Inspection of the fingernails and toenails should be part of a complete physical examination. A basic understanding of nail anatomy and recognition of several basic types of nail lines and bands allow the clinician to properly diagnose and treat the nail disease, to recognize possible underlying systemic diseases, and to know when to refer the patient to a dermatologist for specialized evaluation and biopsy.

In this review, we delineate the three basic types of nail lines­—white lines (leukonychia striata), brown-black lines (longitudinal melanonychia), and red lines (longitudinal erythronychia)—and the differential diagnosis for each type. We also discuss grooves in the nail plate, or Beau lines.

BASIC NAIL ANATOMY

A fundamental understanding of the anatomy of the nail unit is necessary to understand the origin of nail diseases and underlying pathologic conditions.

The nail unit includes the nail matrix, the lunula, the nail fold, the nail plate, and the nail bed. The nail matrix extends from under the proximal nail fold to the half-moon-shaped area (ie, the lunula) and is responsible for nail plate production. The nail bed lies under the nail plate and on top of the distal phalanx and extends from the lunula to just proximal to the free edge of the nail; its rich blood supply gives it its reddish color.

Nails grow slowly, and this should be kept in mind during the examination. Regrowth of a fingernail takes at least 6 months, and regrowth of a toenail may take 12 to 18 months. Therefore, a defect in the nail plate may reveal an injury that occurred—or a condition that began—several months before.¹

NAIL EXAMINATION ESSENTIALS

A complete examination includes all 20 nail units and the periungual skin. Patients should be instructed to remove nail polish from all nails, as it may camouflage dystrophy or disease of the nail. Photography and careful measurement help document changes over time.

LEUKONYCHIA STRIATA: WHITE NAIL LINES

White nail lines or leukonychia is classified as true or apparent, depending on whether the origin is in the nail matrix or the nail bed.

In true leukonychia, there is abnormal keratinization of the underlying nail matrix, resulting in parakeratosis within the nail plate and an opaque appearance on examination.² The white discoloration is unaffected by pressure, and the opacity moves distally as the nail grows out, which can be documented by serial photography on subsequent visits.

Apparent leukonychia involves abnormal nail bed vasculature, which changes the translucency of the nail plate. The whiteness disappears with pressure, is unaffected by nail growth, and will likely show no change on later visits with serial photography.³

True leukonychia

Leukonychia striata, a subtype of true leuko­nychia, is characterized by transverse or longitudinal bands. It is most often associated with microtrauma, such as from a manicure.⁴ Lines due to trauma are typically more apparent in the central part of the nail plate; they spare the lateral portion and lie parallel to the edge of the proximal nail fold.⁵

Onychomycosis. White longitudinal bands may also be seen in onychomycosis, a fungal infection of the nail accounting for up to 50% of all cases of nail disease. The infection may present as irregular dense longitudinal white or yellowish bands or “spikes” on the nail plate with associated hyperkeratosis, known as a dermatophytoma (Figure 1).

If a fungal infection is suspected, a potassium hydroxide stain can be performed on the subungual debris, which is then examined with direct microscopy.⁶ Alternatively, the physician can send a nail plate clipping in a 10% buffered formalin container with a request for a fungal stain such as periodic acid-Schiff.⁷ Microscopic examination of a dermatophytoma shows a dense mass of dermatophyte hyphae, otherwise known as a fungal abscess.⁸

The physician can play an important role in diagnosis because clinical findings suggestive of a dermatophytoma are associated with a poor response to antifungal therapy.⁹

Inherited diseases. White longitudinal bands are also an important clue to the rare autosomal dominant genodermatoses Hailey-Hailey disease (from mutations of the ATP2A2 gene) and Darier disease (from mutations of the ATP2C1 gene). Patients with Hailey-Hailey disease may have nails with multiple parallel longitudinal white stripes of variable width originating in the lunula and most prominent on the thumbs.^10–12 These patients also have recurrent vesicular eruptions in flexural skin areas, such as the groin, axilla, neck, and periumbilical area causing significant morbidity.

Patients with Darier disease may have nails with alternating red and white longitudinal streaks, described as “candy-cane,”¹³ as well as wedge-shaped distal subungual keratosis accompanied by flat keratotic papules on the proximal nail fold.¹⁴ These nail changes are reported in 92% to 95% of patients with Darier disease.^15,16 Patients typically have skin findings characterized by keratotic papules and plaques predominantly in seborrheic areas and palmoplantar pits, as well as secondary infections and malodor causing significant morbidity.¹⁵ Therefore, knowing the characteristic nail findings in these diseases may lead to more rapid diagnosis and treatment.

Mees lines. Leukonychia striata can present as transverse white lines, commonly known as Mees lines. They are 1- to 2-mm wide horizontal parallel white bands that span the width of the nail plate, usually affecting all fingernails.¹⁷ They are not a common finding and are most often associated with arsenic poisoning. They can also be used to identify the time of poisoning, since they tend to appear 2 months after the initial insult.

Mees lines are also associated with acute systemic stresses, such as acute renal failure, heart failure, ulcerative colitis, breast cancer, infections such as measles and tuberculosis, and systemic lupus erythematosus, and with exposure to toxic metals such as thallium.³

Apparent leukonychia

Apparent leukonychia can alert the physician to systemic diseases, infections, drug side effects, and nutrient deficiencies. Specific nail findings include Muehrcke lines, “half-and-half” nails, and Terry nails.

Muehrcke lines are paired white transverse bands that span the width of the nail bed and run parallel to the distal lunula. They were first described in the fingernails of patients with severe hypoalbuminemia, some of whom also had nephrotic syndrome, which resolved with normalization of the serum albumin level. Muehrcke lines have since been reported in patients with liver disease, malnutrition, chemotherapy, organ transplant, human immunodeficiency virus (HIV) infection, and acquired immunodeficiency syndrome.^3,18 They are associated with periods of metabolic stress, ie, when the body’s capacity to synthesize proteins is diminished.¹⁹

Half-and-half nails, or Lindsay nails, are characterized by a white band proximally, a pink or red-brown band distally, and a sharp demarcation between the two (Figure 2). They were originally described in association with chronic renal disease,²⁰ and surprisingly, they resolve with kidney transplant but not with hemodialysis treatment or improvement in hemoglobin or albumin levels.^21–23 Half-and-half nails have been reported with Kawasaki disease, hepatic cirrhosis, Crohn disease, zinc deficiency, chemotherapy, Behçet disease, and pellagra.^3,24,25 They should be distinguished from Terry nails, which are characterized by leukonychia involving more than 80% of the total nail length.²⁶

Terry nails were originally reported in association with hepatic cirrhosis, usually secondary to alcoholism²⁷ but have since been found with heart failure, type 2 diabetes mellitus, pulmonary tuberculosis, reactive arthritis, older age, Hansen disease, and peripheral vascular disease.^3,26,28,29

LONGITUDINAL MELANONYCHIA: VERTICAL BROWN-BLACK NAIL LINES

Longitudinal melanonychia is the presence of black-brown vertical lines in the nail plate. They have a variety of causes, including blood from trauma; bacterial, fungal, or HIV infection; drug therapy (eg, from minocycline); endocrine disorders (Addison disease); exogenous pigmentation; or excess melanin production within the nail matrix.^30–32 They may also be a sign of a benign condition such as benign melanocytic activation, lentigines, or nevi, or a malignant condition such as melanoma (Figure 3).^33,34

When to suspect melanoma and refer

Although melanoma is less commonly associated with brown-black vertical nail lines, awareness of melanoma-associated longitudinal melanonychia reduces the likelihood of delayed diagnosis and improves patient outcomes.³⁵ Also, it is important to remember that although nail melanoma is more common in the 5th and 6th decades of life, it can occur at any age, even in children.³⁶

Findings that raise suspicion of nail melanoma (Table 1)^33,37 and that should prompt referral to a dermatologist who specializes in nails include the following:

• A personal or family history of melanoma
• Involvement of a “high-risk” digit (thumb, index finger, great toe),^30,31,38 although nail melanoma can occur in any digit
• Any new vertical brown-black nail pigmentation in a fair-skinned patient
• Only one nail affected: involvement of more than one nail is common in people with darker skin, and nearly all patients with darker skin exhibit longitudinal melanonychia by age 50³¹
• Changes in the band such as darkening, widening, and bleeding
• A bandwidth greater than 6 mm³³
• A band that is wider proximally than distally³⁴
• Nonuniform color of the line
• Indistinct lateral borders
• Associated with pigmentation of the nail fold (the Hutchinson sign, representing subungual melanoma),^31,39 nail plate dystrophy, bleeding, or ulceration.³³

While these features may help distinguish benign from malignant causes of longitudinal melanonychia, the clinical examination alone may not provide a definitive diagnosis. Delayed diagnosis of nail melanoma carries a high mortality rate; the internist can promote early diagnosis by recognizing the risk factors and clinical signs and referring the patient to a dermatologist for further evaluation with nail biopsy.

LONGITUDINAL ERYTHRONYCHIA: VERTICAL RED NAIL LINES

Longitudinal erythronychia—the presence of one or more linear red bands in the nail unit—can be localized (involving only one nail) or polydactylous (involving more than one nail). The localized form is usually due to a neoplastic process, whereas involvement of more than one nail may indicate an underlying regional or systemic disease.¹³Table 2 lists indications for referral to a nail specialist.

General features on examination

Clinical examination reveals one or more linear, pink-red streaks extending from the proximal nail fold to the distal free edge of the nail plate (Figure 4). The width of the band typically ranges from less than 1 mm to 3 mm.⁴⁰ Other features may include splinter hemorrhages within a red band, a semitransparent distal matrix, distal V-shaped chipping, splitting, onycholysis of the nail plate, and reactive distal nail bed and hyponychial hyperkeratosis. These features can be visible to the naked eye but may be better viewed with a magnifying glass, a 7× loupe, or a dermatoscope.¹³

Localized longitudinal erythronychia is usually seen in middle-aged individuals and is most commonly found on the thumbnail, followed by the index finger.^41,42 The condition may be asymptomatic, but the patient may present with pain or with concern that the split end of the nail catches on fabrics or small objects.⁴²

Glomus tumor

Intense, pulsatile pain with sensitivity to cold and tenderness to palpation is highly suggestive of glomus tumor,⁴³ a benign neoplasm that originates from a neuromyoarterial glomus body. Glomus bodies are located throughout the body but are more highly concentrated in the fingertips, especially beneath the nails, and they regulate skin circulation. Therefore, the nail unit is the most common site for glomus tumor.^44,45 A characteristic feature of subungual glomus tumor is demonstration of tenderness after pin-point palpation of the suspected tumor (positive Love sign).⁴⁵ While it is typical for glomus tumor to affect only one nail, multiple tumors are associated with neurofibromatosis type 1.⁴⁶ Confirmation of this diagnosis requires referral to a dermatologist.

Other causes of localized red nail lines

Onychopapilloma, a benign idiopathic tumor, is the most common cause of localized longitudinal erythronychia. Unlike glomus tumor, it is usually asymptomatic.^42,47 Less common benign conditions are warts, warty dyskeratoma, benign vascular proliferation, a solitary lesion of lichen planus, hemiplegia, and postsurgical scarring of the nail matrix. In some cases, the lines are idiopathic.^42,43

Malignant diseases that can present as localized longitudinal erythronychia include invasive squamous cell carcinoma, squamous cell carcinoma in situ (Bowen disease), and, less frequently, amelanotic melanoma in situ, malignant melanoma, and basal cell carcinoma.⁴² Squamous cell carcinoma in situ most commonly presents in the 5th decade of life and is the malignancy most commonly associated with localized longitudinal erythronychia. Clinically, there is also often nail dystrophy, such as distal subungual keratosis or onycholysis.⁴³

Patients with asymptomatic, stable localized longitudinal erythronychia may be followed closely with photography and measurements. However, any new lesion or a change in an existing lesion should prompt referral to a dermatologist for biopsy.¹³

Red streaks on more than one nail

Polydactylous longitudinal erythronychia usually presents in adults as red streaks on multiple nails and, depending on the presence or absence of symptoms (eg, pain, splitting), may be the patient’s chief complaint or an incidental finding noted by the astute clinician. Often, it is associated with systemic disease, most commonly lichen planus or Darier disease.

Lichen planus is a papulosquamous skin disease with nail involvement in 10% of patients and permanent nail dystrophy in 4%. Common nail findings include thinning, longitudinal ridging, and fissuring, as well as scarring of the nail matrix resulting in pterygium. Linear red streaks may accompany these more typical nail findings.¹³ Patients with Darier disease present with alternating red and white linear bands on multiple nails as in leukonychia striata.

Less frequently, polydactylous longitudinal erythronychia is associated with primary and systemic amyloidosis, hemiplegia, graft-vs-host disease, acantholytic epidermolysis bullosa, acantholytic dyskeratotic epidermal nevus, acrokeratosis verruciformis of Hopf, or pseudobulbar syndrome, or is idiopathic.^13,42,48 Therefore, the physician evaluating a patient with these nail findings should focus on a workup for regional or systemic disease or refer the patient to a dermatologist who specializes in nails.

BEAU LINES

Beau lines are a common finding in clinical practice. They are not true lines, but transverse grooves in the nail plate that arise from the temporary suppression of nail growth within the nail matrix that can occur during periods of acute or chronic stress or systemic illness (Figure 5).⁴⁹

The precipitating event may be local trauma or paronychia, chemotherapeutic agents cytotoxic to the nail matrix, or the abrupt onset of systemic disease.^18,50 The grooves have also been associated with rheumatic fever, malaria, pemphigus, Raynaud disease, and myocardial infarction, as well as following deep-sea dives.^51–53 The distance of a Beau line from the proximal nail fold can provide an estimate of the time of the acute stress, based on an average growth rate of 3 mm per month for fingernails and 1 mm per month for toenails.⁴⁹

Inspection of the fingernails and toenails should be part of a complete physical examination. A basic understanding of nail anatomy and recognition of several basic types of nail lines and bands allow the clinician to properly diagnose and treat the nail disease, to recognize possible underlying systemic diseases, and to know when to refer the patient to a dermatologist for specialized evaluation and biopsy.

In this review, we delineate the three basic types of nail lines­—white lines (leukonychia striata), brown-black lines (longitudinal melanonychia), and red lines (longitudinal erythronychia)—and the differential diagnosis for each type. We also discuss grooves in the nail plate, or Beau lines.

BASIC NAIL ANATOMY

A fundamental understanding of the anatomy of the nail unit is necessary to understand the origin of nail diseases and underlying pathologic conditions.

The nail unit includes the nail matrix, the lunula, the nail fold, the nail plate, and the nail bed. The nail matrix extends from under the proximal nail fold to the half-moon-shaped area (ie, the lunula) and is responsible for nail plate production. The nail bed lies under the nail plate and on top of the distal phalanx and extends from the lunula to just proximal to the free edge of the nail; its rich blood supply gives it its reddish color.

Nails grow slowly, and this should be kept in mind during the examination. Regrowth of a fingernail takes at least 6 months, and regrowth of a toenail may take 12 to 18 months. Therefore, a defect in the nail plate may reveal an injury that occurred—or a condition that began—several months before.¹

NAIL EXAMINATION ESSENTIALS

A complete examination includes all 20 nail units and the periungual skin. Patients should be instructed to remove nail polish from all nails, as it may camouflage dystrophy or disease of the nail. Photography and careful measurement help document changes over time.

LEUKONYCHIA STRIATA: WHITE NAIL LINES

White nail lines or leukonychia is classified as true or apparent, depending on whether the origin is in the nail matrix or the nail bed.

In true leukonychia, there is abnormal keratinization of the underlying nail matrix, resulting in parakeratosis within the nail plate and an opaque appearance on examination.² The white discoloration is unaffected by pressure, and the opacity moves distally as the nail grows out, which can be documented by serial photography on subsequent visits.

Apparent leukonychia involves abnormal nail bed vasculature, which changes the translucency of the nail plate. The whiteness disappears with pressure, is unaffected by nail growth, and will likely show no change on later visits with serial photography.³

True leukonychia

Leukonychia striata, a subtype of true leuko­nychia, is characterized by transverse or longitudinal bands. It is most often associated with microtrauma, such as from a manicure.⁴ Lines due to trauma are typically more apparent in the central part of the nail plate; they spare the lateral portion and lie parallel to the edge of the proximal nail fold.⁵

Onychomycosis. White longitudinal bands may also be seen in onychomycosis, a fungal infection of the nail accounting for up to 50% of all cases of nail disease. The infection may present as irregular dense longitudinal white or yellowish bands or “spikes” on the nail plate with associated hyperkeratosis, known as a dermatophytoma (Figure 1).

If a fungal infection is suspected, a potassium hydroxide stain can be performed on the subungual debris, which is then examined with direct microscopy.⁶ Alternatively, the physician can send a nail plate clipping in a 10% buffered formalin container with a request for a fungal stain such as periodic acid-Schiff.⁷ Microscopic examination of a dermatophytoma shows a dense mass of dermatophyte hyphae, otherwise known as a fungal abscess.⁸

The physician can play an important role in diagnosis because clinical findings suggestive of a dermatophytoma are associated with a poor response to antifungal therapy.⁹

Inherited diseases. White longitudinal bands are also an important clue to the rare autosomal dominant genodermatoses Hailey-Hailey disease (from mutations of the ATP2A2 gene) and Darier disease (from mutations of the ATP2C1 gene). Patients with Hailey-Hailey disease may have nails with multiple parallel longitudinal white stripes of variable width originating in the lunula and most prominent on the thumbs.^10–12 These patients also have recurrent vesicular eruptions in flexural skin areas, such as the groin, axilla, neck, and periumbilical area causing significant morbidity.

Patients with Darier disease may have nails with alternating red and white longitudinal streaks, described as “candy-cane,”¹³ as well as wedge-shaped distal subungual keratosis accompanied by flat keratotic papules on the proximal nail fold.¹⁴ These nail changes are reported in 92% to 95% of patients with Darier disease.^15,16 Patients typically have skin findings characterized by keratotic papules and plaques predominantly in seborrheic areas and palmoplantar pits, as well as secondary infections and malodor causing significant morbidity.¹⁵ Therefore, knowing the characteristic nail findings in these diseases may lead to more rapid diagnosis and treatment.

Mees lines. Leukonychia striata can present as transverse white lines, commonly known as Mees lines. They are 1- to 2-mm wide horizontal parallel white bands that span the width of the nail plate, usually affecting all fingernails.¹⁷ They are not a common finding and are most often associated with arsenic poisoning. They can also be used to identify the time of poisoning, since they tend to appear 2 months after the initial insult.

