assistant professor of medicine, department of emergency medicine; chief of geriatric emergency medicine; and codirector of geriatric emergency medicine fellowship at New York Presbyterian Hospital/Weill Cornell Medical Center, New York.

Case

A 100-year-old woman with a history of hypertension, hypothyroidism, and moderate Alzheimer dementia was brought to the ED by emergency medical services (EMS) for altered mental status after her home health aide (HHA) noted a change in the patient’s behavior. For the past few days, the patient’s appetite waned, and she became progressively more lethargic, not eating for over 24 hours. The aide activated 911 on the direction of the patient’s primary care physician. There were no reported changes to the patient’s medications which included aspirin, levothyroxine, and hydrochlorothiazide. She was unable to provide any meaningful history. 

On arrival to the ED, the patient appeared comfortable in bed. She was sleepy, but easily aroused. Initial vital signs were: heart rate, 110 beats/minute; respiratory rate, 12 breaths/minute; blood pressure, 163/103 mm Hg; oral temperature, 98.2˚F. Oxygen (O₂) saturation was 96% on room air. She was oriented to person only and responded appropriately to simple questions, intermittently following one-step commands. She was unable to attend and required redirection throughout the interview. (According to the aide, this behavior was different than her baseline.)

The patient’s head and neck examination were notable for some mild, boggy, periorbital edema and dry mucous membranes. Her thyroid examination was normal; her lungs were clear; and her cardiac examination revealed a 2/6 systolic ejection murmur over the second right intercostal space. Examination of the abdomen, extremities, and skin was unrevealing, and there were no gross focal neurological deficits. Her reflexes were normal throughout. 

Initial assessment of this patient suggested a diagnosis of dementia and hypoactive delirium—the latter due to one or more of several possible etiologies.

Altered Mental Status

While altered mental status is a billable medical ICD-9-CM code¹ used to specify a diagnosis on a reimbursement claim, it is not a disease state itself. Instead, it is a catchall phrase that incorporates any change in mental status, encompassing symptomatology that may have the largest differential diagnosis encountered in emergency medicine.

Delirium
An important category of altered mental status is delirium. The diagnostic criteria² for delirium in DSM-V have remained essentially unchanged from DSM-IV; however, the prevalence of delirium as one of the key geriatric syndromes has grown as a result of increased research and education, particularly in the emergency medical setting. Distilled down, delirium can be defined as an acute change in mental status not caused by underlying dementia. Its cause is often multifactorial, and it is frequently an underappreciated consequence of both critical illness and the hospital environment.³

Delirium is an emergency unto itself, with an in-hospital mortality rate mirroring that of sepsis or acute myocardial infarction.⁴ The older-adult population is especially at risk of delirium and can present with one of three clinical subtypes: hyperactive (ie, agitated, etc), hypoactive (ie, somnolent, lethargic, stuporous, etc), and mixed type.⁵

Delirium is a form of acute brain dysfunction involving a complex interaction between patient vulnerability factors and precipitating factors,^6,7 resulting in either impaired cerebral metabolism and/or neurotransmitter disequilibrium. Both sets of factors must be considered. For example, delirium may develop in a particularly vulnerable 92-year-old man with moderate dementia and a mild lower leg cellulitis without signs of sepsis.

The composition of factors precipitating the onset of delirium includes age, dementia, alcohol use, depression, illness severity, and drug exposure—notably benzodiazepines, opiates, and medications with anticholinergic properties.⁸ While major precipitants or causes of delirium, traditionally considered as potentially life-threatening acute events, are well known (Table 1), others are often overlooked (Table 2). Yet, ironically, these more frequently bypassed causes are more readily reversible, once discovered, such as inadequate pain control, urinary retention, constipation, dehydration, polypharmacy, and negative environmental conditions in the patient’s immediate surroundings. These findings have recently been corroborated.⁹

Hypoactive or “quiet” delirium is particularly difficult to detect because of the lack of agitation, and can therefore make the evaluation of the underlying precipitant especially challenging.

The Confusion Assessment Method
Identifying delirium can be a particular challenge for emergency physicians (EPs), especially when the patient has an underlying diagnosis of dementia and the specific degree of cognitive impairment is not known. The Confusion Assessment Method (CAM)¹⁰ is the most commonly used tool in the critical care setting11 and is the only validated tool for the ED, with an 86% sensitivity and 100% specificity.12 It evaluates four elements: (1) acute onset and fluctuating course; (2) inattention; (3) disorganized thinking; and (4) altered level of consciousness. A patient must demonstrate the first two elements in addition to either the third or fourth element to be considered to have delirium.¹⁰

The CAM intensive care unit scale has the potential to be even more applicable in the ED. Recent findings support its validation.⁹

 

Case Continued

With respect to the elderly patient in this case with dementia and multiple potential causes for hypoactive delirium, the life-saving measure was the New York City (NYC) mandate that all 911 responding EMS workers wear ambient carbon monoxide (CO) detectors. Though the value of these detectors is controversial due to their low sensitivity, the emergency medical technicians (EMTs) detected an elevated CO level of 300 ppm when they arrived at the patient’s home.

Without this information, the patient’s age and clinical presentation would almost certainly have prompted an extensive evaluation to determine the etiology of her change in mental status and most likely would have missed the true cause of CO toxicity, which was confirmed by venous co-oximetry showing the patient to have a carboxyhemoglobin (HbCO) level of 19.5%.

Carbon Monoxide Toxicity

Carbon monoxide affects multiple cell types. It binds to myoglobin and in high concentrations depresses myocardial contractility. In platelets, CO displaces nitric oxide potentially resulting in vasodilation. Life-threatening CO poisoning causes hypotension, syncope, tachycardia, and an altered mental status. Delayed neuropsychiatric sequelae also may occur as the result of free radical injury to the brain.¹³

Symptoms
Patients with chronic CO poisoning who can adequately communicate may report nausea, headache, lightheadedness, and lethargy mimicking other seasonal illnesses. In debilitated or cognitively impaired patients who are unable to communicate, findings may include tachycardia, a mild change in mental status, and little else. Prolonged exposure and physiologic accumulation of CO may cause depressed mental status, coma, or death.

Although HbCO levels are confirmatory of exposure, venous levels do not necessarily reflect tissue concentrations or outcomes. Patients with a similar level to that of this patient (19.5%) may present with no symptoms, mild headache, or a deep coma depending on the duration of exposure to CO. 

Definitive treatment is removal from the toxic environment and prompt administration of O₂. In some cases, hyperbaric therapy may be beneficial.¹⁴

Diagnosis
Although CO exposure is the most common cause of poisoning death worldwide, its detection requires a high index of suspicion, especially in areas where public-health protection measures are absent.

Although CO exposure is the most common cause of poisoning death worldwide, its detection requires a high index of suspicion, especially in areas where public-health protection measures are absent.

It is rarely easy to diagnose the first case of an illness of which one is unfamiliar or not accustomed to treating. Likewise, it is very difficult to consider, diagnose and, as a result, effectively manage the first presentation of a known condition that is typically seasonal or linked to a different geographic location. Acute presentations of environmental exposures, illicit drug poisonings, and communicable infectious diseases are increasingly the purview of emergency medicine. Whether it is the first case of Ebola, of severe acute respiratory syndrome, the influenza virus, a new lethal street drug overdose, or CO poisoning prior to the onset of winter, maintaining a high index of suspicion for the “index case” is of paramount importance. The patient presented here, the first CO poisoning of the season at the authors’ institution, illustrates the responsibility the EP to consider, diagnose, and prevent a wide-range of deadly consequences—injury prevention as the result of vigilance. Moreover, the consequences of missing the diagnosis would have placed others at risk for continued poisoning and possibly death.