Mees lines are also associated with acute systemic stresses, such as acute renal failure, heart failure, ulcerative colitis, breast cancer, infections such as measles and tuberculosis, and systemic lupus erythematosus, and with exposure to toxic metals such as thallium.³

Apparent leukonychia

Apparent leukonychia can alert the physician to systemic diseases, infections, drug side effects, and nutrient deficiencies. Specific nail findings include Muehrcke lines, “half-and-half” nails, and Terry nails.

Muehrcke lines are paired white transverse bands that span the width of the nail bed and run parallel to the distal lunula. They were first described in the fingernails of patients with severe hypoalbuminemia, some of whom also had nephrotic syndrome, which resolved with normalization of the serum albumin level. Muehrcke lines have since been reported in patients with liver disease, malnutrition, chemotherapy, organ transplant, human immunodeficiency virus (HIV) infection, and acquired immunodeficiency syndrome.^3,18 They are associated with periods of metabolic stress, ie, when the body’s capacity to synthesize proteins is diminished.¹⁹

Half-and-half nails, or Lindsay nails, are characterized by a white band proximally, a pink or red-brown band distally, and a sharp demarcation between the two (Figure 2). They were originally described in association with chronic renal disease,²⁰ and surprisingly, they resolve with kidney transplant but not with hemodialysis treatment or improvement in hemoglobin or albumin levels.^21–23 Half-and-half nails have been reported with Kawasaki disease, hepatic cirrhosis, Crohn disease, zinc deficiency, chemotherapy, Behçet disease, and pellagra.^3,24,25 They should be distinguished from Terry nails, which are characterized by leukonychia involving more than 80% of the total nail length.²⁶

Terry nails were originally reported in association with hepatic cirrhosis, usually secondary to alcoholism²⁷ but have since been found with heart failure, type 2 diabetes mellitus, pulmonary tuberculosis, reactive arthritis, older age, Hansen disease, and peripheral vascular disease.^3,26,28,29

LONGITUDINAL MELANONYCHIA: VERTICAL BROWN-BLACK NAIL LINES

Longitudinal melanonychia is the presence of black-brown vertical lines in the nail plate. They have a variety of causes, including blood from trauma; bacterial, fungal, or HIV infection; drug therapy (eg, from minocycline); endocrine disorders (Addison disease); exogenous pigmentation; or excess melanin production within the nail matrix.^30–32 They may also be a sign of a benign condition such as benign melanocytic activation, lentigines, or nevi, or a malignant condition such as melanoma (Figure 3).^33,34

When to suspect melanoma and refer

Although melanoma is less commonly associated with brown-black vertical nail lines, awareness of melanoma-associated longitudinal melanonychia reduces the likelihood of delayed diagnosis and improves patient outcomes.³⁵ Also, it is important to remember that although nail melanoma is more common in the 5th and 6th decades of life, it can occur at any age, even in children.³⁶

Findings that raise suspicion of nail melanoma (Table 1)^33,37 and that should prompt referral to a dermatologist who specializes in nails include the following:

• A personal or family history of melanoma
• Involvement of a “high-risk” digit (thumb, index finger, great toe),^30,31,38 although nail melanoma can occur in any digit
• Any new vertical brown-black nail pigmentation in a fair-skinned patient
• Only one nail affected: involvement of more than one nail is common in people with darker skin, and nearly all patients with darker skin exhibit longitudinal melanonychia by age 50³¹
• Changes in the band such as darkening, widening, and bleeding
• A bandwidth greater than 6 mm³³
• A band that is wider proximally than distally³⁴
• Nonuniform color of the line
• Indistinct lateral borders
• Associated with pigmentation of the nail fold (the Hutchinson sign, representing subungual melanoma),^31,39 nail plate dystrophy, bleeding, or ulceration.³³

While these features may help distinguish benign from malignant causes of longitudinal melanonychia, the clinical examination alone may not provide a definitive diagnosis. Delayed diagnosis of nail melanoma carries a high mortality rate; the internist can promote early diagnosis by recognizing the risk factors and clinical signs and referring the patient to a dermatologist for further evaluation with nail biopsy.

LONGITUDINAL ERYTHRONYCHIA: VERTICAL RED NAIL LINES

Longitudinal erythronychia—the presence of one or more linear red bands in the nail unit—can be localized (involving only one nail) or polydactylous (involving more than one nail). The localized form is usually due to a neoplastic process, whereas involvement of more than one nail may indicate an underlying regional or systemic disease.¹³Table 2 lists indications for referral to a nail specialist.

General features on examination

Clinical examination reveals one or more linear, pink-red streaks extending from the proximal nail fold to the distal free edge of the nail plate (Figure 4). The width of the band typically ranges from less than 1 mm to 3 mm.⁴⁰ Other features may include splinter hemorrhages within a red band, a semitransparent distal matrix, distal V-shaped chipping, splitting, onycholysis of the nail plate, and reactive distal nail bed and hyponychial hyperkeratosis. These features can be visible to the naked eye but may be better viewed with a magnifying glass, a 7× loupe, or a dermatoscope.¹³

Localized longitudinal erythronychia is usually seen in middle-aged individuals and is most commonly found on the thumbnail, followed by the index finger.^41,42 The condition may be asymptomatic, but the patient may present with pain or with concern that the split end of the nail catches on fabrics or small objects.⁴²

Glomus tumor

Intense, pulsatile pain with sensitivity to cold and tenderness to palpation is highly suggestive of glomus tumor,⁴³ a benign neoplasm that originates from a neuromyoarterial glomus body. Glomus bodies are located throughout the body but are more highly concentrated in the fingertips, especially beneath the nails, and they regulate skin circulation. Therefore, the nail unit is the most common site for glomus tumor.^44,45 A characteristic feature of subungual glomus tumor is demonstration of tenderness after pin-point palpation of the suspected tumor (positive Love sign).⁴⁵ While it is typical for glomus tumor to affect only one nail, multiple tumors are associated with neurofibromatosis type 1.⁴⁶ Confirmation of this diagnosis requires referral to a dermatologist.

Other causes of localized red nail lines

Onychopapilloma, a benign idiopathic tumor, is the most common cause of localized longitudinal erythronychia. Unlike glomus tumor, it is usually asymptomatic.^42,47 Less common benign conditions are warts, warty dyskeratoma, benign vascular proliferation, a solitary lesion of lichen planus, hemiplegia, and postsurgical scarring of the nail matrix. In some cases, the lines are idiopathic.^42,43

Malignant diseases that can present as localized longitudinal erythronychia include invasive squamous cell carcinoma, squamous cell carcinoma in situ (Bowen disease), and, less frequently, amelanotic melanoma in situ, malignant melanoma, and basal cell carcinoma.⁴² Squamous cell carcinoma in situ most commonly presents in the 5th decade of life and is the malignancy most commonly associated with localized longitudinal erythronychia. Clinically, there is also often nail dystrophy, such as distal subungual keratosis or onycholysis.⁴³

Patients with asymptomatic, stable localized longitudinal erythronychia may be followed closely with photography and measurements. However, any new lesion or a change in an existing lesion should prompt referral to a dermatologist for biopsy.¹³

Red streaks on more than one nail

Polydactylous longitudinal erythronychia usually presents in adults as red streaks on multiple nails and, depending on the presence or absence of symptoms (eg, pain, splitting), may be the patient’s chief complaint or an incidental finding noted by the astute clinician. Often, it is associated with systemic disease, most commonly lichen planus or Darier disease.

Lichen planus is a papulosquamous skin disease with nail involvement in 10% of patients and permanent nail dystrophy in 4%. Common nail findings include thinning, longitudinal ridging, and fissuring, as well as scarring of the nail matrix resulting in pterygium. Linear red streaks may accompany these more typical nail findings.¹³ Patients with Darier disease present with alternating red and white linear bands on multiple nails as in leukonychia striata.

Less frequently, polydactylous longitudinal erythronychia is associated with primary and systemic amyloidosis, hemiplegia, graft-vs-host disease, acantholytic epidermolysis bullosa, acantholytic dyskeratotic epidermal nevus, acrokeratosis verruciformis of Hopf, or pseudobulbar syndrome, or is idiopathic.^13,42,48 Therefore, the physician evaluating a patient with these nail findings should focus on a workup for regional or systemic disease or refer the patient to a dermatologist who specializes in nails.

BEAU LINES

Beau lines are a common finding in clinical practice. They are not true lines, but transverse grooves in the nail plate that arise from the temporary suppression of nail growth within the nail matrix that can occur during periods of acute or chronic stress or systemic illness (Figure 5).⁴⁹

The precipitating event may be local trauma or paronychia, chemotherapeutic agents cytotoxic to the nail matrix, or the abrupt onset of systemic disease.^18,50 The grooves have also been associated with rheumatic fever, malaria, pemphigus, Raynaud disease, and myocardial infarction, as well as following deep-sea dives.^51–53 The distance of a Beau line from the proximal nail fold can provide an estimate of the time of the acute stress, based on an average growth rate of 3 mm per month for fingernails and 1 mm per month for toenails.⁴⁹

Inspection of the fingernails and toenails should be part of a complete physical examination. A basic understanding of nail anatomy and recognition of several basic types of nail lines and bands allow the clinician to properly diagnose and treat the nail disease, to recognize possible underlying systemic diseases, and to know when to refer the patient to a dermatologist for specialized evaluation and biopsy.

In this review, we delineate the three basic types of nail lines­—white lines (leukonychia striata), brown-black lines (longitudinal melanonychia), and red lines (longitudinal erythronychia)—and the differential diagnosis for each type. We also discuss grooves in the nail plate, or Beau lines.

BASIC NAIL ANATOMY

A fundamental understanding of the anatomy of the nail unit is necessary to understand the origin of nail diseases and underlying pathologic conditions.

The nail unit includes the nail matrix, the lunula, the nail fold, the nail plate, and the nail bed. The nail matrix extends from under the proximal nail fold to the half-moon-shaped area (ie, the lunula) and is responsible for nail plate production. The nail bed lies under the nail plate and on top of the distal phalanx and extends from the lunula to just proximal to the free edge of the nail; its rich blood supply gives it its reddish color.

Nails grow slowly, and this should be kept in mind during the examination. Regrowth of a fingernail takes at least 6 months, and regrowth of a toenail may take 12 to 18 months. Therefore, a defect in the nail plate may reveal an injury that occurred—or a condition that began—several months before.¹

NAIL EXAMINATION ESSENTIALS

A complete examination includes all 20 nail units and the periungual skin. Patients should be instructed to remove nail polish from all nails, as it may camouflage dystrophy or disease of the nail. Photography and careful measurement help document changes over time.

LEUKONYCHIA STRIATA: WHITE NAIL LINES

White nail lines or leukonychia is classified as true or apparent, depending on whether the origin is in the nail matrix or the nail bed.

In true leukonychia, there is abnormal keratinization of the underlying nail matrix, resulting in parakeratosis within the nail plate and an opaque appearance on examination.² The white discoloration is unaffected by pressure, and the opacity moves distally as the nail grows out, which can be documented by serial photography on subsequent visits.

Apparent leukonychia involves abnormal nail bed vasculature, which changes the translucency of the nail plate. The whiteness disappears with pressure, is unaffected by nail growth, and will likely show no change on later visits with serial photography.³

True leukonychia

Leukonychia striata, a subtype of true leuko­nychia, is characterized by transverse or longitudinal bands. It is most often associated with microtrauma, such as from a manicure.⁴ Lines due to trauma are typically more apparent in the central part of the nail plate; they spare the lateral portion and lie parallel to the edge of the proximal nail fold.⁵

Onychomycosis. White longitudinal bands may also be seen in onychomycosis, a fungal infection of the nail accounting for up to 50% of all cases of nail disease. The infection may present as irregular dense longitudinal white or yellowish bands or “spikes” on the nail plate with associated hyperkeratosis, known as a dermatophytoma (Figure 1).

If a fungal infection is suspected, a potassium hydroxide stain can be performed on the subungual debris, which is then examined with direct microscopy.⁶ Alternatively, the physician can send a nail plate clipping in a 10% buffered formalin container with a request for a fungal stain such as periodic acid-Schiff.⁷ Microscopic examination of a dermatophytoma shows a dense mass of dermatophyte hyphae, otherwise known as a fungal abscess.⁸

The physician can play an important role in diagnosis because clinical findings suggestive of a dermatophytoma are associated with a poor response to antifungal therapy.⁹

Inherited diseases. White longitudinal bands are also an important clue to the rare autosomal dominant genodermatoses Hailey-Hailey disease (from mutations of the ATP2A2 gene) and Darier disease (from mutations of the ATP2C1 gene). Patients with Hailey-Hailey disease may have nails with multiple parallel longitudinal white stripes of variable width originating in the lunula and most prominent on the thumbs.^10–12 These patients also have recurrent vesicular eruptions in flexural skin areas, such as the groin, axilla, neck, and periumbilical area causing significant morbidity.

Patients with Darier disease may have nails with alternating red and white longitudinal streaks, described as “candy-cane,”¹³ as well as wedge-shaped distal subungual keratosis accompanied by flat keratotic papules on the proximal nail fold.¹⁴ These nail changes are reported in 92% to 95% of patients with Darier disease.^15,16 Patients typically have skin findings characterized by keratotic papules and plaques predominantly in seborrheic areas and palmoplantar pits, as well as secondary infections and malodor causing significant morbidity.¹⁵ Therefore, knowing the characteristic nail findings in these diseases may lead to more rapid diagnosis and treatment.

Mees lines. Leukonychia striata can present as transverse white lines, commonly known as Mees lines. They are 1- to 2-mm wide horizontal parallel white bands that span the width of the nail plate, usually affecting all fingernails.¹⁷ They are not a common finding and are most often associated with arsenic poisoning. They can also be used to identify the time of poisoning, since they tend to appear 2 months after the initial insult.

Mees lines are also associated with acute systemic stresses, such as acute renal failure, heart failure, ulcerative colitis, breast cancer, infections such as measles and tuberculosis, and systemic lupus erythematosus, and with exposure to toxic metals such as thallium.³

Apparent leukonychia

Apparent leukonychia can alert the physician to systemic diseases, infections, drug side effects, and nutrient deficiencies. Specific nail findings include Muehrcke lines, “half-and-half” nails, and Terry nails.

Muehrcke lines are paired white transverse bands that span the width of the nail bed and run parallel to the distal lunula. They were first described in the fingernails of patients with severe hypoalbuminemia, some of whom also had nephrotic syndrome, which resolved with normalization of the serum albumin level. Muehrcke lines have since been reported in patients with liver disease, malnutrition, chemotherapy, organ transplant, human immunodeficiency virus (HIV) infection, and acquired immunodeficiency syndrome.^3,18 They are associated with periods of metabolic stress, ie, when the body’s capacity to synthesize proteins is diminished.¹⁹

Half-and-half nails, or Lindsay nails, are characterized by a white band proximally, a pink or red-brown band distally, and a sharp demarcation between the two (Figure 2). They were originally described in association with chronic renal disease,²⁰ and surprisingly, they resolve with kidney transplant but not with hemodialysis treatment or improvement in hemoglobin or albumin levels.^21–23 Half-and-half nails have been reported with Kawasaki disease, hepatic cirrhosis, Crohn disease, zinc deficiency, chemotherapy, Behçet disease, and pellagra.^3,24,25 They should be distinguished from Terry nails, which are characterized by leukonychia involving more than 80% of the total nail length.²⁶

Terry nails were originally reported in association with hepatic cirrhosis, usually secondary to alcoholism²⁷ but have since been found with heart failure, type 2 diabetes mellitus, pulmonary tuberculosis, reactive arthritis, older age, Hansen disease, and peripheral vascular disease.^3,26,28,29

LONGITUDINAL MELANONYCHIA: VERTICAL BROWN-BLACK NAIL LINES

Longitudinal melanonychia is the presence of black-brown vertical lines in the nail plate. They have a variety of causes, including blood from trauma; bacterial, fungal, or HIV infection; drug therapy (eg, from minocycline); endocrine disorders (Addison disease); exogenous pigmentation; or excess melanin production within the nail matrix.^30–32 They may also be a sign of a benign condition such as benign melanocytic activation, lentigines, or nevi, or a malignant condition such as melanoma (Figure 3).^33,34

When to suspect melanoma and refer

Although melanoma is less commonly associated with brown-black vertical nail lines, awareness of melanoma-associated longitudinal melanonychia reduces the likelihood of delayed diagnosis and improves patient outcomes.³⁵ Also, it is important to remember that although nail melanoma is more common in the 5th and 6th decades of life, it can occur at any age, even in children.³⁶

Findings that raise suspicion of nail melanoma (Table 1)^33,37 and that should prompt referral to a dermatologist who specializes in nails include the following:

• A personal or family history of melanoma
• Involvement of a “high-risk” digit (thumb, index finger, great toe),^30,31,38 although nail melanoma can occur in any digit
• Any new vertical brown-black nail pigmentation in a fair-skinned patient
• Only one nail affected: involvement of more than one nail is common in people with darker skin, and nearly all patients with darker skin exhibit longitudinal melanonychia by age 50³¹
• Changes in the band such as darkening, widening, and bleeding
• A bandwidth greater than 6 mm³³
• A band that is wider proximally than distally³⁴
• Nonuniform color of the line
• Indistinct lateral borders
• Associated with pigmentation of the nail fold (the Hutchinson sign, representing subungual melanoma),^31,39 nail plate dystrophy, bleeding, or ulceration.³³

While these features may help distinguish benign from malignant causes of longitudinal melanonychia, the clinical examination alone may not provide a definitive diagnosis. Delayed diagnosis of nail melanoma carries a high mortality rate; the internist can promote early diagnosis by recognizing the risk factors and clinical signs and referring the patient to a dermatologist for further evaluation with nail biopsy.

LONGITUDINAL ERYTHRONYCHIA: VERTICAL RED NAIL LINES

Longitudinal erythronychia—the presence of one or more linear red bands in the nail unit—can be localized (involving only one nail) or polydactylous (involving more than one nail). The localized form is usually due to a neoplastic process, whereas involvement of more than one nail may indicate an underlying regional or systemic disease.¹³Table 2 lists indications for referral to a nail specialist.

General features on examination

Clinical examination reveals one or more linear, pink-red streaks extending from the proximal nail fold to the distal free edge of the nail plate (Figure 4). The width of the band typically ranges from less than 1 mm to 3 mm.⁴⁰ Other features may include splinter hemorrhages within a red band, a semitransparent distal matrix, distal V-shaped chipping, splitting, onycholysis of the nail plate, and reactive distal nail bed and hyponychial hyperkeratosis. These features can be visible to the naked eye but may be better viewed with a magnifying glass, a 7× loupe, or a dermatoscope.¹³

Localized longitudinal erythronychia is usually seen in middle-aged individuals and is most commonly found on the thumbnail, followed by the index finger.^41,42 The condition may be asymptomatic, but the patient may present with pain or with concern that the split end of the nail catches on fabrics or small objects.⁴²

Glomus tumor

Intense, pulsatile pain with sensitivity to cold and tenderness to palpation is highly suggestive of glomus tumor,⁴³ a benign neoplasm that originates from a neuromyoarterial glomus body. Glomus bodies are located throughout the body but are more highly concentrated in the fingertips, especially beneath the nails, and they regulate skin circulation. Therefore, the nail unit is the most common site for glomus tumor.^44,45 A characteristic feature of subungual glomus tumor is demonstration of tenderness after pin-point palpation of the suspected tumor (positive Love sign).⁴⁵ While it is typical for glomus tumor to affect only one nail, multiple tumors are associated with neurofibromatosis type 1.⁴⁶ Confirmation of this diagnosis requires referral to a dermatologist.

Other causes of localized red nail lines

Onychopapilloma, a benign idiopathic tumor, is the most common cause of localized longitudinal erythronychia. Unlike glomus tumor, it is usually asymptomatic.^42,47 Less common benign conditions are warts, warty dyskeratoma, benign vascular proliferation, a solitary lesion of lichen planus, hemiplegia, and postsurgical scarring of the nail matrix. In some cases, the lines are idiopathic.^42,43

Malignant diseases that can present as localized longitudinal erythronychia include invasive squamous cell carcinoma, squamous cell carcinoma in situ (Bowen disease), and, less frequently, amelanotic melanoma in situ, malignant melanoma, and basal cell carcinoma.⁴² Squamous cell carcinoma in situ most commonly presents in the 5th decade of life and is the malignancy most commonly associated with localized longitudinal erythronychia. Clinically, there is also often nail dystrophy, such as distal subungual keratosis or onycholysis.⁴³

Patients with asymptomatic, stable localized longitudinal erythronychia may be followed closely with photography and measurements. However, any new lesion or a change in an existing lesion should prompt referral to a dermatologist for biopsy.¹³

Red streaks on more than one nail

Polydactylous longitudinal erythronychia usually presents in adults as red streaks on multiple nails and, depending on the presence or absence of symptoms (eg, pain, splitting), may be the patient’s chief complaint or an incidental finding noted by the astute clinician. Often, it is associated with systemic disease, most commonly lichen planus or Darier disease.

Lichen planus is a papulosquamous skin disease with nail involvement in 10% of patients and permanent nail dystrophy in 4%. Common nail findings include thinning, longitudinal ridging, and fissuring, as well as scarring of the nail matrix resulting in pterygium. Linear red streaks may accompany these more typical nail findings.¹³ Patients with Darier disease present with alternating red and white linear bands on multiple nails as in leukonychia striata.

Less frequently, polydactylous longitudinal erythronychia is associated with primary and systemic amyloidosis, hemiplegia, graft-vs-host disease, acantholytic epidermolysis bullosa, acantholytic dyskeratotic epidermal nevus, acrokeratosis verruciformis of Hopf, or pseudobulbar syndrome, or is idiopathic.^13,42,48 Therefore, the physician evaluating a patient with these nail findings should focus on a workup for regional or systemic disease or refer the patient to a dermatologist who specializes in nails.

BEAU LINES

Beau lines are a common finding in clinical practice. They are not true lines, but transverse grooves in the nail plate that arise from the temporary suppression of nail growth within the nail matrix that can occur during periods of acute or chronic stress or systemic illness (Figure 5).⁴⁹

The precipitating event may be local trauma or paronychia, chemotherapeutic agents cytotoxic to the nail matrix, or the abrupt onset of systemic disease.^18,50 The grooves have also been associated with rheumatic fever, malaria, pemphigus, Raynaud disease, and myocardial infarction, as well as following deep-sea dives.^51–53 The distance of a Beau line from the proximal nail fold can provide an estimate of the time of the acute stress, based on an average growth rate of 3 mm per month for fingernails and 1 mm per month for toenails.⁴⁹

Pseudomembranous colitis is most often due to Clostridium difficile infection, but it has a variety of other causes, including other infections, ischemia, medications, and inflammatory mucosal diseases (Table 1). When pseudomembranes are found, one should consider these other causes if tests for C difficile are negative or if anti-C difficile therapy does not produce a response.

These less common causes are important to consider to avoid needlessly escalating anti-C difficile antibiotic therapy and to provide appropriate treatment. Pseudomembranous colitis is a nonspecific finding that suggests a larger disease process. Associated signs and symptoms, including fever, abdominal pain, leukocytosis, diarrhea, toxic megacolon, and electrolyte imbalances, may portend a life-threatening condition.¹ Awareness of causes of pseudomembranous colitis other than C difficile infection, the focus of this review, is key to prompt diagnosis and potentially life-saving patient care.

PSEUDOMEMBRANES ARE NONSPECIFIC

A pseudomembrane is a layer of fibropurulent exudate composed of acute inflammatory cells and mucus originating from inflamed and erupting crypts.² Although most often seen in C difficile infection, pseudomembranous colitis is a nonspecific pattern of injury resulting from decreased oxygenation, endothelial damage, and impaired blood flow to the mucosa that can be triggered by a number of disease states.²

On endoscopy, pseudomembranes appear as raised whitish or yellowish plaques that may be scattered or confluent in distribution (Figure 1).² They are usually found in the recto­sigmoid colon but may be isolated to more proximal segments.³ Lower endoscopy is often performed in the diagnostic evaluation of patients with unexplained diarrhea, hematochezia, and abnormal abdominal computed tomographic findings (eg, colonic thickening).

CAUSES OF NON-C DIFFICLE PSEUDOMEMBRANOUS COLITIS

When pseudomembranous colitis is confirmed endoscopically, C difficile infection naturally comes to mind, but the two terms are not interchangeable. A wide differential diagnosis should be maintained, especially when there are clues that C difficile infection may not be the correct diagnosis.

Chemicals and medications

Several chemicals and medications can injure the bowel and predispose to pseudomembrane formation.

Glutaraldehyde has long been used to sanitize endoscopes because of its broad antimicrobial activity. Nevertheless, if the disinfecting solution is not adequately rinsed off the endoscope, direct contact with colonic mucosa can produce an allergic and a chemical reaction, resulting in an acute self-limited colitis with pseudomembrane formation.⁴

Chemotherapeutic and antiproliferative agents can be toxic to the bowel, generally through production of free radicals and up-regulation of inflammatory cytokines. The colonic epithelium is then more susceptible to ulceration and mucosal necrosis with pseudomembrane development. Cisplatin, cyclosporine A, docetaxel, and 5-fluorouracil are prominent examples.^5–8

Nonsteroidal anti-inflammatory drugs can damage the mucosa at all levels of the gastrointestinal tract. Although gastric ulcerations are more typical (from nonselective cyclooxygenase inhibition), colonic ulcerations and colitis can occur.⁹ These drugs, particularly diclofenac and indomethacin, have been associated with non-C difficile pseudomembranous colitis when used by themselves or in conjunction with other agents such as cyclosporine A.^8,10,11

Infections

C difficile is the organism most commonly linked to pseudomembranous colitis, but other bacterial, viral, and parasitic pathogens have also been implicated.

Staphylococcus aureus was believed to be responsible for enterocolitis in a series of 155 surgical patients between 1958 and 1962 receiving antibiotic therapy. All had a positive stool culture for S aureus, and nine were found to have pseudomembranes at autopsy.¹²

Although this finding has been disputed as a misdiagnosis, since C difficile infection was not widely recognized until the 1970s, there is evidence that S aureus may indeed be a cause of non-C difficile pseudomembranous colitis. In a review of 36 cases of methicillin-resistant S aureus bacteremia in Japan, four patients were documented to have intestinal pseudomembranes either by endoscopy or autopsy. In two of these patients, biopsies of the pseudomembranes were positive for methicillin-resistant S aureus.¹³

Escherichia coli O157:H7. Pseudomembranes have been seen endoscopically in several adults and children with enterohemorrhagic E coli O157:H7 infection.^14,15 This invasive gram-negative rod normally resides in the gastrointestinal tract of cattle, sheep, and other animals and can be pathogenic to people who eat undercooked beef. The organism attaches to and effaces intestinal epithelial cells, and bacterial proteins and the Shiga toxin then damage the vasculature, precipitating bloody diarrhea. Colonic damage can range from mild hemorrhagic colitis to severe colitis with ischemic changes. In patients with enterohemorrhagic E coli O157:H7 infection, pseudomembrane formation results from colon ischemia due to microvascular thrombosis or from destructive effects of bacterial enterotoxin.^1,15

Cytomegalovirus is a ubiquitous human herpes virus that can affect nearly all organ systems. Infection is often reported in immunocompromised patients, eg, those with acquired immunodeficiency syndrome, chronic corticosteroid use, inflammatory bowel disease, malignancy, or solid-organ transplants. Gastrointestinal manifestations can be nonspecific and range from abdominal discomfort to diarrhea to tenesmus. Pseudomembranous colitis can be a presenting feature of cytomegalovirus involvement.^16,17 Ulcerative lesions in the colonic mucosa are a frequent accompanying finding on endoscopy.^18,19

Other rarer infectious causes of pseudomembranous colitis identified in case reports include Clostridium ramosum, Entamoeba histolytica, Klebsiella oxytoca, Plesiomonas shigelloides, Schistosomiasis mansoni, and Salmonella and Shigella species.^20–26 Additionally, there have been two reports of Strongyloides stercoralis hyperinfection manifesting as pseudomembranous colitis.^27,28

Ischemia

Colon ischemia usually affects elderly or debilitated patients who have multiple comorbidities. Known risk factors include aortoiliac surgery, cardiovascular disease, diabetes mellitus, hemodialysis, and pulmonary disease. The ischemia can be related to an occlusive arterial or venous thromboembolism, but hypoperfusion without occlusion of the mesenteric or internal iliac arteries is the primary mechanism. Low blood flow states such as atherosclerosis and septic shock can affect the watershed areas, typically the splenic flexure and rectosigmoid junction.

We reported a case of a patient with vascular disease who was incorrectly diagnosed and treated as having refractory C difficile infection when pseudomembranes were seen on flexible sigmoidoscopy. Further investigation revealed ischemic colitis secondary to a high-grade inferior mesenteric artery stenosis as the true cause.²⁹

Microvascular thrombosis is the likely mechanism in a number of non-C difficile causes of pseudomembranous colitis. For example, in most patients with enterohemorrhagic E coli O157:H7 infection, histologic review of colonic mucosal biopsies has revealed fibrin and platelet thrombi in the capillaries, suggesting microvascular thrombosis.^15,30

Cocaine has been associated with pseudomembranes in the setting of ischemia in the cecum and ascending colon. Cocaine can cause vasoconstriction after stimulation of alpha-adrenergic receptors and hence intestinal ischemia, thrombosis of vessels in the large and small intestines, and direct toxic effects.³¹

Inflammatory conditions

Collagenous colitis is an inflammatory disease that often affects middle-aged women and presents with copious watery diarrhea. It is a type of microscopic colitis—the endoscopic appearance is often normal, while the histologic appearance is abnormal and characterized by collagen deposition in the lamina propria. Medications that have been implicated in microscopic colitis include acid-suppressive agents (eg, histamine receptor antagonists, proton pump inhibitors) and nonsteroidal anti-inflammatory drugs.

An increasing number of cases of pseudomembranous changes are being reported in patients diagnosed with collagenous colitis.^32–36 Although the pathophysiologic mechanism is unknown, some authors have suggested that pseudomembrane formation is actually part of the presenting spectrum of collagenous colitis.³⁶

Inflammatory bowel disease. Crohn disease and ulcerative colitis have been associated with pseudomembranous colitis. Pseudomembranes can be found on endoscopy in patients with inflammatory bowel disease during a disease exacerbation with or without C difficile.^37,38 In patients with inflammatory bowel disease and C difficile infection, pseudomembranes can be found endoscopically in up to 13% of cases.³⁹ Pseudomembranous colitis has been reported in a patient with ulcerative colitis exacerbation in association with cytomegalovirus colitis.⁴⁰

Behçet disease is a rare, immune-mediated small-vessel systemic vasculitis. It usually presents with mucous membrane ulcerations and ocular disease but can affect any organ.⁴¹ Pseudomembranous colitis can occur in Behçet disease in the absence of C difficile infection or any infectious colitis. Treatment includes corticosteroids and immunosuppressants such as azathioprine and anti-tumor necrosis factor agents.⁴¹

INITIAL EVALUATION

The initial evaluation of a patient with suspected or confirmed pseudomembranous colitis should include a comprehensive medical history with information on recent hospitalizations or procedures, antibiotic use, infections, exposure to sick contacts, recent travel, and medications taken.

Testing for C difficile

As most patients with pseudomembranous colitis have C difficile infection, it should be excluded first. Empiric anti-C difficile treatment is recommended in seriously ill-appearing patients, ideally starting after a stool sample is obtained.

Diagnosis of C difficile infection requires laboratory demonstration of the toxin or detection of toxigenic organisms. The gold standard test is the cell culture cytotoxicity assay, but it is labor- and time-intensive.⁴² More widely available tests are polymerase chain reaction for the toxin gene or genes, enzyme immunoassay, and stool evaluation for glutamate dehydrogenase, which can yield results readily within hours.

Polymerase chain reaction has a sensitivity of 97% and a specificity of 93%. Results can be falsely positive if empiric treatment is started before specimen collection, in which case C difficile DNA may still be present and detectable, but not the organism itself.⁴³

Enzyme immunoassay for toxins A and B carries a sensitivity of 75% and a specificity of 99%, but 100 to 1,000 pg of toxin must be present for a positive result.^44,45

If the initial enzyme immunoassay or polymerase chain reaction result is negative, current guidelines do not recommend repeat testing, which has limited value.^44,46 Repeat testing after a negative result is positive in less than 5% of samples and greatly increases the chances of false-positive results.^44,46 Nevertheless, if a laboratory’s enzyme immunoassay test has a low sensitivity, repeating negative tests may improve its sensitivity.

Glutamate dehydrogenase is an enzyme produced by both toxigenic and nontoxigenic strains of C difficile. As a result, stool testing for glutamate dehydrogenase is sensitive but not specific for C difficile infection, although multistep testing sequences (glutamate dehydrogenase followed by polymerase chain reaction) have proven to be useful screening tools.⁴⁴

Treatment for C difficile infection

If testing for C difficile is positive, treatment is generally based on the severity and the complications of the illness⁴⁶:

• Mild or moderate C difficile infection should be treated with oral metronidazole 500 mg three times per day for 10 to 14 days.
• Severe infection, which is defined as a white blood cell count of 15.0 × 10⁹/L or higher or a serum creatinine level greater than or equal to 1.5 times the premorbid level, should be treated with oral vancomycin 125 mg four times per day for 10 to 14 days.
• Severe C difficile infection complicated by hypotension, shock, ileus, or megacolon should be treated with a combination of high-dose oral vancomycin (and possibly rectal vancomycin as well) at 500 mg four times per day plus intravenous metronidazole.

Additional treatment recommendations for individualized situations, recurrent C difficile infection, and comorbid conditions are discussed elsewhere.⁴⁶

ENDOSCOPIC EVALUATION

Colonoscopy or flexible sigmoidoscopy is the primary means by which pseudomembranous colitis is diagnosed. Lower endoscopy should be pursued as an adjunctive tool when C difficile infection remains strongly suspected despite negative testing, when presumed C difficile infection does not respond to medical therapy, and when non-C difficile diagnoses are considered. If pseudomembranes are demonstrated on lower endoscopy, obtaining biopsy samples of normal- and abnormal-appearing mucosa is recommended. Pseudomembranes are suggestive but not diagnostic of C difficile infection, and microscopic evaluation of the mucosa is warranted to explore causes of pseudomembranous colitis not related to C difficile.

The pattern and distribution of pseudomembranes may provide clues to the etiology and the degree of mucosal injury. In intestinal ischemia, for example, a localized segment of the bowel is typically involved, and mucosal changes are often well delineated from normal mucosa. On endoscopic examination, mild ischemia is characterized by granular mucosa with decreased vascularity, whereas friable, edematous, ulcerated, and at times necrotic mucosa is evident in severe cases. Punctate pseudomembrane formation is seen in early ischemia, but as injury progresses, the pseudomembranes may grow and merge. In fact, diffuse involvement of the mucosal surface of the biopsy specimen by pseudomembranes has been shown to be more closely associated with ischemic colitis than C difficile infection.⁴⁷

MICROSCOPIC EVALUATION

Histologic study can differentiate the various causes of pseudomembranous colitis.

In C difficile infection and drug reaction, there is acute crypt injury and dilation. The upper lamina propria is usually involved, and affected crypts are filled with an exudate similar to that found in pseudomembranes.¹ However, in drug reaction, there is also prominent apoptosis and increased intraepithelial lymphocytosis.⁹

In colon ischemia, hyalinization of the lamina propria is a sensitive and specific marker.⁴⁷ This has been shown in a study comparing histologic characteristics of colonic biopsies in patients with pseudomembranous colitis due to either known colon ischemia or C difficile infection.⁴⁷ Crypt atrophy, lamina propria hemorrhage, full-thickness mucosal necrosis, and layering of pseudomembranes would further favor the diagnosis.

In collagenous colitis, a thickened subepithelial collagen band and intraepithelial lymphocytosis are often seen.

In inflammatory bowel disease, even with secondary pseudomembranes, ulcerative colitis and Crohn disease retain the characteristics of inflammatory bowel disease with crypt architectural distortion and focal or diffuse basal lymphoplasmacytosis on microscopy.⁴⁸

Pseudomembranous colitis is most often due to Clostridium difficile infection, but it has a variety of other causes, including other infections, ischemia, medications, and inflammatory mucosal diseases (Table 1). When pseudomembranes are found, one should consider these other causes if tests for C difficile are negative or if anti-C difficile therapy does not produce a response.