Portable and Ambient Carboxyhemoglobin Monitors

The NYC Department of Health (NYCDOH) requires that all EMTs and paramedics wear CO detectors and all residential housing contain CO monitors. The NYCDOH also mandates that all identified cases of CO poisoning be reported to the NYC Poison Control Center. This centralization of data on any and all patients exposed to CO can result in an investigation of the source of CO by the fire department and capture symptomatic patients who present for care outside of the 911 response system. The source of CO in this patient was ultimately traced to a faulty furnace that was repaired to prevent others in the building from becoming victims of CO poisoning.

It should be noted that portable noninvasive HbCO monitors may be inadequate to rule out CO poisoning as the sensitivity of such devices can be as low as 48%.¹⁵ Carbon monoxide poisoning can result from brief exposure to a high ambient concentration, such as a fire in which environmental concentrations may exceed 500 ppm or more insidiously, in a setting of a chronic exposure. Faulty furnaces—a common seasonal cause of CO poisoning—may continue to produce adequate heat and fail to prompt any concerns.

Since CO is colorless and odorless, ambient CO detectors stationed in the home are the best means of alerting one to exposure. In this case, though mandated by NYCDOH, a CO detector was not present in the patient’s home.

 

Case Conclusion

Through the rapid identification of CO poisoning in this elderly patient with altered mental status, EMS was able to evacuate the building while bringing the elderly tenant and her home attendant to the ED.

Based on the elderly patient’s elevated HbCO level, she was treated with O₂ and discharged from the hospital the following day feeling well. In addition to the patient’s symptoms, when the aide was interviewed, she reported that she had been experiencing daily headaches, which she said soon resolved on departure from her client’s house. Her symptoms had been bothersome, but not so severe as to prompt her to seek medical attention. The aide was found to have an HbCO level of 12.5% and was discharged from the ED after 6 hours of observation and O₂ therapy. The third occupant of the building, a tenant, was also brought to the ED and found to have an HbCO level of 12%. The tenant was treated with O₂ therapy and discharged to home.

Dr Caldwell is an assistant professor of medicine in the department of emergency medicine, New York Presbyterian Hospital/Weill Cornell Medical Center, New York. Dr Rao is an assistant professor of emergency medicine; and the chief in the division of medical Toxicology, New York Presbyterian Hospital/Weill Cornell Medical Center, New York. Dr Stern is an assistant professor of medicine, department of emergency medicine; chief of geriatric emergency medicine; and codirector of geriatric emergency medicine fellowship at New York Presbyterian Hospital/Weill Cornell Medical Center, New York.

Case

A 58-year-old man presented to the ED via emergency medical services (EMS) with shortness of breath, lightheadedness, and nausea. Upon arrival to the ED, most of his symptoms had resolved. The patient reported that he had taken a 2-hour flight into town the previous day and had spent an uneventful evening at a local hotel. He said that he began experiencing shortness of breath and lightheadedness soon after entering his rental vehicle an hour prior to presentation, explaining that he felt as if he “could not get any air.”

He denied chest pain, leg pain or swelling, abdominal pain, or recent illness. His past medical history was significant only for hypertension, for which he was taking losartan and amlodipine. He had no drug allergies, surgical history, or smoking history. Of note, when the hotel clerk got in the same rental vehicle to move it, he developed symptoms similar to those of the patient. As with the patient, the clerk’s symptoms quickly resolved after he got out of the vehicle.

The patient’s vital signs at examination were: oral temperature, 97.5^oF; pulse, 62 beats/minute; respiratory rate (RR), 18 breaths/minute; and blood pressure, 133/83 mm Hg. Oxygen (O₂) saturation was 100% on room air. He was alert and oriented, in no distress, easily conversational, and without diaphoresis. The lungs were clear to auscultation bilaterally, and there was no calf swelling, tenderness, or palpable cords. The remainder of the physical examination was completely normal.

Ancillary studies revealed a normal chest X-ray. An electrocardiogram demonstrated sinus bradycardia with a rate of 56, but no evidence of ischemia or right heart strain. A complete blood count, troponin I, D-dimer, and creatine phosphokinase (CPK) with MB fraction levels were all within normal limits. A serum chemistry panel was also within normal limits, except for a serum glucose level of 181 mg/dL. Venous co-oximetry showed a carboxyhemoglobin level of 0.0, and methemoglobin level of 0.5 gm% (normal range, 0.4-1.5).

Since both the patient and hotel clerk’s symptoms started when each was in the rental car, the patient was questioned about the vehicle and its contents. The car was a late model rental in good condition per report. The patient informed the treating EP that he worked as a decorative cake salesman and had brought cake samples with him to display at a trade show. He further stated that he had left these samples in the car overnight, packed in dry ice.

Upon learning this information, EMS was contacted and instructed to return to the hotel and rental vehicle. The hotel room was noted to have normal levels of O₂ and carbon monoxide (CO) on measurement. Investigation of the car revealed normal levels of CO, but O₂ levels too low to read on the sensor. The emergency team concluded that the dry ice (the solid form of carbon dioxide [CO₂]), sublimed to CO₂ gas overnight. This displaced the O₂ in the vehicle, resulting in severe hypoxia and the symptoms of both the patient and hotel clerk.

The patient was initially placed on 15 L of O₂ via a nonrebreather mask, but he was switched to 2 L of O₂ via nasal cannula shortly afterward. He was observed for a total of 4 hours after arrival; as he remained symptom-free, he was discharged home. The EP was not able to obtain postdischarge follow up information.

Discussion

Carbon dioxide primarily acts as a simple asphyxiant, but it also dissolves in serum as carbonic acid, resulting in a metabolic acidosis. Carbon dioxide is a prevalent gas that is part of everyday life, from an agent in fire extinguishers and carbonation in beverages to byproducts of cellular metabolism. Similar to CO, it is a colorless and odorless gas. Carbon dioxide is commonly used in the food industry as dry ice to keep items cold. In its solid state, CO₂ can cause severe frostbite with direct contact, similar to a burn. However, when dry ice is warmed and sublimated to a gaseous state, large amounts of CO₂ are generated, and this heavy gas can accumulate and displace air (ie, atmospheric O₂), especially in confined spaces. In low concentrations, gaseous CO₂ appears to have minimal toxicological effects, but at higher concentrations it can cause tachycardia, tachypnea, dyspnea, visual disturbances, arrhythmias, impaired levels of consciousness, and even death.

Carbon dioxide primarily acts as a simple asphyxiant, but it also dissolves in serum as carbonic acid, resulting in a metabolic acidosis. Compensation for this acidosis is accomplished by an increased RR (ie, respiratory alkalosis), which further worsens the intake of CO₂.^1,2

 

The normal concentration of CO₂ in the atmosphere is approximately 0.04% (396 ppm). The Occupational Safety and Health Administration (OSHA) has set a maximum safe exposure level of CO₂ at 0.5% (5,000 ppm) over an 8-hour day.³ Concentrations as low as 1% (10,000 ppm) may cause drowsiness. Exposure to concentrations of 7% to 10% for several minutes to an hour results in headache, tachycardia, dyspnea, and hyperventilation. At levels of 10% to 15%, dizziness, severe muscle twitching, and loss of consciousness can occur after only a few minutes. Death occurs within minutes at concentrations greater than 30%.²

Carbon dioxide also acts as a potent cerebral vasodilator, which may explain symptoms such as headache and dizziness.² The severity of symptoms is dependent on the concentration of CO₂, the length of the exposure, and the underlying health of the patient. Elevated concentrations of CO₂ can occur in areas where there is limited or poor ventilation, such as in a mine (where it is known as blackdamp, stythe, or choke damp),⁴ submarine, grain silo, or a sealed building without mechanical ventilation.