These less common causes are important to consider to avoid needlessly escalating anti-C difficile antibiotic therapy and to provide appropriate treatment. Pseudomembranous colitis is a nonspecific finding that suggests a larger disease process. Associated signs and symptoms, including fever, abdominal pain, leukocytosis, diarrhea, toxic megacolon, and electrolyte imbalances, may portend a life-threatening condition.¹ Awareness of causes of pseudomembranous colitis other than C difficile infection, the focus of this review, is key to prompt diagnosis and potentially life-saving patient care.

PSEUDOMEMBRANES ARE NONSPECIFIC

A pseudomembrane is a layer of fibropurulent exudate composed of acute inflammatory cells and mucus originating from inflamed and erupting crypts.² Although most often seen in C difficile infection, pseudomembranous colitis is a nonspecific pattern of injury resulting from decreased oxygenation, endothelial damage, and impaired blood flow to the mucosa that can be triggered by a number of disease states.²

On endoscopy, pseudomembranes appear as raised whitish or yellowish plaques that may be scattered or confluent in distribution (Figure 1).² They are usually found in the recto­sigmoid colon but may be isolated to more proximal segments.³ Lower endoscopy is often performed in the diagnostic evaluation of patients with unexplained diarrhea, hematochezia, and abnormal abdominal computed tomographic findings (eg, colonic thickening).

CAUSES OF NON-C DIFFICLE PSEUDOMEMBRANOUS COLITIS

When pseudomembranous colitis is confirmed endoscopically, C difficile infection naturally comes to mind, but the two terms are not interchangeable. A wide differential diagnosis should be maintained, especially when there are clues that C difficile infection may not be the correct diagnosis.

Chemicals and medications

Several chemicals and medications can injure the bowel and predispose to pseudomembrane formation.

Glutaraldehyde has long been used to sanitize endoscopes because of its broad antimicrobial activity. Nevertheless, if the disinfecting solution is not adequately rinsed off the endoscope, direct contact with colonic mucosa can produce an allergic and a chemical reaction, resulting in an acute self-limited colitis with pseudomembrane formation.⁴

Chemotherapeutic and antiproliferative agents can be toxic to the bowel, generally through production of free radicals and up-regulation of inflammatory cytokines. The colonic epithelium is then more susceptible to ulceration and mucosal necrosis with pseudomembrane development. Cisplatin, cyclosporine A, docetaxel, and 5-fluorouracil are prominent examples.^5–8

Nonsteroidal anti-inflammatory drugs can damage the mucosa at all levels of the gastrointestinal tract. Although gastric ulcerations are more typical (from nonselective cyclooxygenase inhibition), colonic ulcerations and colitis can occur.⁹ These drugs, particularly diclofenac and indomethacin, have been associated with non-C difficile pseudomembranous colitis when used by themselves or in conjunction with other agents such as cyclosporine A.^8,10,11

Infections

C difficile is the organism most commonly linked to pseudomembranous colitis, but other bacterial, viral, and parasitic pathogens have also been implicated.

Staphylococcus aureus was believed to be responsible for enterocolitis in a series of 155 surgical patients between 1958 and 1962 receiving antibiotic therapy. All had a positive stool culture for S aureus, and nine were found to have pseudomembranes at autopsy.¹²

Although this finding has been disputed as a misdiagnosis, since C difficile infection was not widely recognized until the 1970s, there is evidence that S aureus may indeed be a cause of non-C difficile pseudomembranous colitis. In a review of 36 cases of methicillin-resistant S aureus bacteremia in Japan, four patients were documented to have intestinal pseudomembranes either by endoscopy or autopsy. In two of these patients, biopsies of the pseudomembranes were positive for methicillin-resistant S aureus.¹³

Escherichia coli O157:H7. Pseudomembranes have been seen endoscopically in several adults and children with enterohemorrhagic E coli O157:H7 infection.^14,15 This invasive gram-negative rod normally resides in the gastrointestinal tract of cattle, sheep, and other animals and can be pathogenic to people who eat undercooked beef. The organism attaches to and effaces intestinal epithelial cells, and bacterial proteins and the Shiga toxin then damage the vasculature, precipitating bloody diarrhea. Colonic damage can range from mild hemorrhagic colitis to severe colitis with ischemic changes. In patients with enterohemorrhagic E coli O157:H7 infection, pseudomembrane formation results from colon ischemia due to microvascular thrombosis or from destructive effects of bacterial enterotoxin.^1,15

Cytomegalovirus is a ubiquitous human herpes virus that can affect nearly all organ systems. Infection is often reported in immunocompromised patients, eg, those with acquired immunodeficiency syndrome, chronic corticosteroid use, inflammatory bowel disease, malignancy, or solid-organ transplants. Gastrointestinal manifestations can be nonspecific and range from abdominal discomfort to diarrhea to tenesmus. Pseudomembranous colitis can be a presenting feature of cytomegalovirus involvement.^16,17 Ulcerative lesions in the colonic mucosa are a frequent accompanying finding on endoscopy.^18,19

Other rarer infectious causes of pseudomembranous colitis identified in case reports include Clostridium ramosum, Entamoeba histolytica, Klebsiella oxytoca, Plesiomonas shigelloides, Schistosomiasis mansoni, and Salmonella and Shigella species.^20–26 Additionally, there have been two reports of Strongyloides stercoralis hyperinfection manifesting as pseudomembranous colitis.^27,28

Ischemia

Colon ischemia usually affects elderly or debilitated patients who have multiple comorbidities. Known risk factors include aortoiliac surgery, cardiovascular disease, diabetes mellitus, hemodialysis, and pulmonary disease. The ischemia can be related to an occlusive arterial or venous thromboembolism, but hypoperfusion without occlusion of the mesenteric or internal iliac arteries is the primary mechanism. Low blood flow states such as atherosclerosis and septic shock can affect the watershed areas, typically the splenic flexure and rectosigmoid junction.

We reported a case of a patient with vascular disease who was incorrectly diagnosed and treated as having refractory C difficile infection when pseudomembranes were seen on flexible sigmoidoscopy. Further investigation revealed ischemic colitis secondary to a high-grade inferior mesenteric artery stenosis as the true cause.²⁹

Microvascular thrombosis is the likely mechanism in a number of non-C difficile causes of pseudomembranous colitis. For example, in most patients with enterohemorrhagic E coli O157:H7 infection, histologic review of colonic mucosal biopsies has revealed fibrin and platelet thrombi in the capillaries, suggesting microvascular thrombosis.^15,30

Cocaine has been associated with pseudomembranes in the setting of ischemia in the cecum and ascending colon. Cocaine can cause vasoconstriction after stimulation of alpha-adrenergic receptors and hence intestinal ischemia, thrombosis of vessels in the large and small intestines, and direct toxic effects.³¹

Inflammatory conditions

Collagenous colitis is an inflammatory disease that often affects middle-aged women and presents with copious watery diarrhea. It is a type of microscopic colitis—the endoscopic appearance is often normal, while the histologic appearance is abnormal and characterized by collagen deposition in the lamina propria. Medications that have been implicated in microscopic colitis include acid-suppressive agents (eg, histamine receptor antagonists, proton pump inhibitors) and nonsteroidal anti-inflammatory drugs.

An increasing number of cases of pseudomembranous changes are being reported in patients diagnosed with collagenous colitis.^32–36 Although the pathophysiologic mechanism is unknown, some authors have suggested that pseudomembrane formation is actually part of the presenting spectrum of collagenous colitis.³⁶

Inflammatory bowel disease. Crohn disease and ulcerative colitis have been associated with pseudomembranous colitis. Pseudomembranes can be found on endoscopy in patients with inflammatory bowel disease during a disease exacerbation with or without C difficile.^37,38 In patients with inflammatory bowel disease and C difficile infection, pseudomembranes can be found endoscopically in up to 13% of cases.³⁹ Pseudomembranous colitis has been reported in a patient with ulcerative colitis exacerbation in association with cytomegalovirus colitis.⁴⁰

Behçet disease is a rare, immune-mediated small-vessel systemic vasculitis. It usually presents with mucous membrane ulcerations and ocular disease but can affect any organ.⁴¹ Pseudomembranous colitis can occur in Behçet disease in the absence of C difficile infection or any infectious colitis. Treatment includes corticosteroids and immunosuppressants such as azathioprine and anti-tumor necrosis factor agents.⁴¹

INITIAL EVALUATION

The initial evaluation of a patient with suspected or confirmed pseudomembranous colitis should include a comprehensive medical history with information on recent hospitalizations or procedures, antibiotic use, infections, exposure to sick contacts, recent travel, and medications taken.

Testing for C difficile

As most patients with pseudomembranous colitis have C difficile infection, it should be excluded first. Empiric anti-C difficile treatment is recommended in seriously ill-appearing patients, ideally starting after a stool sample is obtained.

Diagnosis of C difficile infection requires laboratory demonstration of the toxin or detection of toxigenic organisms. The gold standard test is the cell culture cytotoxicity assay, but it is labor- and time-intensive.⁴² More widely available tests are polymerase chain reaction for the toxin gene or genes, enzyme immunoassay, and stool evaluation for glutamate dehydrogenase, which can yield results readily within hours.

Polymerase chain reaction has a sensitivity of 97% and a specificity of 93%. Results can be falsely positive if empiric treatment is started before specimen collection, in which case C difficile DNA may still be present and detectable, but not the organism itself.⁴³

Enzyme immunoassay for toxins A and B carries a sensitivity of 75% and a specificity of 99%, but 100 to 1,000 pg of toxin must be present for a positive result.^44,45

If the initial enzyme immunoassay or polymerase chain reaction result is negative, current guidelines do not recommend repeat testing, which has limited value.^44,46 Repeat testing after a negative result is positive in less than 5% of samples and greatly increases the chances of false-positive results.^44,46 Nevertheless, if a laboratory’s enzyme immunoassay test has a low sensitivity, repeating negative tests may improve its sensitivity.

Glutamate dehydrogenase is an enzyme produced by both toxigenic and nontoxigenic strains of C difficile. As a result, stool testing for glutamate dehydrogenase is sensitive but not specific for C difficile infection, although multistep testing sequences (glutamate dehydrogenase followed by polymerase chain reaction) have proven to be useful screening tools.⁴⁴

Treatment for C difficile infection

If testing for C difficile is positive, treatment is generally based on the severity and the complications of the illness⁴⁶:

• Mild or moderate C difficile infection should be treated with oral metronidazole 500 mg three times per day for 10 to 14 days.
• Severe infection, which is defined as a white blood cell count of 15.0 × 10⁹/L or higher or a serum creatinine level greater than or equal to 1.5 times the premorbid level, should be treated with oral vancomycin 125 mg four times per day for 10 to 14 days.
• Severe C difficile infection complicated by hypotension, shock, ileus, or megacolon should be treated with a combination of high-dose oral vancomycin (and possibly rectal vancomycin as well) at 500 mg four times per day plus intravenous metronidazole.

Additional treatment recommendations for individualized situations, recurrent C difficile infection, and comorbid conditions are discussed elsewhere.⁴⁶

ENDOSCOPIC EVALUATION

Colonoscopy or flexible sigmoidoscopy is the primary means by which pseudomembranous colitis is diagnosed. Lower endoscopy should be pursued as an adjunctive tool when C difficile infection remains strongly suspected despite negative testing, when presumed C difficile infection does not respond to medical therapy, and when non-C difficile diagnoses are considered. If pseudomembranes are demonstrated on lower endoscopy, obtaining biopsy samples of normal- and abnormal-appearing mucosa is recommended. Pseudomembranes are suggestive but not diagnostic of C difficile infection, and microscopic evaluation of the mucosa is warranted to explore causes of pseudomembranous colitis not related to C difficile.

The pattern and distribution of pseudomembranes may provide clues to the etiology and the degree of mucosal injury. In intestinal ischemia, for example, a localized segment of the bowel is typically involved, and mucosal changes are often well delineated from normal mucosa. On endoscopic examination, mild ischemia is characterized by granular mucosa with decreased vascularity, whereas friable, edematous, ulcerated, and at times necrotic mucosa is evident in severe cases. Punctate pseudomembrane formation is seen in early ischemia, but as injury progresses, the pseudomembranes may grow and merge. In fact, diffuse involvement of the mucosal surface of the biopsy specimen by pseudomembranes has been shown to be more closely associated with ischemic colitis than C difficile infection.⁴⁷

MICROSCOPIC EVALUATION

Histologic study can differentiate the various causes of pseudomembranous colitis.

In C difficile infection and drug reaction, there is acute crypt injury and dilation. The upper lamina propria is usually involved, and affected crypts are filled with an exudate similar to that found in pseudomembranes.¹ However, in drug reaction, there is also prominent apoptosis and increased intraepithelial lymphocytosis.⁹

In colon ischemia, hyalinization of the lamina propria is a sensitive and specific marker.⁴⁷ This has been shown in a study comparing histologic characteristics of colonic biopsies in patients with pseudomembranous colitis due to either known colon ischemia or C difficile infection.⁴⁷ Crypt atrophy, lamina propria hemorrhage, full-thickness mucosal necrosis, and layering of pseudomembranes would further favor the diagnosis.

In collagenous colitis, a thickened subepithelial collagen band and intraepithelial lymphocytosis are often seen.

In inflammatory bowel disease, even with secondary pseudomembranes, ulcerative colitis and Crohn disease retain the characteristics of inflammatory bowel disease with crypt architectural distortion and focal or diffuse basal lymphoplasmacytosis on microscopy.⁴⁸

Pseudomembranous colitis is most often due to Clostridium difficile infection, but it has a variety of other causes, including other infections, ischemia, medications, and inflammatory mucosal diseases (Table 1). When pseudomembranes are found, one should consider these other causes if tests for C difficile are negative or if anti-C difficile therapy does not produce a response.

These less common causes are important to consider to avoid needlessly escalating anti-C difficile antibiotic therapy and to provide appropriate treatment. Pseudomembranous colitis is a nonspecific finding that suggests a larger disease process. Associated signs and symptoms, including fever, abdominal pain, leukocytosis, diarrhea, toxic megacolon, and electrolyte imbalances, may portend a life-threatening condition.¹ Awareness of causes of pseudomembranous colitis other than C difficile infection, the focus of this review, is key to prompt diagnosis and potentially life-saving patient care.

PSEUDOMEMBRANES ARE NONSPECIFIC

A pseudomembrane is a layer of fibropurulent exudate composed of acute inflammatory cells and mucus originating from inflamed and erupting crypts.² Although most often seen in C difficile infection, pseudomembranous colitis is a nonspecific pattern of injury resulting from decreased oxygenation, endothelial damage, and impaired blood flow to the mucosa that can be triggered by a number of disease states.²

On endoscopy, pseudomembranes appear as raised whitish or yellowish plaques that may be scattered or confluent in distribution (Figure 1).² They are usually found in the recto­sigmoid colon but may be isolated to more proximal segments.³ Lower endoscopy is often performed in the diagnostic evaluation of patients with unexplained diarrhea, hematochezia, and abnormal abdominal computed tomographic findings (eg, colonic thickening).

CAUSES OF NON-C DIFFICLE PSEUDOMEMBRANOUS COLITIS

When pseudomembranous colitis is confirmed endoscopically, C difficile infection naturally comes to mind, but the two terms are not interchangeable. A wide differential diagnosis should be maintained, especially when there are clues that C difficile infection may not be the correct diagnosis.

Chemicals and medications

Several chemicals and medications can injure the bowel and predispose to pseudomembrane formation.

Glutaraldehyde has long been used to sanitize endoscopes because of its broad antimicrobial activity. Nevertheless, if the disinfecting solution is not adequately rinsed off the endoscope, direct contact with colonic mucosa can produce an allergic and a chemical reaction, resulting in an acute self-limited colitis with pseudomembrane formation.⁴

Chemotherapeutic and antiproliferative agents can be toxic to the bowel, generally through production of free radicals and up-regulation of inflammatory cytokines. The colonic epithelium is then more susceptible to ulceration and mucosal necrosis with pseudomembrane development. Cisplatin, cyclosporine A, docetaxel, and 5-fluorouracil are prominent examples.^5–8

Nonsteroidal anti-inflammatory drugs can damage the mucosa at all levels of the gastrointestinal tract. Although gastric ulcerations are more typical (from nonselective cyclooxygenase inhibition), colonic ulcerations and colitis can occur.⁹ These drugs, particularly diclofenac and indomethacin, have been associated with non-C difficile pseudomembranous colitis when used by themselves or in conjunction with other agents such as cyclosporine A.^8,10,11

Infections

C difficile is the organism most commonly linked to pseudomembranous colitis, but other bacterial, viral, and parasitic pathogens have also been implicated.

Staphylococcus aureus was believed to be responsible for enterocolitis in a series of 155 surgical patients between 1958 and 1962 receiving antibiotic therapy. All had a positive stool culture for S aureus, and nine were found to have pseudomembranes at autopsy.¹²

Although this finding has been disputed as a misdiagnosis, since C difficile infection was not widely recognized until the 1970s, there is evidence that S aureus may indeed be a cause of non-C difficile pseudomembranous colitis. In a review of 36 cases of methicillin-resistant S aureus bacteremia in Japan, four patients were documented to have intestinal pseudomembranes either by endoscopy or autopsy. In two of these patients, biopsies of the pseudomembranes were positive for methicillin-resistant S aureus.¹³

Escherichia coli O157:H7. Pseudomembranes have been seen endoscopically in several adults and children with enterohemorrhagic E coli O157:H7 infection.^14,15 This invasive gram-negative rod normally resides in the gastrointestinal tract of cattle, sheep, and other animals and can be pathogenic to people who eat undercooked beef. The organism attaches to and effaces intestinal epithelial cells, and bacterial proteins and the Shiga toxin then damage the vasculature, precipitating bloody diarrhea. Colonic damage can range from mild hemorrhagic colitis to severe colitis with ischemic changes. In patients with enterohemorrhagic E coli O157:H7 infection, pseudomembrane formation results from colon ischemia due to microvascular thrombosis or from destructive effects of bacterial enterotoxin.^1,15

Cytomegalovirus is a ubiquitous human herpes virus that can affect nearly all organ systems. Infection is often reported in immunocompromised patients, eg, those with acquired immunodeficiency syndrome, chronic corticosteroid use, inflammatory bowel disease, malignancy, or solid-organ transplants. Gastrointestinal manifestations can be nonspecific and range from abdominal discomfort to diarrhea to tenesmus. Pseudomembranous colitis can be a presenting feature of cytomegalovirus involvement.^16,17 Ulcerative lesions in the colonic mucosa are a frequent accompanying finding on endoscopy.^18,19

Other rarer infectious causes of pseudomembranous colitis identified in case reports include Clostridium ramosum, Entamoeba histolytica, Klebsiella oxytoca, Plesiomonas shigelloides, Schistosomiasis mansoni, and Salmonella and Shigella species.^20–26 Additionally, there have been two reports of Strongyloides stercoralis hyperinfection manifesting as pseudomembranous colitis.^27,28

Ischemia

Colon ischemia usually affects elderly or debilitated patients who have multiple comorbidities. Known risk factors include aortoiliac surgery, cardiovascular disease, diabetes mellitus, hemodialysis, and pulmonary disease. The ischemia can be related to an occlusive arterial or venous thromboembolism, but hypoperfusion without occlusion of the mesenteric or internal iliac arteries is the primary mechanism. Low blood flow states such as atherosclerosis and septic shock can affect the watershed areas, typically the splenic flexure and rectosigmoid junction.