Other Case Presentations

Similar cases as the one presented in this article have been described in the literature. In one such case, following Hurricane Ivan, a 34-year-old-man placed four 25-pound blocks of dry ice wrapped in paper in the front seat of his truck with the windows closed.⁵ After driving less than one quarter of a mile, he developed dyspnea and telephoned for help before losing consciousness. Fortunately, he was found in time and recovered soon after the doors to his truck were opened.⁵

In another case, a 59-year-old man entered a walk-in freezer that contained dry ice wrapped loosely in plastic. He was found inside the freezer 20 minutes later in cardiac arrest; resuscitation efforts were unsuccessful. Investigation of the freezer found the initial O₂ concentration to be 13% (normal level, 20.93%) and an estimated CO₂ level of 40%.⁵

Similarly, a 35-year-old woman was inadvertently locked in a bank vault while storing receipts. In a bid for help, she pulled the fire alarm, which triggered a CO₂-based fire-extinguishing system. The fire department responded and found the woman dead in the vault 30 minutes later. The cause of death was labeled as CO₂ intoxication.⁶

Natural Phenomenon

In addition to the foregoing case examples, there have been documented cases of CO₂ toxicity associated with volcanic eruption and other natural phenomena. One example is the Lake Nyos, Cameroon, West Africa incident in 1986. In this event, a magma pocket underlying the lake saturated the water with CO₂ stored as carbonic acid in the water. When a landslide hit the lake, it caused the carbonic acid stored in the depths of the lake to be upheaved to the surface, where it turned back into CO₂ and was released into the atmosphere. Since CO₂ is heavier than O₂, it displaced the O₂ near the ground, resulting in the suffocation and death of 1,700 people in the surrounding villages.²

Differential Diagnosis

When CO₂ toxicity is suspected, other conditions should be considered as there may be more than one process involved. For example, other causes of coma or dyspnea should be investigated, including trauma, hypoglycemia, CO, methemoglobinemia, or other metabolic processes. In addition, a patient may have a pre-existing condition, such as a trauma or an altered mental status due to drugs or alcohol, all of which can increase his or her susceptibility to the effects of CO₂.

Evaluation and Treatment

Useful laboratory testing includes arterial blood gas, venous co-oximetry for carboxyhemoglobin, chemistry panels, ethanol testing, and radiographs or computed tomography scanning, as indicated. 

Initial management of suspected CO₂ toxicity includes first removing the patient from the source of the gas. Rescuers must exercise caution so as to prevent a mass casualty incident. Once out of the dangerous environment, as long as the patient is conscious and spontaneously breathing, supportive measures are generally all that are necessary. Oxygen should be applied, after which the spontaneously breathing patient without underlying lung disease should rapidly return to normal. If there is marked decrease in mental status or poor respiratory drive despite O₂ administration, intubation with mechanical ventilation may be required. A higher than normal RR will help remove excessive CO₂ if mechanical ventilation is required. If a respiratory acidosis is present, intravenous sodium bicarbonate should be avoided as this may increase the level of serum CO₂. Intravenous fluids and other supportive measures, including treatment for any concurrent conditions, may be indicated.

Dr Schreckengaust is an emergency physician in the department of emergency medicine at Camp Lejune, North Carolina. Dr Lang is an assistant professor in the department of emergency medicine at Eastern Virginia Medical School, Norfolk; and a physician at Emergency Physicians of Tidewater, Norfolk, Virginia. Dr Counselman is the distinguished professor and chairman of the department of emergency medicine at Eastern Virginia Medical School, Norfolk; and a physician at Emergency Physicians of Tidewater, Norfolk, Virginia. He is also the associate editor in chief of EMERGENCY MEDICINE editorial board.

Case

A 58-year-old man presented to the ED via emergency medical services (EMS) with shortness of breath, lightheadedness, and nausea. Upon arrival to the ED, most of his symptoms had resolved. The patient reported that he had taken a 2-hour flight into town the previous day and had spent an uneventful evening at a local hotel. He said that he began experiencing shortness of breath and lightheadedness soon after entering his rental vehicle an hour prior to presentation, explaining that he felt as if he “could not get any air.”

He denied chest pain, leg pain or swelling, abdominal pain, or recent illness. His past medical history was significant only for hypertension, for which he was taking losartan and amlodipine. He had no drug allergies, surgical history, or smoking history. Of note, when the hotel clerk got in the same rental vehicle to move it, he developed symptoms similar to those of the patient. As with the patient, the clerk’s symptoms quickly resolved after he got out of the vehicle.

The patient’s vital signs at examination were: oral temperature, 97.5^oF; pulse, 62 beats/minute; respiratory rate (RR), 18 breaths/minute; and blood pressure, 133/83 mm Hg. Oxygen (O₂) saturation was 100% on room air. He was alert and oriented, in no distress, easily conversational, and without diaphoresis. The lungs were clear to auscultation bilaterally, and there was no calf swelling, tenderness, or palpable cords. The remainder of the physical examination was completely normal.

Ancillary studies revealed a normal chest X-ray. An electrocardiogram demonstrated sinus bradycardia with a rate of 56, but no evidence of ischemia or right heart strain. A complete blood count, troponin I, D-dimer, and creatine phosphokinase (CPK) with MB fraction levels were all within normal limits. A serum chemistry panel was also within normal limits, except for a serum glucose level of 181 mg/dL. Venous co-oximetry showed a carboxyhemoglobin level of 0.0, and methemoglobin level of 0.5 gm% (normal range, 0.4-1.5).

Since both the patient and hotel clerk’s symptoms started when each was in the rental car, the patient was questioned about the vehicle and its contents. The car was a late model rental in good condition per report. The patient informed the treating EP that he worked as a decorative cake salesman and had brought cake samples with him to display at a trade show. He further stated that he had left these samples in the car overnight, packed in dry ice.

Upon learning this information, EMS was contacted and instructed to return to the hotel and rental vehicle. The hotel room was noted to have normal levels of O₂ and carbon monoxide (CO) on measurement. Investigation of the car revealed normal levels of CO, but O₂ levels too low to read on the sensor. The emergency team concluded that the dry ice (the solid form of carbon dioxide [CO₂]), sublimed to CO₂ gas overnight. This displaced the O₂ in the vehicle, resulting in severe hypoxia and the symptoms of both the patient and hotel clerk.

The patient was initially placed on 15 L of O₂ via a nonrebreather mask, but he was switched to 2 L of O₂ via nasal cannula shortly afterward. He was observed for a total of 4 hours after arrival; as he remained symptom-free, he was discharged home. The EP was not able to obtain postdischarge follow up information.

Discussion

Carbon dioxide primarily acts as a simple asphyxiant, but it also dissolves in serum as carbonic acid, resulting in a metabolic acidosis. Carbon dioxide is a prevalent gas that is part of everyday life, from an agent in fire extinguishers and carbonation in beverages to byproducts of cellular metabolism. Similar to CO, it is a colorless and odorless gas. Carbon dioxide is commonly used in the food industry as dry ice to keep items cold. In its solid state, CO₂ can cause severe frostbite with direct contact, similar to a burn. However, when dry ice is warmed and sublimated to a gaseous state, large amounts of CO₂ are generated, and this heavy gas can accumulate and displace air (ie, atmospheric O₂), especially in confined spaces. In low concentrations, gaseous CO₂ appears to have minimal toxicological effects, but at higher concentrations it can cause tachycardia, tachypnea, dyspnea, visual disturbances, arrhythmias, impaired levels of consciousness, and even death.

Carbon dioxide primarily acts as a simple asphyxiant, but it also dissolves in serum as carbonic acid, resulting in a metabolic acidosis. Compensation for this acidosis is accomplished by an increased RR (ie, respiratory alkalosis), which further worsens the intake of CO₂.^1,2

 

The normal concentration of CO₂ in the atmosphere is approximately 0.04% (396 ppm). The Occupational Safety and Health Administration (OSHA) has set a maximum safe exposure level of CO₂ at 0.5% (5,000 ppm) over an 8-hour day.³ Concentrations as low as 1% (10,000 ppm) may cause drowsiness. Exposure to concentrations of 7% to 10% for several minutes to an hour results in headache, tachycardia, dyspnea, and hyperventilation. At levels of 10% to 15%, dizziness, severe muscle twitching, and loss of consciousness can occur after only a few minutes. Death occurs within minutes at concentrations greater than 30%.²

Carbon dioxide also acts as a potent cerebral vasodilator, which may explain symptoms such as headache and dizziness.² The severity of symptoms is dependent on the concentration of CO₂, the length of the exposure, and the underlying health of the patient. Elevated concentrations of CO₂ can occur in areas where there is limited or poor ventilation, such as in a mine (where it is known as blackdamp, stythe, or choke damp),⁴ submarine, grain silo, or a sealed building without mechanical ventilation.

Other Case Presentations

Similar cases as the one presented in this article have been described in the literature. In one such case, following Hurricane Ivan, a 34-year-old-man placed four 25-pound blocks of dry ice wrapped in paper in the front seat of his truck with the windows closed.⁵ After driving less than one quarter of a mile, he developed dyspnea and telephoned for help before losing consciousness. Fortunately, he was found in time and recovered soon after the doors to his truck were opened.⁵

In another case, a 59-year-old man entered a walk-in freezer that contained dry ice wrapped loosely in plastic. He was found inside the freezer 20 minutes later in cardiac arrest; resuscitation efforts were unsuccessful. Investigation of the freezer found the initial O₂ concentration to be 13% (normal level, 20.93%) and an estimated CO₂ level of 40%.⁵

Similarly, a 35-year-old woman was inadvertently locked in a bank vault while storing receipts. In a bid for help, she pulled the fire alarm, which triggered a CO₂-based fire-extinguishing system. The fire department responded and found the woman dead in the vault 30 minutes later. The cause of death was labeled as CO₂ intoxication.⁶

Natural Phenomenon

In addition to the foregoing case examples, there have been documented cases of CO₂ toxicity associated with volcanic eruption and other natural phenomena. One example is the Lake Nyos, Cameroon, West Africa incident in 1986. In this event, a magma pocket underlying the lake saturated the water with CO₂ stored as carbonic acid in the water. When a landslide hit the lake, it caused the carbonic acid stored in the depths of the lake to be upheaved to the surface, where it turned back into CO₂ and was released into the atmosphere. Since CO₂ is heavier than O₂, it displaced the O₂ near the ground, resulting in the suffocation and death of 1,700 people in the surrounding villages.²

Differential Diagnosis

When CO₂ toxicity is suspected, other conditions should be considered as there may be more than one process involved. For example, other causes of coma or dyspnea should be investigated, including trauma, hypoglycemia, CO, methemoglobinemia, or other metabolic processes. In addition, a patient may have a pre-existing condition, such as a trauma or an altered mental status due to drugs or alcohol, all of which can increase his or her susceptibility to the effects of CO₂.

Evaluation and Treatment

Useful laboratory testing includes arterial blood gas, venous co-oximetry for carboxyhemoglobin, chemistry panels, ethanol testing, and radiographs or computed tomography scanning, as indicated. 

Initial management of suspected CO₂ toxicity includes first removing the patient from the source of the gas. Rescuers must exercise caution so as to prevent a mass casualty incident. Once out of the dangerous environment, as long as the patient is conscious and spontaneously breathing, supportive measures are generally all that are necessary. Oxygen should be applied, after which the spontaneously breathing patient without underlying lung disease should rapidly return to normal. If there is marked decrease in mental status or poor respiratory drive despite O₂ administration, intubation with mechanical ventilation may be required. A higher than normal RR will help remove excessive CO₂ if mechanical ventilation is required. If a respiratory acidosis is present, intravenous sodium bicarbonate should be avoided as this may increase the level of serum CO₂. Intravenous fluids and other supportive measures, including treatment for any concurrent conditions, may be indicated.

Dr Schreckengaust is an emergency physician in the department of emergency medicine at Camp Lejune, North Carolina. Dr Lang is an assistant professor in the department of emergency medicine at Eastern Virginia Medical School, Norfolk; and a physician at Emergency Physicians of Tidewater, Norfolk, Virginia. Dr Counselman is the distinguished professor and chairman of the department of emergency medicine at Eastern Virginia Medical School, Norfolk; and a physician at Emergency Physicians of Tidewater, Norfolk, Virginia. He is also the associate editor in chief of EMERGENCY MEDICINE editorial board.

Case

A 58-year-old man presented to the ED via emergency medical services (EMS) with shortness of breath, lightheadedness, and nausea. Upon arrival to the ED, most of his symptoms had resolved. The patient reported that he had taken a 2-hour flight into town the previous day and had spent an uneventful evening at a local hotel. He said that he began experiencing shortness of breath and lightheadedness soon after entering his rental vehicle an hour prior to presentation, explaining that he felt as if he “could not get any air.”

He denied chest pain, leg pain or swelling, abdominal pain, or recent illness. His past medical history was significant only for hypertension, for which he was taking losartan and amlodipine. He had no drug allergies, surgical history, or smoking history. Of note, when the hotel clerk got in the same rental vehicle to move it, he developed symptoms similar to those of the patient. As with the patient, the clerk’s symptoms quickly resolved after he got out of the vehicle.

The patient’s vital signs at examination were: oral temperature, 97.5^oF; pulse, 62 beats/minute; respiratory rate (RR), 18 breaths/minute; and blood pressure, 133/83 mm Hg. Oxygen (O₂) saturation was 100% on room air. He was alert and oriented, in no distress, easily conversational, and without diaphoresis. The lungs were clear to auscultation bilaterally, and there was no calf swelling, tenderness, or palpable cords. The remainder of the physical examination was completely normal.

Ancillary studies revealed a normal chest X-ray. An electrocardiogram demonstrated sinus bradycardia with a rate of 56, but no evidence of ischemia or right heart strain. A complete blood count, troponin I, D-dimer, and creatine phosphokinase (CPK) with MB fraction levels were all within normal limits. A serum chemistry panel was also within normal limits, except for a serum glucose level of 181 mg/dL. Venous co-oximetry showed a carboxyhemoglobin level of 0.0, and methemoglobin level of 0.5 gm% (normal range, 0.4-1.5).

Since both the patient and hotel clerk’s symptoms started when each was in the rental car, the patient was questioned about the vehicle and its contents. The car was a late model rental in good condition per report. The patient informed the treating EP that he worked as a decorative cake salesman and had brought cake samples with him to display at a trade show. He further stated that he had left these samples in the car overnight, packed in dry ice.

Upon learning this information, EMS was contacted and instructed to return to the hotel and rental vehicle. The hotel room was noted to have normal levels of O₂ and carbon monoxide (CO) on measurement. Investigation of the car revealed normal levels of CO, but O₂ levels too low to read on the sensor. The emergency team concluded that the dry ice (the solid form of carbon dioxide [CO₂]), sublimed to CO₂ gas overnight. This displaced the O₂ in the vehicle, resulting in severe hypoxia and the symptoms of both the patient and hotel clerk.

The patient was initially placed on 15 L of O₂ via a nonrebreather mask, but he was switched to 2 L of O₂ via nasal cannula shortly afterward. He was observed for a total of 4 hours after arrival; as he remained symptom-free, he was discharged home. The EP was not able to obtain postdischarge follow up information.

Discussion

Carbon dioxide primarily acts as a simple asphyxiant, but it also dissolves in serum as carbonic acid, resulting in a metabolic acidosis. Carbon dioxide is a prevalent gas that is part of everyday life, from an agent in fire extinguishers and carbonation in beverages to byproducts of cellular metabolism. Similar to CO, it is a colorless and odorless gas. Carbon dioxide is commonly used in the food industry as dry ice to keep items cold. In its solid state, CO₂ can cause severe frostbite with direct contact, similar to a burn. However, when dry ice is warmed and sublimated to a gaseous state, large amounts of CO₂ are generated, and this heavy gas can accumulate and displace air (ie, atmospheric O₂), especially in confined spaces. In low concentrations, gaseous CO₂ appears to have minimal toxicological effects, but at higher concentrations it can cause tachycardia, tachypnea, dyspnea, visual disturbances, arrhythmias, impaired levels of consciousness, and even death.