We reported a case of a patient with vascular disease who was incorrectly diagnosed and treated as having refractory C difficile infection when pseudomembranes were seen on flexible sigmoidoscopy. Further investigation revealed ischemic colitis secondary to a high-grade inferior mesenteric artery stenosis as the true cause.²⁹

Microvascular thrombosis is the likely mechanism in a number of non-C difficile causes of pseudomembranous colitis. For example, in most patients with enterohemorrhagic E coli O157:H7 infection, histologic review of colonic mucosal biopsies has revealed fibrin and platelet thrombi in the capillaries, suggesting microvascular thrombosis.^15,30

Cocaine has been associated with pseudomembranes in the setting of ischemia in the cecum and ascending colon. Cocaine can cause vasoconstriction after stimulation of alpha-adrenergic receptors and hence intestinal ischemia, thrombosis of vessels in the large and small intestines, and direct toxic effects.³¹

Inflammatory conditions

Collagenous colitis is an inflammatory disease that often affects middle-aged women and presents with copious watery diarrhea. It is a type of microscopic colitis—the endoscopic appearance is often normal, while the histologic appearance is abnormal and characterized by collagen deposition in the lamina propria. Medications that have been implicated in microscopic colitis include acid-suppressive agents (eg, histamine receptor antagonists, proton pump inhibitors) and nonsteroidal anti-inflammatory drugs.

An increasing number of cases of pseudomembranous changes are being reported in patients diagnosed with collagenous colitis.^32–36 Although the pathophysiologic mechanism is unknown, some authors have suggested that pseudomembrane formation is actually part of the presenting spectrum of collagenous colitis.³⁶

Inflammatory bowel disease. Crohn disease and ulcerative colitis have been associated with pseudomembranous colitis. Pseudomembranes can be found on endoscopy in patients with inflammatory bowel disease during a disease exacerbation with or without C difficile.^37,38 In patients with inflammatory bowel disease and C difficile infection, pseudomembranes can be found endoscopically in up to 13% of cases.³⁹ Pseudomembranous colitis has been reported in a patient with ulcerative colitis exacerbation in association with cytomegalovirus colitis.⁴⁰

Behçet disease is a rare, immune-mediated small-vessel systemic vasculitis. It usually presents with mucous membrane ulcerations and ocular disease but can affect any organ.⁴¹ Pseudomembranous colitis can occur in Behçet disease in the absence of C difficile infection or any infectious colitis. Treatment includes corticosteroids and immunosuppressants such as azathioprine and anti-tumor necrosis factor agents.⁴¹

INITIAL EVALUATION

The initial evaluation of a patient with suspected or confirmed pseudomembranous colitis should include a comprehensive medical history with information on recent hospitalizations or procedures, antibiotic use, infections, exposure to sick contacts, recent travel, and medications taken.

Testing for C difficile

As most patients with pseudomembranous colitis have C difficile infection, it should be excluded first. Empiric anti-C difficile treatment is recommended in seriously ill-appearing patients, ideally starting after a stool sample is obtained.

Diagnosis of C difficile infection requires laboratory demonstration of the toxin or detection of toxigenic organisms. The gold standard test is the cell culture cytotoxicity assay, but it is labor- and time-intensive.⁴² More widely available tests are polymerase chain reaction for the toxin gene or genes, enzyme immunoassay, and stool evaluation for glutamate dehydrogenase, which can yield results readily within hours.

Polymerase chain reaction has a sensitivity of 97% and a specificity of 93%. Results can be falsely positive if empiric treatment is started before specimen collection, in which case C difficile DNA may still be present and detectable, but not the organism itself.⁴³

Enzyme immunoassay for toxins A and B carries a sensitivity of 75% and a specificity of 99%, but 100 to 1,000 pg of toxin must be present for a positive result.^44,45

If the initial enzyme immunoassay or polymerase chain reaction result is negative, current guidelines do not recommend repeat testing, which has limited value.^44,46 Repeat testing after a negative result is positive in less than 5% of samples and greatly increases the chances of false-positive results.^44,46 Nevertheless, if a laboratory’s enzyme immunoassay test has a low sensitivity, repeating negative tests may improve its sensitivity.

Glutamate dehydrogenase is an enzyme produced by both toxigenic and nontoxigenic strains of C difficile. As a result, stool testing for glutamate dehydrogenase is sensitive but not specific for C difficile infection, although multistep testing sequences (glutamate dehydrogenase followed by polymerase chain reaction) have proven to be useful screening tools.⁴⁴

Treatment for C difficile infection

If testing for C difficile is positive, treatment is generally based on the severity and the complications of the illness⁴⁶:

• Mild or moderate C difficile infection should be treated with oral metronidazole 500 mg three times per day for 10 to 14 days.
• Severe infection, which is defined as a white blood cell count of 15.0 × 10⁹/L or higher or a serum creatinine level greater than or equal to 1.5 times the premorbid level, should be treated with oral vancomycin 125 mg four times per day for 10 to 14 days.
• Severe C difficile infection complicated by hypotension, shock, ileus, or megacolon should be treated with a combination of high-dose oral vancomycin (and possibly rectal vancomycin as well) at 500 mg four times per day plus intravenous metronidazole.

Additional treatment recommendations for individualized situations, recurrent C difficile infection, and comorbid conditions are discussed elsewhere.⁴⁶

ENDOSCOPIC EVALUATION

Colonoscopy or flexible sigmoidoscopy is the primary means by which pseudomembranous colitis is diagnosed. Lower endoscopy should be pursued as an adjunctive tool when C difficile infection remains strongly suspected despite negative testing, when presumed C difficile infection does not respond to medical therapy, and when non-C difficile diagnoses are considered. If pseudomembranes are demonstrated on lower endoscopy, obtaining biopsy samples of normal- and abnormal-appearing mucosa is recommended. Pseudomembranes are suggestive but not diagnostic of C difficile infection, and microscopic evaluation of the mucosa is warranted to explore causes of pseudomembranous colitis not related to C difficile.

The pattern and distribution of pseudomembranes may provide clues to the etiology and the degree of mucosal injury. In intestinal ischemia, for example, a localized segment of the bowel is typically involved, and mucosal changes are often well delineated from normal mucosa. On endoscopic examination, mild ischemia is characterized by granular mucosa with decreased vascularity, whereas friable, edematous, ulcerated, and at times necrotic mucosa is evident in severe cases. Punctate pseudomembrane formation is seen in early ischemia, but as injury progresses, the pseudomembranes may grow and merge. In fact, diffuse involvement of the mucosal surface of the biopsy specimen by pseudomembranes has been shown to be more closely associated with ischemic colitis than C difficile infection.⁴⁷

MICROSCOPIC EVALUATION

Histologic study can differentiate the various causes of pseudomembranous colitis.

In C difficile infection and drug reaction, there is acute crypt injury and dilation. The upper lamina propria is usually involved, and affected crypts are filled with an exudate similar to that found in pseudomembranes.¹ However, in drug reaction, there is also prominent apoptosis and increased intraepithelial lymphocytosis.⁹

In colon ischemia, hyalinization of the lamina propria is a sensitive and specific marker.⁴⁷ This has been shown in a study comparing histologic characteristics of colonic biopsies in patients with pseudomembranous colitis due to either known colon ischemia or C difficile infection.⁴⁷ Crypt atrophy, lamina propria hemorrhage, full-thickness mucosal necrosis, and layering of pseudomembranes would further favor the diagnosis.

In collagenous colitis, a thickened subepithelial collagen band and intraepithelial lymphocytosis are often seen.

In inflammatory bowel disease, even with secondary pseudomembranes, ulcerative colitis and Crohn disease retain the characteristics of inflammatory bowel disease with crypt architectural distortion and focal or diffuse basal lymphoplasmacytosis on microscopy.⁴⁸

Extracorporeal membrance oxygenation (ECMO) provides temporary cardiorespiratory support for patients with severe respiratory or cardiac failure refractory to conventional therapy.¹ It can be configured to provide oxygen, remove carbon dioxide, support perfusion, or all of the above. It may provide a bridge to recovery in patients with acute cardiopulmonary failure or to heart or lung transplant.

Developed in the 1970s, ECMO has proven effective and is widely used in children with respiratory and cardiopulmonary failure.²  However, it remained little used in adults, as early randomized trials showed higher rates of complications in adults who received it and no survival advantage.^3,4 Proponents of using it in adult patients believe that these poor outcomes were at least partially due to limited training, intensive anticoagulation, and excessive volume and pressure during mechanical ventilation. Although ECMO technology has improved substantially in the last decade and survival rates have improved (www.elso.org), evidence to support its routine use in adults remains limited.

Nevertheless, about 14,000 adult patients received ECMO between 1990 and 2014, with a rate of survival to discharge of 57% for those in respiratory failure and 41% for those in cardiac failure.⁵ Its use increased 433% in the United States from 2006 to 2011.⁶

A national survey of critical care physicians and trainees in the United States found they had limited knowledge about ECMO technology and wanted to include specific educational objectives about it in their training.⁷

This article summarizes the principles of ECMO, including practical aspects such as patient selection, monitoring, and complications.

LIMITED EVIDENCE OF BENEFIT FROM CONTROLLED TRIALS

There is limited evidence from randomized controlled trials that ECMO is beneficial in adults.

In acute respiratory failure, the first randomized trial of ECMO in adults was conducted in 1979 in multiple medical centers.³ The survival rate was no higher with ECMO than with mechanical ventilation alone, and complication rates were very high.

Similarly, Morris et al⁴ performed a single-center trial comparing pressure-controlled inverse-ratio ventilation and extracorporeal carbon dioxide removal in patients with acute respiratory distress syndrome, which showed no survival benefit.

After these two early trials, ECMO was largely abandoned, and not until 2009 did a multicenter randomized trial in acute respiratory distress syndrome⁸ rejuvenate interest in its use. Although the trial did not conclusively prove that ECMO was more effective than conventional mechanical ventilation, the findings supported early referral to tertiary care centers with ECMO expertise, and the survival rate was substantially higher than in previous studies. A concise summary of randomized trials and retrospective studies utilizing ECMO in respiratory failure is shown in Table 1.^8–14

During the global pandemic of influenza H1N1 in 2009–2010, several centers reported survival benefits from ECMO in patients with severe acute respiratory distress syndrome secondary to influenza.^{9–12,15–19} Two retrospective case-control studies reported lower mortality rates when H1N1 patients were transferred to ECMO centers¹⁰ and among younger patients with H1N1 who received ECMO.¹²

Ongoing trials (ClinicalTrials.gov identifier NCT01470703) may provide definitive evidence for the effectiveness of ECMO as a rescue therapy in acute respiratory distress syndrome.

In cardiogenic shock, single-center retrospective and observational studies have reported better outcomes for patients who received ECMO for cardiogenic shock secondary to myocardial infarction, pulmonary embolism, sepsis-related cardiomyopathy, and even extracorporeal cardiopulmonary resuscitation.²⁰

WHAT IS ECMO?

In ECMO, venous blood is shunted through a machine to add oxygen, remove carbon dioxide, and regulate temperature (Figure 1). The components of an ECMO circuit are as follows:

• Blood pump
• Membrane oxygenator
• Gas mixer
• Cannulas
• Heater/cooler
• Console.

TWO BASIC CONFIGURATIONS

Two basic ECMO configurations are used in adults: venoarterial and venovenous,²¹ although combinations of the two—hybrid configurations—are sometimes used (Figure 2).

Venoarterial ECMO

Venoarterial ECMO provides complete or partial support to the heart and lungs and is the configuration of choice in patients with isolated cardiac failure that is refractory to other treatments. It takes deoxygenated blood from the venous system and returns oxygenated blood to the arterial circulation.

In the central venoarterial configuration, the intake cannula is most often surgically placed in the right atrium and the return cannula is placed in the proximal ascending aorta.

In the peripheral femoral configuration, the drainage cannula is placed in the femoral vein and advanced to the right atrium, and the return cannula is placed in either the ipsilateral or contralateral femoral artery. However, this configuration provides the patient with retrograde flow (against the native cardiac output), and oxygen delivery to the upper body may be impeded.

Axillary cannulation, in which the return cannula is placed directly into the axillary artery to provide antegrade flow, has been used recently in patients with pulmonary hypertension or right ventricular failure.²²

Venovenous ECMO

Venovenous ECMO provides complete or partial support to the lungs and is the configuration of choice in isolated respiratory failure when cardiac function is preserved. It takes deoxygenated blood from the central venous system—either the femoral vein or internal jugular vein—and returns oxygenated blood to the venous circulation directed into the right atrium. It can be delivered by different cannula configurations based on the patient’s size and clinical requirements.

In the past, the most commonly used configuration was the femoral-atrial, in which the drainage cannula was placed in the femoral vein with the tip advanced to the level of the diaphragm in the inferior vena cava, and the return cannula was placed in the right internal jugular vein with its tip at the junction of the superior vena cava and right atrium. In this configuration, some of the oxygenated blood delivered by the superior vena cava cannula reaches the inferior vena cava cannula, creating a “shunt,” also known as “recirculation.”

Currently, a double-lumen cannula is preferred. This type of cannula is placed in the right internal jugular vein with the tip advanced to the inferior vena cava so that blood is drained through one lumen from both the inferior and superior vena cavas and returned via the other lumen with the jet directed over the tricuspid valve. Advantages of this system are that as it delivers more oxygen to the pulmonary arteries it reduces recirculation, it requires only a single cannula to be inserted, and it facilitates ambulation and rehabilitation in patients requiring long-term ECMO.

A newer double-lumen cannula designed to drain venous blood from the right atrium and reinfuse it directly into the pulmonary artery may provide an alternative for patients with right ventricular failure.

Extracorporeal removal of carbon dioxide

ECMO can remove carbon dioxide in patients with hypercapneic respiratory failure. Early technology used a variation of venovenous ECMO with very low blood flow rates through the pump, which allowed use of smaller cannulas while efficiently removing carbon dioxide.²³

Since then, a pumpless extracorporeal lung-assist device has been developed that uses an arteriovenous configuration with two smaller cannulas inserted into the femoral artery and vein (Novalung, Germany).²⁴ Lacking a pump, it avoids the complications associated with pumps such as hemolysis and clotting. It effectively removes carbon dioxide and helps reduce the frequency and intensity of mechanical ventilation. Since the flow is driven by the patient’s arteriovenous pressure gradient, good cardiac output is a prerequisite for its use.

A portable low-blood-flow machine that uses a very small (ie, 15-F) catheter in the venovenous configuration is under investigation (Hemolung RAS, Alung Technologies).

WHO CAN BENEFIT FROM ECMO?

Although evidence to support the routine use of ECMO is limited, tools and guidelines have been developed to help clinicians decide if a patient might benefit from it. Indications for and contraindications to ECMO are shown in Table 2.

The Extracorporeal Life Support Organization recommends considering ECMO if the predicted risk of death is greater than 50% without it, and says ECMO is indicated if the predicted risk exceeds 80%. A scoring system has been developed to help predict the risk of death in patients on ECMO.¹⁴ This system has been validated using a historical cohort of patients, and current studies are ongoing for prospective validation.

Many centers are now using ECMO as a salvage therapy in patients with severe respiratory failure when conventional mechanical ventilation and adjunctive therapies such as neuromuscular blockade, inhaled nitric oxide, steroids, prone positioning, and high-frequency oscillation therapy fail to improve gas exchange.^25,26

ECMO is also indicated in hypercapneic respiratory failure secondary to status asthmaticus and exacerbation of chronic obstructive pulmonary disease, permissive hypercapnea with a Paco₂ greater than 80 mm Hg, or inability to achieve safe inflation pressures with plateau pressures of 30 cm H₂O or higher, refractory to conventional therapy.²⁷

Sometimes, delay in referral leads to irreversible ventilator-induced lung injury due to intense mechanical ventilation, thus limiting the utility of ECMO.⁸ Early referral should be considered if the patient does not improve after a few days on optimal ventilator settings. In centers where this technology is not available, referral to the nearest ECMO center should be considered. A list of certified ECMO centers is available at www.elso.org/Members/CenterDirectory.aspx.

Contraindications to ECMO

Advanced age, comorbid conditions such as malignancy, nonpulmonary organ dysfunction (including complications of critical illness), and immunodeficiency or pharmacologic immune suppression have been associated with poor outcomes in ECMO patients.²⁸ Severe aortic incompetence and aortic dissection are contraindications, since ventricular end-diastolic pressure can be increased with resultant ventricular distention, compromised myocardial oxygenation, and worsening of left heart failure.

ECMO is increasingly being used in situations in which it was previously considered contraindicated. Pregnant and postpartum patients with cardiorespiratory failure were previously not considered for ECMO because of a possible increased risk of coagulopathy and complications. However, a recent review showed that the outcomes of ECMO in pregnancy and postpartum were similar to those in nonpregnant patients, and the risk of catastrophic bleeding was minor.²⁹

Similarly, ECMO is also being used increasingly in posttrauma patients and patients with other bleeding risks.³⁰

Morbid obesity was once considered a contraindication because of difficulty in cannulation, but with newer types of cannulas, even patients with a body mass index greater than 60 kg/m² are receiving ECMO.³¹

HOW DO YOU DO IT?

Figures 3 and 4 depict clinical decision-making in starting and weaning from ECMO in respiratory failure and cardiogenic shock, respectively.

Management of patients on ECMO

Appropriate patient selection and initiation of ECMO are only the beginning of a tough journey. Successful management requires minimizing lung injury from mechanical ventilation, careful monitoring of anticoagulation, and instituting adequate physical therapy, including ambulation when possible (Table 3).