Carbon dioxide primarily acts as a simple asphyxiant, but it also dissolves in serum as carbonic acid, resulting in a metabolic acidosis. Compensation for this acidosis is accomplished by an increased RR (ie, respiratory alkalosis), which further worsens the intake of CO₂.^1,2

 

The normal concentration of CO₂ in the atmosphere is approximately 0.04% (396 ppm). The Occupational Safety and Health Administration (OSHA) has set a maximum safe exposure level of CO₂ at 0.5% (5,000 ppm) over an 8-hour day.³ Concentrations as low as 1% (10,000 ppm) may cause drowsiness. Exposure to concentrations of 7% to 10% for several minutes to an hour results in headache, tachycardia, dyspnea, and hyperventilation. At levels of 10% to 15%, dizziness, severe muscle twitching, and loss of consciousness can occur after only a few minutes. Death occurs within minutes at concentrations greater than 30%.²

Carbon dioxide also acts as a potent cerebral vasodilator, which may explain symptoms such as headache and dizziness.² The severity of symptoms is dependent on the concentration of CO₂, the length of the exposure, and the underlying health of the patient. Elevated concentrations of CO₂ can occur in areas where there is limited or poor ventilation, such as in a mine (where it is known as blackdamp, stythe, or choke damp),⁴ submarine, grain silo, or a sealed building without mechanical ventilation.

Other Case Presentations

Similar cases as the one presented in this article have been described in the literature. In one such case, following Hurricane Ivan, a 34-year-old-man placed four 25-pound blocks of dry ice wrapped in paper in the front seat of his truck with the windows closed.⁵ After driving less than one quarter of a mile, he developed dyspnea and telephoned for help before losing consciousness. Fortunately, he was found in time and recovered soon after the doors to his truck were opened.⁵

In another case, a 59-year-old man entered a walk-in freezer that contained dry ice wrapped loosely in plastic. He was found inside the freezer 20 minutes later in cardiac arrest; resuscitation efforts were unsuccessful. Investigation of the freezer found the initial O₂ concentration to be 13% (normal level, 20.93%) and an estimated CO₂ level of 40%.⁵

Similarly, a 35-year-old woman was inadvertently locked in a bank vault while storing receipts. In a bid for help, she pulled the fire alarm, which triggered a CO₂-based fire-extinguishing system. The fire department responded and found the woman dead in the vault 30 minutes later. The cause of death was labeled as CO₂ intoxication.⁶

Natural Phenomenon

In addition to the foregoing case examples, there have been documented cases of CO₂ toxicity associated with volcanic eruption and other natural phenomena. One example is the Lake Nyos, Cameroon, West Africa incident in 1986. In this event, a magma pocket underlying the lake saturated the water with CO₂ stored as carbonic acid in the water. When a landslide hit the lake, it caused the carbonic acid stored in the depths of the lake to be upheaved to the surface, where it turned back into CO₂ and was released into the atmosphere. Since CO₂ is heavier than O₂, it displaced the O₂ near the ground, resulting in the suffocation and death of 1,700 people in the surrounding villages.²

Differential Diagnosis

When CO₂ toxicity is suspected, other conditions should be considered as there may be more than one process involved. For example, other causes of coma or dyspnea should be investigated, including trauma, hypoglycemia, CO, methemoglobinemia, or other metabolic processes. In addition, a patient may have a pre-existing condition, such as a trauma or an altered mental status due to drugs or alcohol, all of which can increase his or her susceptibility to the effects of CO₂.

Evaluation and Treatment

Useful laboratory testing includes arterial blood gas, venous co-oximetry for carboxyhemoglobin, chemistry panels, ethanol testing, and radiographs or computed tomography scanning, as indicated. 

Initial management of suspected CO₂ toxicity includes first removing the patient from the source of the gas. Rescuers must exercise caution so as to prevent a mass casualty incident. Once out of the dangerous environment, as long as the patient is conscious and spontaneously breathing, supportive measures are generally all that are necessary. Oxygen should be applied, after which the spontaneously breathing patient without underlying lung disease should rapidly return to normal. If there is marked decrease in mental status or poor respiratory drive despite O₂ administration, intubation with mechanical ventilation may be required. A higher than normal RR will help remove excessive CO₂ if mechanical ventilation is required. If a respiratory acidosis is present, intravenous sodium bicarbonate should be avoided as this may increase the level of serum CO₂. Intravenous fluids and other supportive measures, including treatment for any concurrent conditions, may be indicated.

Dr Schreckengaust is an emergency physician in the department of emergency medicine at Camp Lejune, North Carolina. Dr Lang is an assistant professor in the department of emergency medicine at Eastern Virginia Medical School, Norfolk; and a physician at Emergency Physicians of Tidewater, Norfolk, Virginia. Dr Counselman is the distinguished professor and chairman of the department of emergency medicine at Eastern Virginia Medical School, Norfolk; and a physician at Emergency Physicians of Tidewater, Norfolk, Virginia. He is also the associate editor in chief of EMERGENCY MEDICINE editorial board.

Case

A 23-year-old Hispanic woman with no past medical history is brought to the ED for the second time in one day. On her first presentation, which was for a fever and a headache, meningitis was excluded with normal laboratory tests that included a lumbar puncture. She was administered acetaminophen for fever and pain control, and was discharged with a diagnosis of viral illness. On this second visit, 10 hours after being discharged, she presented because her family noted convulsions that began 3 hours after taking an herbal headache remedy given to her by a naturopath. 

The patient arrived to the ED with a persistent seizure that terminated following administration of 2 mg of lorazepam. Her initial vital signs were: blood pressure, 115/51 mm Hg; heart rate, 121 beats/minute; respiratory rate, 24 breaths/minute; temperature, 97.6^oF. Oxygen (O₂) saturation was 100% with 2 L of O₂ administered via nasal cannula. Her neurological examination was significant for a depressed mental status, pupils that were 6 mm and minimally reactive, clonus, and hyperreflexia. Repeat laboratory evaluation found a leukocytosis of 22.0 x 10³/µL, serum bicarbonate of 9 mEq/L, and an anion gap of 22 with a normal serum lactate. 

What is the differential diagnosis of this patient? 

The history of medicinal plant ingestion raises the possibility of a toxicologic etiology. However, because the patient took the “medication” to treat another disorder, a search for an alternate cause should be performed. The differential diagnosis of a toxin-induced seizure is broad and includes pharmaceuticals (eg, tramadol, antihistamines), which may be surreptitiously added to herbal medication to assure efficacy. Plants associated with seizures include those containing antimuscarinic tropane alkaloids such as Jimsonweed (though a rare side effect from this plant product) or the water hemlock (Cicuta maculata). Contaminants of the plant itself may include pesticides such as organophosphates.

Although unlikely in a 21 year old, with­drawal from benzodiazepines, ethanol, baclofen, or gamma hydroxybutyrate are other possible etiologies. In addition to pharmaceutical and plant-derived causes, carbon monoxide poisoning should be a consideration in any patient with headache and flu-like illness.  

This patient also presented with a constellation of other findings that included hyperreflexia, clonus, tachycardia, and altered mental status. Together these signs are expected in patients with serotonin toxicity (also referred to as serotonin syndrome), neuroleptic malignant syndrome, exogenous thyrotoxicosis, and lithium poisoning.

Case Continuation

The naturopathic practitioner arrived at the ED concerned about the patient, informing the ED team that she had given the patient 2 ounces of ayahuasca tea.

What is ayahuasca? What is the mechanism by which it exerts toxic effects?