Initial ECMO settings and monitoring

The cannulas for venovenous ECMO are frequently inserted under fluoroscopic or transesophageal echocardiographic guidance, whereas venoarterial ECMO cannulation does not require imaging and can be performed at the bedside in the intensive care unit or operating room.

The initial ECMO settings are titrated according to the patient’s hemodynamic and respiratory needs. There are three main variables: blood flow, fraction of oxygen in the sweep gas, and sweep gas flow rate. These are adjusted to achieve desirable levels of oxygen and carbon dioxide in the blood.

Blood flow is determined by the revolutions per minute of the pump, preload, and afterload of the circuit. Common patient conditions that may reduce flow are systemic hypertension, hypovolemia, cardiac tamponade, and tension pneumothorax, depending on the modality. In addition, mechanical factors such as clots in the oxygenator or kinks in the circuit can increase resistance and reduce flow. Resistance to flow is directly proportional to cannula lengths and inversely proportional to cannula radius to the fourth power. The greater the flow, the greater the oxygen delivery.

Fraction of oxygen in the sweep gas. The oxygenator has a gas blender that mixes air and oxygen and allows for a range of oxygen concentrations. Increases in fraction of oxygen increase the partial pressure of oxygen in the blood.

Sweep gas flow rate. Venous blood in the extracorporeal circuit is exposed to fresh gas (or sweep gas) that oxygenates the blood and removes carbon dioxide by diffusion. Increasing the sweep gas flow rate results in greater carbon dioxide elimination from the blood.

Laboratory monitoring. During ECMO, the following values are monitored frequently:

• Arterial blood gases
• Blood gases in the ECMO circuit before and after going through the oxygenator— to monitor the efficacy of the oxygenator membrane
• Lactic acid—to monitor for tissue hypoxia
• Plasma free hemoglobin (a marker of hemolysis)—to monitor for hemolysis.

Mechanical ventilation on ECMO

Low tidal volume ventilation greatly reduces the risk of death in patients on ECMO by reducing ventilator-induced lung injury. Proponents of ECMO believe that ECMO provides “lung rest,” and thus it is imperative that lung-protective ventilation strategies be followed in patients on ECMO.⁸ In most cases, after ECMO is started, low tidal volume ventilation (6 mL/kg) is possible and should be used—or even very low tidal volume ventilation (3–6 mL/kg).^32,33 Many cases have also been described in which patients have been safely extubated while on ECMO to prevent ventilator-induced lung injury.^34,35

If hypoxemia persists

Despite full support with venovenous ECMO, some patients remain hypoxemic due to inadequate blood flow to match metabolic demands, eg, patients with morbid obesity or severe sepsis and fever. The physician should ensure there is no recirculation, maximize blood flow, optimize the hematocrit to increase oxygen delivery, and consider ways to decrease oxygen consumption, including sedation, paralysis, and hypothermia.

Recirculation can be calculated by measuring the oxygen saturation of the blood in the ECMO machine before and after it goes through the oxygenator, and also in the central venous blood. Recirculation has been reduced by using double-lumen cannulas but can also be reduced by manipulation of the reinfusion cannula or increasing the distance between drainage and reinfusion ports in other configurations of venovenous ECMO.

Expert opinion suggests that oxygen saturation of 86% or more and Pao₂ of 55 mm Hg or more in patients on venovenous ECMO are sufficient to prevent hypoxia-related end-organ injury.³⁶ Venoarterial ECMO should be considered in patients on venovenous ECMO with refractory hypoxemia with the above measures.

Harlequin syndrome is characterized by upper body hypoxia resulting in cerebral hypoxemia due to poorly oxygenated blood in the coronary and cerebral circulations, especially in patients on peripheral venoarterial ECMO. It can be detected by sampling the blood in the arm (where the oxygen isn’t going) instead of the leg (where the oxygen is going), and it can be corrected by adjusting the Fio₂, using positive end-expiratory pressure, or both to increase oxygenation. If ventilator settings do not improve this syndrome, the arterial cannulation site can be switched from the femoral artery to the axillary or carotid artery.

Alternatively, a mixed-configuration venoarterial-venous ECMO can also be created, in which a portion of arterialized blood from the arterial outflow cannula is diverted via the right internal jugular artery to the right heart. This enriches the blood traveling through the pulmonary circulation and to the left ventricle to provide better oxygen delivery to the coronary and cerebral circulations.

Anticoagulation monitoring and transfusions

Anticoagulation is necessary to maintain a clot-free and functional circuit. Most clots develop in the oxygenator membrane, where they can prevent optimal gas exchange and, rarely, lead to embolization to the systemic circulation. However, reports have suggested that anticoagulation can be held for short periods on ECMO if necessary.

Unfractionated heparin is usually used for anticoagulation. Commonly used tests to monitor anticoagulation are the augmented partial thromboplastin time, activated clotting time, and anti-factor Xa levels. Lately, thromboelastography analysis is being used to comprehensively monitor various components of the coagulation cascade.³⁷ Anticoagulation is usually tailored to whether there are clots in the circuit, coagulopathy, and bleeding while on ECMO.³⁸

Traditionally, blood products were used liberally during ECMO to maintain a normal hematocrit and improve oxygen delivery, although recent data suggest that outcomes may be similar with conservative use of blood products.^39,40

Fluid management on ECMO

ECMO patients are fluid-overloaded due to a profound inflammatory response, cardiac failure, or both. Studies have shown that conservative fluid management improve lung function and shortens the duration of mechanical ventilation and intensive care in patients with lung injury.⁴¹ Hence, the patient’s net fluid balance should be kept negative, provided renal and hemodynamic parameters remain stable.

There is a high incidence of acute kidney injury in ECMO patients, and fluid overload is one of the main indications for renal replacement therapy.⁴² Continuous renal replacement therapy can be provided either by an in-line hemofilter or by incorporating a standard continuous renal replacement therapy machine into the ECMO circuit. There are no studies comparing the efficacy of these techniques, but they allow for rapid improvement in fluid balance and electrolyte disturbances and are commonly used in ECMO patients.^42,43

Physical rehabilitation and ambulation on ECMO

Physical rehabilitation in mechanically ventilated patients has been shown to reduce ventilator days and stay in the intensive care unit.⁴⁴ With the use of internal jugular double-lumen cannulas for venovenous ECMO and improvement in durability of the ECMO circuit, several centers are implementing physical rehabilitation and ambulation for patients while on ECMO. Current data suggest that physical therapy is safe for patients receiving ECMO and may accelerate the weaning process and shorten length of stay in the hospital after ECMO.^45,46 Aggressive rehabilitation is especially beneficial in patients awaiting lung transplant and may improve posttransplant recovery and outcomes.⁴⁷

Weaning from ECMO

There are no standard guidelines for weaning from venovenous or venoarterial ECMO. Once the underlying condition for which ECMO was initiated has improved, weaning can begin by reducing the blood flow rate, the flow rate of the sweep gas, or both.

Weaning from venovenous ECMO should be started when there is improvement in lung compliance, tidal volumes, and oxygenation. Once the circuit flow rate is reduced to less than 3 L/minute, ventilator settings are adjusted to standard lung-protective settings. ECMO support is gradually decreased by reducing the flow rate of sweep gas to less than 2 L/minute. If tidal volumes, respiratory rate, and gas exchange remain adequate after approximately 2 to 4 hours on a low rate of sweep gas, the patient can be weaned off the venovenous ECMO circuit.

Weaning from venoarterial ECMO should be considered when there is myocardial recovery with improved pulse pressure and contractility on echocardiography. This is done by reducing flow rates in increments of 0.5 to 2 L/minute over 24 to 36 hours and monitoring mean arterial pressures, central venous pressure, and myocardial contractility. When acceptable, patients are mostly weaned in a surgical setting. Prolonged periods on a low rate of blood flow are avoided to prevent thrombus formation in the circuit.

COMPLICATIONS OF ECMO

ECMO use can be associated with a myriad of patient and mechanical complications.

Hemorrhage is the most common complication encountered in ECMO, occurring in  approximately 43% of patients.²⁹ It occurs most frequently from cannulation and surgical sites. Although rare, potentially life-threatening pulmonary hemorrhage (including bleeding at the tracheostomy site), intracranial hemorrhage, and gastrointestinal hemorrhage have also been reported.³⁰

Infections, including new infection and worsening sepsis in patients with acute respiratory distress syndrome secondary to infection, are common in patients on ECMO.⁴⁸

Renal failure secondary to acute tubular necrosis requiring hemodialysis has been reported to occur in 13% of patients on ECMO.³⁰

Other complications of concern, especially in patients on venoarterial ECMO, are lower limb ischemia and thromboembolism associated with site of cannulation and direction of blood flow.⁴⁹ Mechanical complications include inappropriate placement of the cannula leading to insufficient oxygenation, injury to vessel walls, and rarely myocardial wall rupture; thrombus formation within the circuit causing failure of the oxygenator and sometimes, pulmonary or systemic embolism; and air embolism from the circuit.³⁶

NOT SUITED FOR ALL

Despite limited data to support its use, there has been a recent increase in utilization of ECMO to support critically ill adult patients with cardiopulmonary failure. ECMO support is not suited for all patients. Careful selection of patients should be done to optimize resource utilization and provide the best opportunity for recovery or transplant.

Extracorporeal membrance oxygenation (ECMO) provides temporary cardiorespiratory support for patients with severe respiratory or cardiac failure refractory to conventional therapy.¹ It can be configured to provide oxygen, remove carbon dioxide, support perfusion, or all of the above. It may provide a bridge to recovery in patients with acute cardiopulmonary failure or to heart or lung transplant.

Developed in the 1970s, ECMO has proven effective and is widely used in children with respiratory and cardiopulmonary failure.²  However, it remained little used in adults, as early randomized trials showed higher rates of complications in adults who received it and no survival advantage.^3,4 Proponents of using it in adult patients believe that these poor outcomes were at least partially due to limited training, intensive anticoagulation, and excessive volume and pressure during mechanical ventilation. Although ECMO technology has improved substantially in the last decade and survival rates have improved (www.elso.org), evidence to support its routine use in adults remains limited.

Nevertheless, about 14,000 adult patients received ECMO between 1990 and 2014, with a rate of survival to discharge of 57% for those in respiratory failure and 41% for those in cardiac failure.⁵ Its use increased 433% in the United States from 2006 to 2011.⁶

A national survey of critical care physicians and trainees in the United States found they had limited knowledge about ECMO technology and wanted to include specific educational objectives about it in their training.⁷

This article summarizes the principles of ECMO, including practical aspects such as patient selection, monitoring, and complications.

LIMITED EVIDENCE OF BENEFIT FROM CONTROLLED TRIALS

There is limited evidence from randomized controlled trials that ECMO is beneficial in adults.

In acute respiratory failure, the first randomized trial of ECMO in adults was conducted in 1979 in multiple medical centers.³ The survival rate was no higher with ECMO than with mechanical ventilation alone, and complication rates were very high.

Similarly, Morris et al⁴ performed a single-center trial comparing pressure-controlled inverse-ratio ventilation and extracorporeal carbon dioxide removal in patients with acute respiratory distress syndrome, which showed no survival benefit.

After these two early trials, ECMO was largely abandoned, and not until 2009 did a multicenter randomized trial in acute respiratory distress syndrome⁸ rejuvenate interest in its use. Although the trial did not conclusively prove that ECMO was more effective than conventional mechanical ventilation, the findings supported early referral to tertiary care centers with ECMO expertise, and the survival rate was substantially higher than in previous studies. A concise summary of randomized trials and retrospective studies utilizing ECMO in respiratory failure is shown in Table 1.^8–14

During the global pandemic of influenza H1N1 in 2009–2010, several centers reported survival benefits from ECMO in patients with severe acute respiratory distress syndrome secondary to influenza.^{9–12,15–19} Two retrospective case-control studies reported lower mortality rates when H1N1 patients were transferred to ECMO centers¹⁰ and among younger patients with H1N1 who received ECMO.¹²

Ongoing trials (ClinicalTrials.gov identifier NCT01470703) may provide definitive evidence for the effectiveness of ECMO as a rescue therapy in acute respiratory distress syndrome.

In cardiogenic shock, single-center retrospective and observational studies have reported better outcomes for patients who received ECMO for cardiogenic shock secondary to myocardial infarction, pulmonary embolism, sepsis-related cardiomyopathy, and even extracorporeal cardiopulmonary resuscitation.²⁰

WHAT IS ECMO?

In ECMO, venous blood is shunted through a machine to add oxygen, remove carbon dioxide, and regulate temperature (Figure 1). The components of an ECMO circuit are as follows:

• Blood pump
• Membrane oxygenator
• Gas mixer
• Cannulas
• Heater/cooler
• Console.

TWO BASIC CONFIGURATIONS

Two basic ECMO configurations are used in adults: venoarterial and venovenous,²¹ although combinations of the two—hybrid configurations—are sometimes used (Figure 2).

Venoarterial ECMO

Venoarterial ECMO provides complete or partial support to the heart and lungs and is the configuration of choice in patients with isolated cardiac failure that is refractory to other treatments. It takes deoxygenated blood from the venous system and returns oxygenated blood to the arterial circulation.

In the central venoarterial configuration, the intake cannula is most often surgically placed in the right atrium and the return cannula is placed in the proximal ascending aorta.

In the peripheral femoral configuration, the drainage cannula is placed in the femoral vein and advanced to the right atrium, and the return cannula is placed in either the ipsilateral or contralateral femoral artery. However, this configuration provides the patient with retrograde flow (against the native cardiac output), and oxygen delivery to the upper body may be impeded.

Axillary cannulation, in which the return cannula is placed directly into the axillary artery to provide antegrade flow, has been used recently in patients with pulmonary hypertension or right ventricular failure.²²

Venovenous ECMO

Venovenous ECMO provides complete or partial support to the lungs and is the configuration of choice in isolated respiratory failure when cardiac function is preserved. It takes deoxygenated blood from the central venous system—either the femoral vein or internal jugular vein—and returns oxygenated blood to the venous circulation directed into the right atrium. It can be delivered by different cannula configurations based on the patient’s size and clinical requirements.

In the past, the most commonly used configuration was the femoral-atrial, in which the drainage cannula was placed in the femoral vein with the tip advanced to the level of the diaphragm in the inferior vena cava, and the return cannula was placed in the right internal jugular vein with its tip at the junction of the superior vena cava and right atrium. In this configuration, some of the oxygenated blood delivered by the superior vena cava cannula reaches the inferior vena cava cannula, creating a “shunt,” also known as “recirculation.”

Currently, a double-lumen cannula is preferred. This type of cannula is placed in the right internal jugular vein with the tip advanced to the inferior vena cava so that blood is drained through one lumen from both the inferior and superior vena cavas and returned via the other lumen with the jet directed over the tricuspid valve. Advantages of this system are that as it delivers more oxygen to the pulmonary arteries it reduces recirculation, it requires only a single cannula to be inserted, and it facilitates ambulation and rehabilitation in patients requiring long-term ECMO.

A newer double-lumen cannula designed to drain venous blood from the right atrium and reinfuse it directly into the pulmonary artery may provide an alternative for patients with right ventricular failure.

Extracorporeal removal of carbon dioxide

ECMO can remove carbon dioxide in patients with hypercapneic respiratory failure. Early technology used a variation of venovenous ECMO with very low blood flow rates through the pump, which allowed use of smaller cannulas while efficiently removing carbon dioxide.²³

Since then, a pumpless extracorporeal lung-assist device has been developed that uses an arteriovenous configuration with two smaller cannulas inserted into the femoral artery and vein (Novalung, Germany).²⁴ Lacking a pump, it avoids the complications associated with pumps such as hemolysis and clotting. It effectively removes carbon dioxide and helps reduce the frequency and intensity of mechanical ventilation. Since the flow is driven by the patient’s arteriovenous pressure gradient, good cardiac output is a prerequisite for its use.

A portable low-blood-flow machine that uses a very small (ie, 15-F) catheter in the venovenous configuration is under investigation (Hemolung RAS, Alung Technologies).

WHO CAN BENEFIT FROM ECMO?

Although evidence to support the routine use of ECMO is limited, tools and guidelines have been developed to help clinicians decide if a patient might benefit from it. Indications for and contraindications to ECMO are shown in Table 2.

The Extracorporeal Life Support Organization recommends considering ECMO if the predicted risk of death is greater than 50% without it, and says ECMO is indicated if the predicted risk exceeds 80%. A scoring system has been developed to help predict the risk of death in patients on ECMO.¹⁴ This system has been validated using a historical cohort of patients, and current studies are ongoing for prospective validation.

Many centers are now using ECMO as a salvage therapy in patients with severe respiratory failure when conventional mechanical ventilation and adjunctive therapies such as neuromuscular blockade, inhaled nitric oxide, steroids, prone positioning, and high-frequency oscillation therapy fail to improve gas exchange.^25,26

ECMO is also indicated in hypercapneic respiratory failure secondary to status asthmaticus and exacerbation of chronic obstructive pulmonary disease, permissive hypercapnea with a Paco₂ greater than 80 mm Hg, or inability to achieve safe inflation pressures with plateau pressures of 30 cm H₂O or higher, refractory to conventional therapy.²⁷

Sometimes, delay in referral leads to irreversible ventilator-induced lung injury due to intense mechanical ventilation, thus limiting the utility of ECMO.⁸ Early referral should be considered if the patient does not improve after a few days on optimal ventilator settings. In centers where this technology is not available, referral to the nearest ECMO center should be considered. A list of certified ECMO centers is available at www.elso.org/Members/CenterDirectory.aspx.

Contraindications to ECMO

Advanced age, comorbid conditions such as malignancy, nonpulmonary organ dysfunction (including complications of critical illness), and immunodeficiency or pharmacologic immune suppression have been associated with poor outcomes in ECMO patients.²⁸ Severe aortic incompetence and aortic dissection are contraindications, since ventricular end-diastolic pressure can be increased with resultant ventricular distention, compromised myocardial oxygenation, and worsening of left heart failure.

ECMO is increasingly being used in situations in which it was previously considered contraindicated. Pregnant and postpartum patients with cardiorespiratory failure were previously not considered for ECMO because of a possible increased risk of coagulopathy and complications. However, a recent review showed that the outcomes of ECMO in pregnancy and postpartum were similar to those in nonpregnant patients, and the risk of catastrophic bleeding was minor.²⁹

Similarly, ECMO is also being used increasingly in posttrauma patients and patients with other bleeding risks.³⁰

Morbid obesity was once considered a contraindication because of difficulty in cannulation, but with newer types of cannulas, even patients with a body mass index greater than 60 kg/m² are receiving ECMO.³¹

HOW DO YOU DO IT?