Ayahuasca is a plant-derived psychotropic beverage that is used for religious purposes by members of two Brazilian churches—Centro Espírita Beneficente União do Vegetal (UDV) and Santo Daime. The ayahuasca beverage consists of two pharmacologically active compounds that together, but not individually, are psychoactive. The desired active effects for church participants include hallucinations, and vomiting to bring about a “religious purge.”¹  

Ayahuasca is prepared by combining two plants indigenous to the Amazon Basin area: Banisteriopsis caapi and either Psychotria viridis or Diplopterys cabrerana. B caapi contains the β-carboline alkaloids harmine, harmaline, and tetrahydroharmine. These alkaloids act as reversible inhibitors of the monoamine oxidase A (MAO-A) enzyme. The bark and stems of B caapi are boiled along with either P viridis or D cabrerana, both of which contain the potent hallucinogen N-N dimethyltryptamine (DMT).² Normally, DMT is not active orally because it is enzymatically metabolized by MAO-A. However, when taken in the presence of the B caapi-derived MAO-A–inhibiting harmine alkaloids, DMT reaches the systemic circulation and produces its clinical effects.³

N-N dimethyltryptamine is structurally and functionally related to other proserotonergic psychedelics such as lysergic acid diethylamide and psylocibin (from the liberty cap mushroom) (Figure 1). Specifically, DMT is an agonist at the 5-hydroxytryptamine_2A (5-HT_2A) receptor.

What are the clinical findings of serotonin toxicity? 

Serotonin toxicity is a collection of clinical findings that fall under three main categories: autonomic hyperactivity, altered mental status, and muscle rigidity.⁵ The autonomic findings may include tachycardia, hypertension, hyperthermia, shivering, diaphoresis, or mydriasis. Altered mental status ranges from mild agitation and hypervigilance to agitated delirium to obtundation. Other neurological findings may include tremor, myoclonus, hyperreflexia, or seizures. The onset of these signs is rapid, usually occurring within minutes after exposure to one or more serotonergic compounds. Although rare, severe serotonin toxicity may be associated with hypotension and shock, leading to death.⁴ 

 

The diagnosis of serotonin toxicity is based on the history and physical examination of the patient. Diagnostic criteria that have been suggested include the following: (1) a recent addition or increase in a known serotonergic agent; (2) absence of other possible etiologies; (3) no recent increase or addition of a neuroleptic agent (suggesting neuroleptic malignant syndrome); and/or (4) at least 3 of the following symptoms—mental status changes, myoclonus, agitation, hyperreflexia, diaphoresis, shivering, tremor, diarrhea, incoordination, fever⁵ (Figure 2).

The above criteria, though still used by many clinicians, have been criticized by others as not being specific enough for serotonin toxicity, particularly with the heavy emphasis on altered mental status. A later study developed a clinical decision rule that showed ocular clonus, spontaneous clonus, inducible clonus, agitation, diaphoresis, tremor, and hyperreflexia were more accurate in predicting serotonin toxicity.⁶ In summary, it is important for clinicians to be aware that patients with serotonin toxicity may present with a wide range of clinical findings, but the history may be the most important factor. 

How should this patient be managed?

The management of serotonin toxicity is primarily supportive with aggressive control of hyperthermia and autonomic instability. The precipitating xenobiotic agent should be immediately discontinued. In general, treatment with intravenous fluids, cooling measures, benzodiazepines, and a nonspecific 5-HT antagonist such as cyproheptadine should greatly improve the patient’s clinical status. Patients with severe toxicity may require induced paralysis and intubation.⁴ It is not clear in this case if the serotonin hyperactivation was due to the DMT (5-HT_2A is associated with serotonin toxicity) or another serotonergic agent (eg, dextromethorphan from a cough and cold preparation) in combination with the MAO-inhibiting harmine alkaloids.

What is the availability of ayahuasca in the United States? How is it used in its nonherbal form?

...[Ayahuasca] is currently available in the United States and is legal for use by members of the UDV and Santo Daime churches. Many clinicians are becoming increasingly familiar with this herbal preparation since the recreational use of ayahuasca is gaining popularity in the United States. Internet fora with information on how to safely use ayahuasca, such as avoiding aged cheeses, are becoming more prevalent.⁷ A recent article in the New York Times described an ayahuasca gathering in Brooklyn, New York, where participants use the herb in a communal fashion.⁸ This herbal product is also associated with the Hollywood social scene and has received celebrity endorsements.⁸

The National Survey on Drug Use and Health found that the number of people in the United States who have used DMT has gone up almost every year since 2006, from an estimated 688,000 in 2006 to 1,475,000 in 2012.⁹ When used alone (not as ayahuasca), DMT is almost exclusively insufflated as a nasal snuff, bypassing hepatic elimination. It has an onset of around 45 seconds and a duration of 5 to 10 minutes.  Insufflating DMT was historically referred to as a “businessman’s trip” because users were able to have a brief hallucinogenic experience on a lunch break and recover rapidly to perform their normal work.¹⁰

International law declares that DMT is an illegal substance and its importation is banned. However, its use for religious purposes, as is allowed for mescaline found in peyote, remains controversial.⁷ The UDV brought suit in United States federal court to prevent interference with the church’s use of ayahuasca during religious ceremonies based on the Religious Freedom Restoration Act. This act states that the government should not cause substantial imposition on religious practices in the absence of a compelling government interest. The court sided with the UDV, finding that the government had not sufficiently proved the alleged health risks posed by ayahuasca and could not show a substantial risk that the drug would be abused recreationally.¹¹ Thus it is currently available in the United States and is legal for use by members of the UDV and Santo Daime churches.

Ayahuasca is not regulated by the US Food and Drug Administration. Many different types of preparations with different ingredients as well as different concentrations may exist, and clinical variability should be expected. Understanding that ayahuasca is capable of inhibiting MAO is important in order to avoid foods and medications, such as dextromethorphan, that may trigger adverse effects.

Case Conclusion

The patient’s hospital course was complicated by an additional seizure 12 hours after her initial presentation. By 36 hours she was back to her baseline mental status with a normal neurological examination.

Dr Fil is a senior fellow in medical toxicology at North Shore University Hospital, Manhasset, 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 23-year-old Hispanic woman with no past medical history is brought to the ED for the second time in one day. On her first presentation, which was for a fever and a headache, meningitis was excluded with normal laboratory tests that included a lumbar puncture. She was administered acetaminophen for fever and pain control, and was discharged with a diagnosis of viral illness. On this second visit, 10 hours after being discharged, she presented because her family noted convulsions that began 3 hours after taking an herbal headache remedy given to her by a naturopath. 

The patient arrived to the ED with a persistent seizure that terminated following administration of 2 mg of lorazepam. Her initial vital signs were: blood pressure, 115/51 mm Hg; heart rate, 121 beats/minute; respiratory rate, 24 breaths/minute; temperature, 97.6^oF. Oxygen (O₂) saturation was 100% with 2 L of O₂ administered via nasal cannula. Her neurological examination was significant for a depressed mental status, pupils that were 6 mm and minimally reactive, clonus, and hyperreflexia. Repeat laboratory evaluation found a leukocytosis of 22.0 x 10³/µL, serum bicarbonate of 9 mEq/L, and an anion gap of 22 with a normal serum lactate. 

What is the differential diagnosis of this patient? 

The history of medicinal plant ingestion raises the possibility of a toxicologic etiology. However, because the patient took the “medication” to treat another disorder, a search for an alternate cause should be performed. The differential diagnosis of a toxin-induced seizure is broad and includes pharmaceuticals (eg, tramadol, antihistamines), which may be surreptitiously added to herbal medication to assure efficacy. Plants associated with seizures include those containing antimuscarinic tropane alkaloids such as Jimsonweed (though a rare side effect from this plant product) or the water hemlock (Cicuta maculata). Contaminants of the plant itself may include pesticides such as organophosphates.

Although unlikely in a 21 year old, with­drawal from benzodiazepines, ethanol, baclofen, or gamma hydroxybutyrate are other possible etiologies. In addition to pharmaceutical and plant-derived causes, carbon monoxide poisoning should be a consideration in any patient with headache and flu-like illness.  

This patient also presented with a constellation of other findings that included hyperreflexia, clonus, tachycardia, and altered mental status. Together these signs are expected in patients with serotonin toxicity (also referred to as serotonin syndrome), neuroleptic malignant syndrome, exogenous thyrotoxicosis, and lithium poisoning.

Case Continuation

The naturopathic practitioner arrived at the ED concerned about the patient, informing the ED team that she had given the patient 2 ounces of ayahuasca tea.

What is ayahuasca? What is the mechanism by which it exerts toxic effects?

Ayahuasca is a plant-derived psychotropic beverage that is used for religious purposes by members of two Brazilian churches—Centro Espírita Beneficente União do Vegetal (UDV) and Santo Daime. The ayahuasca beverage consists of two pharmacologically active compounds that together, but not individually, are psychoactive. The desired active effects for church participants include hallucinations, and vomiting to bring about a “religious purge.”¹  

Ayahuasca is prepared by combining two plants indigenous to the Amazon Basin area: Banisteriopsis caapi and either Psychotria viridis or Diplopterys cabrerana. B caapi contains the β-carboline alkaloids harmine, harmaline, and tetrahydroharmine. These alkaloids act as reversible inhibitors of the monoamine oxidase A (MAO-A) enzyme. The bark and stems of B caapi are boiled along with either P viridis or D cabrerana, both of which contain the potent hallucinogen N-N dimethyltryptamine (DMT).² Normally, DMT is not active orally because it is enzymatically metabolized by MAO-A. However, when taken in the presence of the B caapi-derived MAO-A–inhibiting harmine alkaloids, DMT reaches the systemic circulation and produces its clinical effects.³

N-N dimethyltryptamine is structurally and functionally related to other proserotonergic psychedelics such as lysergic acid diethylamide and psylocibin (from the liberty cap mushroom) (Figure 1). Specifically, DMT is an agonist at the 5-hydroxytryptamine_2A (5-HT_2A) receptor.

What are the clinical findings of serotonin toxicity? 

Serotonin toxicity is a collection of clinical findings that fall under three main categories: autonomic hyperactivity, altered mental status, and muscle rigidity.⁵ The autonomic findings may include tachycardia, hypertension, hyperthermia, shivering, diaphoresis, or mydriasis. Altered mental status ranges from mild agitation and hypervigilance to agitated delirium to obtundation. Other neurological findings may include tremor, myoclonus, hyperreflexia, or seizures. The onset of these signs is rapid, usually occurring within minutes after exposure to one or more serotonergic compounds. Although rare, severe serotonin toxicity may be associated with hypotension and shock, leading to death.⁴ 

 

The diagnosis of serotonin toxicity is based on the history and physical examination of the patient. Diagnostic criteria that have been suggested include the following: (1) a recent addition or increase in a known serotonergic agent; (2) absence of other possible etiologies; (3) no recent increase or addition of a neuroleptic agent (suggesting neuroleptic malignant syndrome); and/or (4) at least 3 of the following symptoms—mental status changes, myoclonus, agitation, hyperreflexia, diaphoresis, shivering, tremor, diarrhea, incoordination, fever⁵ (Figure 2).

The above criteria, though still used by many clinicians, have been criticized by others as not being specific enough for serotonin toxicity, particularly with the heavy emphasis on altered mental status. A later study developed a clinical decision rule that showed ocular clonus, spontaneous clonus, inducible clonus, agitation, diaphoresis, tremor, and hyperreflexia were more accurate in predicting serotonin toxicity.⁶ In summary, it is important for clinicians to be aware that patients with serotonin toxicity may present with a wide range of clinical findings, but the history may be the most important factor. 

How should this patient be managed?

The management of serotonin toxicity is primarily supportive with aggressive control of hyperthermia and autonomic instability. The precipitating xenobiotic agent should be immediately discontinued. In general, treatment with intravenous fluids, cooling measures, benzodiazepines, and a nonspecific 5-HT antagonist such as cyproheptadine should greatly improve the patient’s clinical status. Patients with severe toxicity may require induced paralysis and intubation.⁴ It is not clear in this case if the serotonin hyperactivation was due to the DMT (5-HT_2A is associated with serotonin toxicity) or another serotonergic agent (eg, dextromethorphan from a cough and cold preparation) in combination with the MAO-inhibiting harmine alkaloids.

What is the availability of ayahuasca in the United States? How is it used in its nonherbal form?

...[Ayahuasca] is currently available in the United States and is legal for use by members of the UDV and Santo Daime churches. Many clinicians are becoming increasingly familiar with this herbal preparation since the recreational use of ayahuasca is gaining popularity in the United States. Internet fora with information on how to safely use ayahuasca, such as avoiding aged cheeses, are becoming more prevalent.⁷ A recent article in the New York Times described an ayahuasca gathering in Brooklyn, New York, where participants use the herb in a communal fashion.⁸ This herbal product is also associated with the Hollywood social scene and has received celebrity endorsements.⁸

The National Survey on Drug Use and Health found that the number of people in the United States who have used DMT has gone up almost every year since 2006, from an estimated 688,000 in 2006 to 1,475,000 in 2012.⁹ When used alone (not as ayahuasca), DMT is almost exclusively insufflated as a nasal snuff, bypassing hepatic elimination. It has an onset of around 45 seconds and a duration of 5 to 10 minutes.  Insufflating DMT was historically referred to as a “businessman’s trip” because users were able to have a brief hallucinogenic experience on a lunch break and recover rapidly to perform their normal work.¹⁰

International law declares that DMT is an illegal substance and its importation is banned. However, its use for religious purposes, as is allowed for mescaline found in peyote, remains controversial.⁷ The UDV brought suit in United States federal court to prevent interference with the church’s use of ayahuasca during religious ceremonies based on the Religious Freedom Restoration Act. This act states that the government should not cause substantial imposition on religious practices in the absence of a compelling government interest. The court sided with the UDV, finding that the government had not sufficiently proved the alleged health risks posed by ayahuasca and could not show a substantial risk that the drug would be abused recreationally.¹¹ Thus it is currently available in the United States and is legal for use by members of the UDV and Santo Daime churches.

Ayahuasca is not regulated by the US Food and Drug Administration. Many different types of preparations with different ingredients as well as different concentrations may exist, and clinical variability should be expected. Understanding that ayahuasca is capable of inhibiting MAO is important in order to avoid foods and medications, such as dextromethorphan, that may trigger adverse effects.

Case Conclusion

The patient’s hospital course was complicated by an additional seizure 12 hours after her initial presentation. By 36 hours she was back to her baseline mental status with a normal neurological examination.

Dr Fil is a senior fellow in medical toxicology at North Shore University Hospital, Manhasset, 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 23-year-old Hispanic woman with no past medical history is brought to the ED for the second time in one day. On her first presentation, which was for a fever and a headache, meningitis was excluded with normal laboratory tests that included a lumbar puncture. She was administered acetaminophen for fever and pain control, and was discharged with a diagnosis of viral illness. On this second visit, 10 hours after being discharged, she presented because her family noted convulsions that began 3 hours after taking an herbal headache remedy given to her by a naturopath. 

The patient arrived to the ED with a persistent seizure that terminated following administration of 2 mg of lorazepam. Her initial vital signs were: blood pressure, 115/51 mm Hg; heart rate, 121 beats/minute; respiratory rate, 24 breaths/minute; temperature, 97.6^oF. Oxygen (O₂) saturation was 100% with 2 L of O₂ administered via nasal cannula. Her neurological examination was significant for a depressed mental status, pupils that were 6 mm and minimally reactive, clonus, and hyperreflexia. Repeat laboratory evaluation found a leukocytosis of 22.0 x 10³/µL, serum bicarbonate of 9 mEq/L, and an anion gap of 22 with a normal serum lactate. 