Figures 3 and 4 depict clinical decision-making in starting and weaning from ECMO in respiratory failure and cardiogenic shock, respectively.

Management of patients on ECMO

Appropriate patient selection and initiation of ECMO are only the beginning of a tough journey. Successful management requires minimizing lung injury from mechanical ventilation, careful monitoring of anticoagulation, and instituting adequate physical therapy, including ambulation when possible (Table 3).

Initial ECMO settings and monitoring

The cannulas for venovenous ECMO are frequently inserted under fluoroscopic or transesophageal echocardiographic guidance, whereas venoarterial ECMO cannulation does not require imaging and can be performed at the bedside in the intensive care unit or operating room.

The initial ECMO settings are titrated according to the patient’s hemodynamic and respiratory needs. There are three main variables: blood flow, fraction of oxygen in the sweep gas, and sweep gas flow rate. These are adjusted to achieve desirable levels of oxygen and carbon dioxide in the blood.

Blood flow is determined by the revolutions per minute of the pump, preload, and afterload of the circuit. Common patient conditions that may reduce flow are systemic hypertension, hypovolemia, cardiac tamponade, and tension pneumothorax, depending on the modality. In addition, mechanical factors such as clots in the oxygenator or kinks in the circuit can increase resistance and reduce flow. Resistance to flow is directly proportional to cannula lengths and inversely proportional to cannula radius to the fourth power. The greater the flow, the greater the oxygen delivery.

Fraction of oxygen in the sweep gas. The oxygenator has a gas blender that mixes air and oxygen and allows for a range of oxygen concentrations. Increases in fraction of oxygen increase the partial pressure of oxygen in the blood.

Sweep gas flow rate. Venous blood in the extracorporeal circuit is exposed to fresh gas (or sweep gas) that oxygenates the blood and removes carbon dioxide by diffusion. Increasing the sweep gas flow rate results in greater carbon dioxide elimination from the blood.

Laboratory monitoring. During ECMO, the following values are monitored frequently:

• Arterial blood gases
• Blood gases in the ECMO circuit before and after going through the oxygenator— to monitor the efficacy of the oxygenator membrane
• Lactic acid—to monitor for tissue hypoxia
• Plasma free hemoglobin (a marker of hemolysis)—to monitor for hemolysis.

Mechanical ventilation on ECMO

Low tidal volume ventilation greatly reduces the risk of death in patients on ECMO by reducing ventilator-induced lung injury. Proponents of ECMO believe that ECMO provides “lung rest,” and thus it is imperative that lung-protective ventilation strategies be followed in patients on ECMO.⁸ In most cases, after ECMO is started, low tidal volume ventilation (6 mL/kg) is possible and should be used—or even very low tidal volume ventilation (3–6 mL/kg).^32,33 Many cases have also been described in which patients have been safely extubated while on ECMO to prevent ventilator-induced lung injury.^34,35

If hypoxemia persists

Despite full support with venovenous ECMO, some patients remain hypoxemic due to inadequate blood flow to match metabolic demands, eg, patients with morbid obesity or severe sepsis and fever. The physician should ensure there is no recirculation, maximize blood flow, optimize the hematocrit to increase oxygen delivery, and consider ways to decrease oxygen consumption, including sedation, paralysis, and hypothermia.

Recirculation can be calculated by measuring the oxygen saturation of the blood in the ECMO machine before and after it goes through the oxygenator, and also in the central venous blood. Recirculation has been reduced by using double-lumen cannulas but can also be reduced by manipulation of the reinfusion cannula or increasing the distance between drainage and reinfusion ports in other configurations of venovenous ECMO.

Expert opinion suggests that oxygen saturation of 86% or more and Pao₂ of 55 mm Hg or more in patients on venovenous ECMO are sufficient to prevent hypoxia-related end-organ injury.³⁶ Venoarterial ECMO should be considered in patients on venovenous ECMO with refractory hypoxemia with the above measures.

Harlequin syndrome is characterized by upper body hypoxia resulting in cerebral hypoxemia due to poorly oxygenated blood in the coronary and cerebral circulations, especially in patients on peripheral venoarterial ECMO. It can be detected by sampling the blood in the arm (where the oxygen isn’t going) instead of the leg (where the oxygen is going), and it can be corrected by adjusting the Fio₂, using positive end-expiratory pressure, or both to increase oxygenation. If ventilator settings do not improve this syndrome, the arterial cannulation site can be switched from the femoral artery to the axillary or carotid artery.

Alternatively, a mixed-configuration venoarterial-venous ECMO can also be created, in which a portion of arterialized blood from the arterial outflow cannula is diverted via the right internal jugular artery to the right heart. This enriches the blood traveling through the pulmonary circulation and to the left ventricle to provide better oxygen delivery to the coronary and cerebral circulations.

Anticoagulation monitoring and transfusions

Anticoagulation is necessary to maintain a clot-free and functional circuit. Most clots develop in the oxygenator membrane, where they can prevent optimal gas exchange and, rarely, lead to embolization to the systemic circulation. However, reports have suggested that anticoagulation can be held for short periods on ECMO if necessary.

Unfractionated heparin is usually used for anticoagulation. Commonly used tests to monitor anticoagulation are the augmented partial thromboplastin time, activated clotting time, and anti-factor Xa levels. Lately, thromboelastography analysis is being used to comprehensively monitor various components of the coagulation cascade.³⁷ Anticoagulation is usually tailored to whether there are clots in the circuit, coagulopathy, and bleeding while on ECMO.³⁸

Traditionally, blood products were used liberally during ECMO to maintain a normal hematocrit and improve oxygen delivery, although recent data suggest that outcomes may be similar with conservative use of blood products.^39,40

Fluid management on ECMO

ECMO patients are fluid-overloaded due to a profound inflammatory response, cardiac failure, or both. Studies have shown that conservative fluid management improve lung function and shortens the duration of mechanical ventilation and intensive care in patients with lung injury.⁴¹ Hence, the patient’s net fluid balance should be kept negative, provided renal and hemodynamic parameters remain stable.

There is a high incidence of acute kidney injury in ECMO patients, and fluid overload is one of the main indications for renal replacement therapy.⁴² Continuous renal replacement therapy can be provided either by an in-line hemofilter or by incorporating a standard continuous renal replacement therapy machine into the ECMO circuit. There are no studies comparing the efficacy of these techniques, but they allow for rapid improvement in fluid balance and electrolyte disturbances and are commonly used in ECMO patients.^42,43

Physical rehabilitation and ambulation on ECMO

Physical rehabilitation in mechanically ventilated patients has been shown to reduce ventilator days and stay in the intensive care unit.⁴⁴ With the use of internal jugular double-lumen cannulas for venovenous ECMO and improvement in durability of the ECMO circuit, several centers are implementing physical rehabilitation and ambulation for patients while on ECMO. Current data suggest that physical therapy is safe for patients receiving ECMO and may accelerate the weaning process and shorten length of stay in the hospital after ECMO.^45,46 Aggressive rehabilitation is especially beneficial in patients awaiting lung transplant and may improve posttransplant recovery and outcomes.⁴⁷

Weaning from ECMO

There are no standard guidelines for weaning from venovenous or venoarterial ECMO. Once the underlying condition for which ECMO was initiated has improved, weaning can begin by reducing the blood flow rate, the flow rate of the sweep gas, or both.

Weaning from venovenous ECMO should be started when there is improvement in lung compliance, tidal volumes, and oxygenation. Once the circuit flow rate is reduced to less than 3 L/minute, ventilator settings are adjusted to standard lung-protective settings. ECMO support is gradually decreased by reducing the flow rate of sweep gas to less than 2 L/minute. If tidal volumes, respiratory rate, and gas exchange remain adequate after approximately 2 to 4 hours on a low rate of sweep gas, the patient can be weaned off the venovenous ECMO circuit.

Weaning from venoarterial ECMO should be considered when there is myocardial recovery with improved pulse pressure and contractility on echocardiography. This is done by reducing flow rates in increments of 0.5 to 2 L/minute over 24 to 36 hours and monitoring mean arterial pressures, central venous pressure, and myocardial contractility. When acceptable, patients are mostly weaned in a surgical setting. Prolonged periods on a low rate of blood flow are avoided to prevent thrombus formation in the circuit.

COMPLICATIONS OF ECMO

ECMO use can be associated with a myriad of patient and mechanical complications.

Hemorrhage is the most common complication encountered in ECMO, occurring in  approximately 43% of patients.²⁹ It occurs most frequently from cannulation and surgical sites. Although rare, potentially life-threatening pulmonary hemorrhage (including bleeding at the tracheostomy site), intracranial hemorrhage, and gastrointestinal hemorrhage have also been reported.³⁰

Infections, including new infection and worsening sepsis in patients with acute respiratory distress syndrome secondary to infection, are common in patients on ECMO.⁴⁸

Renal failure secondary to acute tubular necrosis requiring hemodialysis has been reported to occur in 13% of patients on ECMO.³⁰

Other complications of concern, especially in patients on venoarterial ECMO, are lower limb ischemia and thromboembolism associated with site of cannulation and direction of blood flow.⁴⁹ Mechanical complications include inappropriate placement of the cannula leading to insufficient oxygenation, injury to vessel walls, and rarely myocardial wall rupture; thrombus formation within the circuit causing failure of the oxygenator and sometimes, pulmonary or systemic embolism; and air embolism from the circuit.³⁶

NOT SUITED FOR ALL

Despite limited data to support its use, there has been a recent increase in utilization of ECMO to support critically ill adult patients with cardiopulmonary failure. ECMO support is not suited for all patients. Careful selection of patients should be done to optimize resource utilization and provide the best opportunity for recovery or transplant.

Extracorporeal membrance oxygenation (ECMO) provides temporary cardiorespiratory support for patients with severe respiratory or cardiac failure refractory to conventional therapy.¹ It can be configured to provide oxygen, remove carbon dioxide, support perfusion, or all of the above. It may provide a bridge to recovery in patients with acute cardiopulmonary failure or to heart or lung transplant.

Developed in the 1970s, ECMO has proven effective and is widely used in children with respiratory and cardiopulmonary failure.²  However, it remained little used in adults, as early randomized trials showed higher rates of complications in adults who received it and no survival advantage.^3,4 Proponents of using it in adult patients believe that these poor outcomes were at least partially due to limited training, intensive anticoagulation, and excessive volume and pressure during mechanical ventilation. Although ECMO technology has improved substantially in the last decade and survival rates have improved (www.elso.org), evidence to support its routine use in adults remains limited.

Nevertheless, about 14,000 adult patients received ECMO between 1990 and 2014, with a rate of survival to discharge of 57% for those in respiratory failure and 41% for those in cardiac failure.⁵ Its use increased 433% in the United States from 2006 to 2011.⁶

A national survey of critical care physicians and trainees in the United States found they had limited knowledge about ECMO technology and wanted to include specific educational objectives about it in their training.⁷

This article summarizes the principles of ECMO, including practical aspects such as patient selection, monitoring, and complications.

LIMITED EVIDENCE OF BENEFIT FROM CONTROLLED TRIALS

There is limited evidence from randomized controlled trials that ECMO is beneficial in adults.

In acute respiratory failure, the first randomized trial of ECMO in adults was conducted in 1979 in multiple medical centers.³ The survival rate was no higher with ECMO than with mechanical ventilation alone, and complication rates were very high.

Similarly, Morris et al⁴ performed a single-center trial comparing pressure-controlled inverse-ratio ventilation and extracorporeal carbon dioxide removal in patients with acute respiratory distress syndrome, which showed no survival benefit.

After these two early trials, ECMO was largely abandoned, and not until 2009 did a multicenter randomized trial in acute respiratory distress syndrome⁸ rejuvenate interest in its use. Although the trial did not conclusively prove that ECMO was more effective than conventional mechanical ventilation, the findings supported early referral to tertiary care centers with ECMO expertise, and the survival rate was substantially higher than in previous studies. A concise summary of randomized trials and retrospective studies utilizing ECMO in respiratory failure is shown in Table 1.^8–14

During the global pandemic of influenza H1N1 in 2009–2010, several centers reported survival benefits from ECMO in patients with severe acute respiratory distress syndrome secondary to influenza.^{9–12,15–19} Two retrospective case-control studies reported lower mortality rates when H1N1 patients were transferred to ECMO centers¹⁰ and among younger patients with H1N1 who received ECMO.¹²

Ongoing trials (ClinicalTrials.gov identifier NCT01470703) may provide definitive evidence for the effectiveness of ECMO as a rescue therapy in acute respiratory distress syndrome.

In cardiogenic shock, single-center retrospective and observational studies have reported better outcomes for patients who received ECMO for cardiogenic shock secondary to myocardial infarction, pulmonary embolism, sepsis-related cardiomyopathy, and even extracorporeal cardiopulmonary resuscitation.²⁰

WHAT IS ECMO?

In ECMO, venous blood is shunted through a machine to add oxygen, remove carbon dioxide, and regulate temperature (Figure 1). The components of an ECMO circuit are as follows:

• Blood pump
• Membrane oxygenator
• Gas mixer
• Cannulas
• Heater/cooler
• Console.

TWO BASIC CONFIGURATIONS

Two basic ECMO configurations are used in adults: venoarterial and venovenous,²¹ although combinations of the two—hybrid configurations—are sometimes used (Figure 2).

Venoarterial ECMO

Venoarterial ECMO provides complete or partial support to the heart and lungs and is the configuration of choice in patients with isolated cardiac failure that is refractory to other treatments. It takes deoxygenated blood from the venous system and returns oxygenated blood to the arterial circulation.

In the central venoarterial configuration, the intake cannula is most often surgically placed in the right atrium and the return cannula is placed in the proximal ascending aorta.

In the peripheral femoral configuration, the drainage cannula is placed in the femoral vein and advanced to the right atrium, and the return cannula is placed in either the ipsilateral or contralateral femoral artery. However, this configuration provides the patient with retrograde flow (against the native cardiac output), and oxygen delivery to the upper body may be impeded.

Axillary cannulation, in which the return cannula is placed directly into the axillary artery to provide antegrade flow, has been used recently in patients with pulmonary hypertension or right ventricular failure.²²

Venovenous ECMO

Venovenous ECMO provides complete or partial support to the lungs and is the configuration of choice in isolated respiratory failure when cardiac function is preserved. It takes deoxygenated blood from the central venous system—either the femoral vein or internal jugular vein—and returns oxygenated blood to the venous circulation directed into the right atrium. It can be delivered by different cannula configurations based on the patient’s size and clinical requirements.

In the past, the most commonly used configuration was the femoral-atrial, in which the drainage cannula was placed in the femoral vein with the tip advanced to the level of the diaphragm in the inferior vena cava, and the return cannula was placed in the right internal jugular vein with its tip at the junction of the superior vena cava and right atrium. In this configuration, some of the oxygenated blood delivered by the superior vena cava cannula reaches the inferior vena cava cannula, creating a “shunt,” also known as “recirculation.”

Currently, a double-lumen cannula is preferred. This type of cannula is placed in the right internal jugular vein with the tip advanced to the inferior vena cava so that blood is drained through one lumen from both the inferior and superior vena cavas and returned via the other lumen with the jet directed over the tricuspid valve. Advantages of this system are that as it delivers more oxygen to the pulmonary arteries it reduces recirculation, it requires only a single cannula to be inserted, and it facilitates ambulation and rehabilitation in patients requiring long-term ECMO.

A newer double-lumen cannula designed to drain venous blood from the right atrium and reinfuse it directly into the pulmonary artery may provide an alternative for patients with right ventricular failure.

Extracorporeal removal of carbon dioxide

ECMO can remove carbon dioxide in patients with hypercapneic respiratory failure. Early technology used a variation of venovenous ECMO with very low blood flow rates through the pump, which allowed use of smaller cannulas while efficiently removing carbon dioxide.²³

Since then, a pumpless extracorporeal lung-assist device has been developed that uses an arteriovenous configuration with two smaller cannulas inserted into the femoral artery and vein (Novalung, Germany).²⁴ Lacking a pump, it avoids the complications associated with pumps such as hemolysis and clotting. It effectively removes carbon dioxide and helps reduce the frequency and intensity of mechanical ventilation. Since the flow is driven by the patient’s arteriovenous pressure gradient, good cardiac output is a prerequisite for its use.

A portable low-blood-flow machine that uses a very small (ie, 15-F) catheter in the venovenous configuration is under investigation (Hemolung RAS, Alung Technologies).

WHO CAN BENEFIT FROM ECMO?

Although evidence to support the routine use of ECMO is limited, tools and guidelines have been developed to help clinicians decide if a patient might benefit from it. Indications for and contraindications to ECMO are shown in Table 2.

The Extracorporeal Life Support Organization recommends considering ECMO if the predicted risk of death is greater than 50% without it, and says ECMO is indicated if the predicted risk exceeds 80%. A scoring system has been developed to help predict the risk of death in patients on ECMO.¹⁴ This system has been validated using a historical cohort of patients, and current studies are ongoing for prospective validation.

Many centers are now using ECMO as a salvage therapy in patients with severe respiratory failure when conventional mechanical ventilation and adjunctive therapies such as neuromuscular blockade, inhaled nitric oxide, steroids, prone positioning, and high-frequency oscillation therapy fail to improve gas exchange.^25,26

ECMO is also indicated in hypercapneic respiratory failure secondary to status asthmaticus and exacerbation of chronic obstructive pulmonary disease, permissive hypercapnea with a Paco₂ greater than 80 mm Hg, or inability to achieve safe inflation pressures with plateau pressures of 30 cm H₂O or higher, refractory to conventional therapy.²⁷

Sometimes, delay in referral leads to irreversible ventilator-induced lung injury due to intense mechanical ventilation, thus limiting the utility of ECMO.⁸ Early referral should be considered if the patient does not improve after a few days on optimal ventilator settings. In centers where this technology is not available, referral to the nearest ECMO center should be considered. A list of certified ECMO centers is available at www.elso.org/Members/CenterDirectory.aspx.

Contraindications to ECMO

Advanced age, comorbid conditions such as malignancy, nonpulmonary organ dysfunction (including complications of critical illness), and immunodeficiency or pharmacologic immune suppression have been associated with poor outcomes in ECMO patients.²⁸ Severe aortic incompetence and aortic dissection are contraindications, since ventricular end-diastolic pressure can be increased with resultant ventricular distention, compromised myocardial oxygenation, and worsening of left heart failure.