What is the differential diagnosis of this patient? 

The history of medicinal plant ingestion raises the possibility of a toxicologic etiology. However, because the patient took the “medication” to treat another disorder, a search for an alternate cause should be performed. The differential diagnosis of a toxin-induced seizure is broad and includes pharmaceuticals (eg, tramadol, antihistamines), which may be surreptitiously added to herbal medication to assure efficacy. Plants associated with seizures include those containing antimuscarinic tropane alkaloids such as Jimsonweed (though a rare side effect from this plant product) or the water hemlock (Cicuta maculata). Contaminants of the plant itself may include pesticides such as organophosphates.

Although unlikely in a 21 year old, with­drawal from benzodiazepines, ethanol, baclofen, or gamma hydroxybutyrate are other possible etiologies. In addition to pharmaceutical and plant-derived causes, carbon monoxide poisoning should be a consideration in any patient with headache and flu-like illness.  

This patient also presented with a constellation of other findings that included hyperreflexia, clonus, tachycardia, and altered mental status. Together these signs are expected in patients with serotonin toxicity (also referred to as serotonin syndrome), neuroleptic malignant syndrome, exogenous thyrotoxicosis, and lithium poisoning.

Case Continuation

The naturopathic practitioner arrived at the ED concerned about the patient, informing the ED team that she had given the patient 2 ounces of ayahuasca tea.

What is ayahuasca? What is the mechanism by which it exerts toxic effects?

Ayahuasca is a plant-derived psychotropic beverage that is used for religious purposes by members of two Brazilian churches—Centro Espírita Beneficente União do Vegetal (UDV) and Santo Daime. The ayahuasca beverage consists of two pharmacologically active compounds that together, but not individually, are psychoactive. The desired active effects for church participants include hallucinations, and vomiting to bring about a “religious purge.”¹  

Ayahuasca is prepared by combining two plants indigenous to the Amazon Basin area: Banisteriopsis caapi and either Psychotria viridis or Diplopterys cabrerana. B caapi contains the β-carboline alkaloids harmine, harmaline, and tetrahydroharmine. These alkaloids act as reversible inhibitors of the monoamine oxidase A (MAO-A) enzyme. The bark and stems of B caapi are boiled along with either P viridis or D cabrerana, both of which contain the potent hallucinogen N-N dimethyltryptamine (DMT).² Normally, DMT is not active orally because it is enzymatically metabolized by MAO-A. However, when taken in the presence of the B caapi-derived MAO-A–inhibiting harmine alkaloids, DMT reaches the systemic circulation and produces its clinical effects.³

N-N dimethyltryptamine is structurally and functionally related to other proserotonergic psychedelics such as lysergic acid diethylamide and psylocibin (from the liberty cap mushroom) (Figure 1). Specifically, DMT is an agonist at the 5-hydroxytryptamine_2A (5-HT_2A) receptor.

What are the clinical findings of serotonin toxicity? 

Serotonin toxicity is a collection of clinical findings that fall under three main categories: autonomic hyperactivity, altered mental status, and muscle rigidity.⁵ The autonomic findings may include tachycardia, hypertension, hyperthermia, shivering, diaphoresis, or mydriasis. Altered mental status ranges from mild agitation and hypervigilance to agitated delirium to obtundation. Other neurological findings may include tremor, myoclonus, hyperreflexia, or seizures. The onset of these signs is rapid, usually occurring within minutes after exposure to one or more serotonergic compounds. Although rare, severe serotonin toxicity may be associated with hypotension and shock, leading to death.⁴ 

 

The diagnosis of serotonin toxicity is based on the history and physical examination of the patient. Diagnostic criteria that have been suggested include the following: (1) a recent addition or increase in a known serotonergic agent; (2) absence of other possible etiologies; (3) no recent increase or addition of a neuroleptic agent (suggesting neuroleptic malignant syndrome); and/or (4) at least 3 of the following symptoms—mental status changes, myoclonus, agitation, hyperreflexia, diaphoresis, shivering, tremor, diarrhea, incoordination, fever⁵ (Figure 2).

The above criteria, though still used by many clinicians, have been criticized by others as not being specific enough for serotonin toxicity, particularly with the heavy emphasis on altered mental status. A later study developed a clinical decision rule that showed ocular clonus, spontaneous clonus, inducible clonus, agitation, diaphoresis, tremor, and hyperreflexia were more accurate in predicting serotonin toxicity.⁶ In summary, it is important for clinicians to be aware that patients with serotonin toxicity may present with a wide range of clinical findings, but the history may be the most important factor. 

How should this patient be managed?

The management of serotonin toxicity is primarily supportive with aggressive control of hyperthermia and autonomic instability. The precipitating xenobiotic agent should be immediately discontinued. In general, treatment with intravenous fluids, cooling measures, benzodiazepines, and a nonspecific 5-HT antagonist such as cyproheptadine should greatly improve the patient’s clinical status. Patients with severe toxicity may require induced paralysis and intubation.⁴ It is not clear in this case if the serotonin hyperactivation was due to the DMT (5-HT_2A is associated with serotonin toxicity) or another serotonergic agent (eg, dextromethorphan from a cough and cold preparation) in combination with the MAO-inhibiting harmine alkaloids.

What is the availability of ayahuasca in the United States? How is it used in its nonherbal form?

...[Ayahuasca] is currently available in the United States and is legal for use by members of the UDV and Santo Daime churches. Many clinicians are becoming increasingly familiar with this herbal preparation since the recreational use of ayahuasca is gaining popularity in the United States. Internet fora with information on how to safely use ayahuasca, such as avoiding aged cheeses, are becoming more prevalent.⁷ A recent article in the New York Times described an ayahuasca gathering in Brooklyn, New York, where participants use the herb in a communal fashion.⁸ This herbal product is also associated with the Hollywood social scene and has received celebrity endorsements.⁸

The National Survey on Drug Use and Health found that the number of people in the United States who have used DMT has gone up almost every year since 2006, from an estimated 688,000 in 2006 to 1,475,000 in 2012.⁹ When used alone (not as ayahuasca), DMT is almost exclusively insufflated as a nasal snuff, bypassing hepatic elimination. It has an onset of around 45 seconds and a duration of 5 to 10 minutes.  Insufflating DMT was historically referred to as a “businessman’s trip” because users were able to have a brief hallucinogenic experience on a lunch break and recover rapidly to perform their normal work.¹⁰

International law declares that DMT is an illegal substance and its importation is banned. However, its use for religious purposes, as is allowed for mescaline found in peyote, remains controversial.⁷ The UDV brought suit in United States federal court to prevent interference with the church’s use of ayahuasca during religious ceremonies based on the Religious Freedom Restoration Act. This act states that the government should not cause substantial imposition on religious practices in the absence of a compelling government interest. The court sided with the UDV, finding that the government had not sufficiently proved the alleged health risks posed by ayahuasca and could not show a substantial risk that the drug would be abused recreationally.¹¹ Thus it is currently available in the United States and is legal for use by members of the UDV and Santo Daime churches.

Ayahuasca is not regulated by the US Food and Drug Administration. Many different types of preparations with different ingredients as well as different concentrations may exist, and clinical variability should be expected. Understanding that ayahuasca is capable of inhibiting MAO is important in order to avoid foods and medications, such as dextromethorphan, that may trigger adverse effects.

Case Conclusion

The patient’s hospital course was complicated by an additional seizure 12 hours after her initial presentation. By 36 hours she was back to her baseline mental status with a normal neurological examination.

Dr Fil is a senior fellow in medical toxicology at North Shore University Hospital, Manhasset, 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.