ECMO is increasingly being used in situations in which it was previously considered contraindicated. Pregnant and postpartum patients with cardiorespiratory failure were previously not considered for ECMO because of a possible increased risk of coagulopathy and complications. However, a recent review showed that the outcomes of ECMO in pregnancy and postpartum were similar to those in nonpregnant patients, and the risk of catastrophic bleeding was minor.²⁹

Similarly, ECMO is also being used increasingly in posttrauma patients and patients with other bleeding risks.³⁰

Morbid obesity was once considered a contraindication because of difficulty in cannulation, but with newer types of cannulas, even patients with a body mass index greater than 60 kg/m² are receiving ECMO.³¹

HOW DO YOU DO IT?

Figures 3 and 4 depict clinical decision-making in starting and weaning from ECMO in respiratory failure and cardiogenic shock, respectively.

Management of patients on ECMO

Appropriate patient selection and initiation of ECMO are only the beginning of a tough journey. Successful management requires minimizing lung injury from mechanical ventilation, careful monitoring of anticoagulation, and instituting adequate physical therapy, including ambulation when possible (Table 3).

Initial ECMO settings and monitoring

The cannulas for venovenous ECMO are frequently inserted under fluoroscopic or transesophageal echocardiographic guidance, whereas venoarterial ECMO cannulation does not require imaging and can be performed at the bedside in the intensive care unit or operating room.

The initial ECMO settings are titrated according to the patient’s hemodynamic and respiratory needs. There are three main variables: blood flow, fraction of oxygen in the sweep gas, and sweep gas flow rate. These are adjusted to achieve desirable levels of oxygen and carbon dioxide in the blood.

Blood flow is determined by the revolutions per minute of the pump, preload, and afterload of the circuit. Common patient conditions that may reduce flow are systemic hypertension, hypovolemia, cardiac tamponade, and tension pneumothorax, depending on the modality. In addition, mechanical factors such as clots in the oxygenator or kinks in the circuit can increase resistance and reduce flow. Resistance to flow is directly proportional to cannula lengths and inversely proportional to cannula radius to the fourth power. The greater the flow, the greater the oxygen delivery.

Fraction of oxygen in the sweep gas. The oxygenator has a gas blender that mixes air and oxygen and allows for a range of oxygen concentrations. Increases in fraction of oxygen increase the partial pressure of oxygen in the blood.

Sweep gas flow rate. Venous blood in the extracorporeal circuit is exposed to fresh gas (or sweep gas) that oxygenates the blood and removes carbon dioxide by diffusion. Increasing the sweep gas flow rate results in greater carbon dioxide elimination from the blood.

Laboratory monitoring. During ECMO, the following values are monitored frequently:

• Arterial blood gases
• Blood gases in the ECMO circuit before and after going through the oxygenator— to monitor the efficacy of the oxygenator membrane
• Lactic acid—to monitor for tissue hypoxia
• Plasma free hemoglobin (a marker of hemolysis)—to monitor for hemolysis.

Mechanical ventilation on ECMO

Low tidal volume ventilation greatly reduces the risk of death in patients on ECMO by reducing ventilator-induced lung injury. Proponents of ECMO believe that ECMO provides “lung rest,” and thus it is imperative that lung-protective ventilation strategies be followed in patients on ECMO.⁸ In most cases, after ECMO is started, low tidal volume ventilation (6 mL/kg) is possible and should be used—or even very low tidal volume ventilation (3–6 mL/kg).^32,33 Many cases have also been described in which patients have been safely extubated while on ECMO to prevent ventilator-induced lung injury.^34,35

If hypoxemia persists

Despite full support with venovenous ECMO, some patients remain hypoxemic due to inadequate blood flow to match metabolic demands, eg, patients with morbid obesity or severe sepsis and fever. The physician should ensure there is no recirculation, maximize blood flow, optimize the hematocrit to increase oxygen delivery, and consider ways to decrease oxygen consumption, including sedation, paralysis, and hypothermia.

Recirculation can be calculated by measuring the oxygen saturation of the blood in the ECMO machine before and after it goes through the oxygenator, and also in the central venous blood. Recirculation has been reduced by using double-lumen cannulas but can also be reduced by manipulation of the reinfusion cannula or increasing the distance between drainage and reinfusion ports in other configurations of venovenous ECMO.

Expert opinion suggests that oxygen saturation of 86% or more and Pao₂ of 55 mm Hg or more in patients on venovenous ECMO are sufficient to prevent hypoxia-related end-organ injury.³⁶ Venoarterial ECMO should be considered in patients on venovenous ECMO with refractory hypoxemia with the above measures.

Harlequin syndrome is characterized by upper body hypoxia resulting in cerebral hypoxemia due to poorly oxygenated blood in the coronary and cerebral circulations, especially in patients on peripheral venoarterial ECMO. It can be detected by sampling the blood in the arm (where the oxygen isn’t going) instead of the leg (where the oxygen is going), and it can be corrected by adjusting the Fio₂, using positive end-expiratory pressure, or both to increase oxygenation. If ventilator settings do not improve this syndrome, the arterial cannulation site can be switched from the femoral artery to the axillary or carotid artery.

Alternatively, a mixed-configuration venoarterial-venous ECMO can also be created, in which a portion of arterialized blood from the arterial outflow cannula is diverted via the right internal jugular artery to the right heart. This enriches the blood traveling through the pulmonary circulation and to the left ventricle to provide better oxygen delivery to the coronary and cerebral circulations.

Anticoagulation monitoring and transfusions

Anticoagulation is necessary to maintain a clot-free and functional circuit. Most clots develop in the oxygenator membrane, where they can prevent optimal gas exchange and, rarely, lead to embolization to the systemic circulation. However, reports have suggested that anticoagulation can be held for short periods on ECMO if necessary.

Unfractionated heparin is usually used for anticoagulation. Commonly used tests to monitor anticoagulation are the augmented partial thromboplastin time, activated clotting time, and anti-factor Xa levels. Lately, thromboelastography analysis is being used to comprehensively monitor various components of the coagulation cascade.³⁷ Anticoagulation is usually tailored to whether there are clots in the circuit, coagulopathy, and bleeding while on ECMO.³⁸

Traditionally, blood products were used liberally during ECMO to maintain a normal hematocrit and improve oxygen delivery, although recent data suggest that outcomes may be similar with conservative use of blood products.^39,40

Fluid management on ECMO

ECMO patients are fluid-overloaded due to a profound inflammatory response, cardiac failure, or both. Studies have shown that conservative fluid management improve lung function and shortens the duration of mechanical ventilation and intensive care in patients with lung injury.⁴¹ Hence, the patient’s net fluid balance should be kept negative, provided renal and hemodynamic parameters remain stable.

There is a high incidence of acute kidney injury in ECMO patients, and fluid overload is one of the main indications for renal replacement therapy.⁴² Continuous renal replacement therapy can be provided either by an in-line hemofilter or by incorporating a standard continuous renal replacement therapy machine into the ECMO circuit. There are no studies comparing the efficacy of these techniques, but they allow for rapid improvement in fluid balance and electrolyte disturbances and are commonly used in ECMO patients.^42,43

Physical rehabilitation and ambulation on ECMO

Physical rehabilitation in mechanically ventilated patients has been shown to reduce ventilator days and stay in the intensive care unit.⁴⁴ With the use of internal jugular double-lumen cannulas for venovenous ECMO and improvement in durability of the ECMO circuit, several centers are implementing physical rehabilitation and ambulation for patients while on ECMO. Current data suggest that physical therapy is safe for patients receiving ECMO and may accelerate the weaning process and shorten length of stay in the hospital after ECMO.^45,46 Aggressive rehabilitation is especially beneficial in patients awaiting lung transplant and may improve posttransplant recovery and outcomes.⁴⁷

Weaning from ECMO

There are no standard guidelines for weaning from venovenous or venoarterial ECMO. Once the underlying condition for which ECMO was initiated has improved, weaning can begin by reducing the blood flow rate, the flow rate of the sweep gas, or both.

Weaning from venovenous ECMO should be started when there is improvement in lung compliance, tidal volumes, and oxygenation. Once the circuit flow rate is reduced to less than 3 L/minute, ventilator settings are adjusted to standard lung-protective settings. ECMO support is gradually decreased by reducing the flow rate of sweep gas to less than 2 L/minute. If tidal volumes, respiratory rate, and gas exchange remain adequate after approximately 2 to 4 hours on a low rate of sweep gas, the patient can be weaned off the venovenous ECMO circuit.

Weaning from venoarterial ECMO should be considered when there is myocardial recovery with improved pulse pressure and contractility on echocardiography. This is done by reducing flow rates in increments of 0.5 to 2 L/minute over 24 to 36 hours and monitoring mean arterial pressures, central venous pressure, and myocardial contractility. When acceptable, patients are mostly weaned in a surgical setting. Prolonged periods on a low rate of blood flow are avoided to prevent thrombus formation in the circuit.

COMPLICATIONS OF ECMO

ECMO use can be associated with a myriad of patient and mechanical complications.

Hemorrhage is the most common complication encountered in ECMO, occurring in  approximately 43% of patients.²⁹ It occurs most frequently from cannulation and surgical sites. Although rare, potentially life-threatening pulmonary hemorrhage (including bleeding at the tracheostomy site), intracranial hemorrhage, and gastrointestinal hemorrhage have also been reported.³⁰

Infections, including new infection and worsening sepsis in patients with acute respiratory distress syndrome secondary to infection, are common in patients on ECMO.⁴⁸

Renal failure secondary to acute tubular necrosis requiring hemodialysis has been reported to occur in 13% of patients on ECMO.³⁰

Other complications of concern, especially in patients on venoarterial ECMO, are lower limb ischemia and thromboembolism associated with site of cannulation and direction of blood flow.⁴⁹ Mechanical complications include inappropriate placement of the cannula leading to insufficient oxygenation, injury to vessel walls, and rarely myocardial wall rupture; thrombus formation within the circuit causing failure of the oxygenator and sometimes, pulmonary or systemic embolism; and air embolism from the circuit.³⁶

NOT SUITED FOR ALL

Despite limited data to support its use, there has been a recent increase in utilization of ECMO to support critically ill adult patients with cardiopulmonary failure. ECMO support is not suited for all patients. Careful selection of patients should be done to optimize resource utilization and provide the best opportunity for recovery or transplant.

A 28-year-old woman presents for routine follow-up of asthma. She has a 1-year-old son and is considering a second pregnancy. She says she read on the Internet that the US Food and Drug Administration recently banned baby bottles and “sippy” cups that contain bisphenol A (BPA), as recommended by the American Medical Association. She wonders if there are other sources of BPA and if they pose a health risk to her, her son, and possible future children.

Environmental toxins have been linked to pregnancy complications and poor birth outcomes. The past several decades have seen a significant rise in reproductive disorders such as early onset of puberty, low sperm count, and birth defects such as cryptorchidism and hypospadias.^1–4 These changes may be partly explained by other trends over time, such as older maternal age, rising incidence of obesity and maternal diabetes, and demographic changes that exacerbate health disparities. However, these well-recognized factors explain only some of the wide range of reproductive health problems that have been identified.

This review focuses on emerging evidence of the adverse reproductive effects of human-produced endocrine-disrupting chemicals (EDCs).

ENDOCRINE-DISRUPTING CHEMICALS ARE UBIQUITOUS

EDCs are exogenous substances that alter normal functioning of the endocrine system and consequently have the potential to cause adverse health effects in an intact organism and its progeny.^2,3,5 Substances classified as EDCs are divided into two groups:

• Human-produced compounds, such as pesticides, industrial solvents and lubricants, plastics (eg, BPA), and plasticizers (eg, phthalates)
• Phytoestrogens, which are naturally occurring plant compounds that bind to estrogen receptors. The major dietary source of phytoestrogens is soy.⁶

Many human-produced EDCs are lipid-soluble. Some bioconcentrate in fat. The dietary route of exposure is the most common, primarily from fat-containing foods such as meats and seafood. Common predatory fish such as tuna and swordfish, as well as other large fish, are more likely to have high levels of EDCs accumulated from contaminated water.⁷ Water, air, soil, and dust in communities, schools, and workplaces may also carry EDCs.⁸

Some human-produced EDCs have a very long half-life and can remain in the environment for years or even decades. Because of their long half-life, some EDCs can “travel” and become part of the food chain, even in “pristine” areas where manufactured substances are not normally found.²

Nearly all pregnant women in the United States have detectable seum levels of EDCs, including BPA

Women of childbearing age and pregnant women come in contact with EDCs on a daily basis. According to the Fourth National Report on Human Exposure to Environmental Chemicals, conducted by the US Centers for Disease Control and Prevention (CDC) in 2009, nearly all pregnant women in the United States have detectable serum levels of EDCs, including BPA, perchlorates, phthalates, polybrominated diphenyl ethers, and pesticides.⁹ Furthermore, BPA was found in more than 90% of the urine samples tested from random subsamples of 2,500 participants in the 2009 National Health and Nutrition Examination Survey report.¹⁰

EDCs AFFECT MULTIPLE PATHWAYS

Laboratory and animal research suggests that both manufactured and naturally occurring EDCs primarily affect sex steroid hormone pathways. However, some EDCs can affect adrenal, thyroid, and other endocrine pathways.

EDCs can potentially alter normal endocrine functioning by several mechanisms. They can bind to nuclear hormone receptors such as estrogen receptors, androgen receptors, progesterone receptors, thyroid hormone receptors, and peroxisome proliferator-activated receptor gamma. They can also bind to nonnuclear steroid hormone receptors (membrane estrogen receptors), nonsteroid receptors such as those for dopamine, serotonin, and norepinephrine, and orphan receptors such as aryl hydrocarbon receptor. In addition, some affect enzymatic pathways involved in steroid biosynthesis and metabolism and disrupt centralized endocrine pathways through positive and negative feedback.² There is also evidence that EDCs have epigenetic and transgenerational effects, as they seem to be able to influence DNA methylation and histone acetylation.¹¹

EVIDENCE OF HARM, LIMITATIONS OF EVIDENCE

In 2009, the Endocrine Society found convincing evidence that EDCs affect the male and female reproductive systems, breast health, oncogenesis, neuroendocrine function, thyroid function, metabolism, and cardiovascular health.² The first evidence of harm from EDCs came from long-term follow-up of randomized trials of diethylstilbestrol, a synthetic estrogen used as a pharmacologic agent between 1938 and 1971 to prevent miscarriage.¹²

In 2013, the American College of Obstetricians and Gynecologists and the American Society for Reproductive Medicine issued a joint opinion report on exposure to toxic environmental agents. The report acknowledged that environmental toxins are ubiquitous and that they can negatively affect health from preconception through adult life. “[We] join leading scientists and other clinical practitioners in calling for timely action to identify and reduce exposure to toxic environmental agents while addressing the consequences of such exposure.”¹³

Because the effects of endocrine disruption can occur during fetal and embryonic development, EDC exposure before and during pregnancy is of special concern.¹⁴ Long-term effects of exposure to hazardous chemicals such as cadmium, lead, and mercury in breast milk can lead to negative consequences in adulthood.^15,16

Animal studies have pointed to the negative health effects of human-produced EDCs. For example, animals exposed to BPA have aberrant breast and genitourinary development and exhibit abnormal endometrial stimulation and early puberty.²

Despite these suggestive findings, the health effects of environmental toxins on humans are difficult to determine definitively. Evidence comes mainly from epidemiologic studies showing distribution of particular medical conditions based on different levels of exposure. More research is needed on human exposure to environmental toxicants and their effects.^2,17,18 The National Children’s Study, a cohort collaborative study being conducted by the Environmental Protection Agency, CDC, and National Institutes of Health, is examining the effects of environmental exposures on more than 100,000 children across the United States, following them from before birth until age 21.¹⁹

THE PRECAUTIONARY PRINCIPLE

The precautionary principle is a fairly new concept in environmental science, although clinicians have been adhering to it for a long time. As proposed in a meeting of scientists, lawyers, policy makers, and environmentalists in 1998, “When an activity raises threats of harm to human health or the environment, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically.”²⁰

For example, studies show that breast milk in the United States commonly contains many chemicals that do not meet US Food and Drug Administration standards for baby food.^15,16 Although medical ethics prohibits randomized clinical trials in humans to examine the potential harms of these chemicals, proponents of the precautionary principle advocate that it is reasonable to be concerned that these chemicals may be harmful.^21–23

Applying the precautionary principle: Recommendations for patients

Without categorizing all chemicals as toxicants, nor all toxicants as worrisome for reproductive health, physicians can provide access to a growing body of information about reasonable strategies to reduce risk.

Educating patients about potential risks of exposure needs to be tempered with an appreciation of the economic and social barriers to avoiding or reducing risks. Often, people do not make conscious decisions about exposure. Rather, they are unknowingly exposed or have no access to safe, equivalent alternatives. Substituting suspected or known harmful products with potentially safer ones can be financially prohibitive. Where a patient lives and works may result in unavoidable exposure to environmental toxins.

Nevertheless, there are simple, economical alternatives that reduce or eliminate exposure to EDCs (Table 1).^24–26 These include avoiding plastic and metal containers that are not marked “BPA-free,” eating locally grown fruits and vegetables, controlling pests through frequent cleaning and trapping, and avoiding pesticides.

Additionally, the flame retardant tris (1,3-dichloro-2-propyl) phosphate may be an endocrine disrupter. It is found in numerous products for infants and young children, as well as in dust, automobiles, and furniture. Flame retardants are no longer used in infant clothing but may still be found in foam products such as upholstered furniture, automobiles, and children’s nap pads.^27,28

National organizations that provide resources for counseling women about human-produced EDCs as well as patient education materials are listed in Table 2.

CLINICAL CASE RESOLUTION

You counsel the patient that, although more research in humans is needed, because of safety concerns, avoiding exposing herself and her baby to BPA when possible is a reasonable approach. You review recommendations including avoiding BPA-containing plastics for food preparation (recycle code #7) and BPA-containing toys that might be mouthed by the baby, as well as minimizing the consumption of canned foods when possible. You further advise her to avoid products with flame retardants.

TAKE THE POST-TEST AND COMPLETE THE CME PROCESS

A 28-year-old woman presents for routine follow-up of asthma. She has a 1-year-old son and is considering a second pregnancy. She says she read on the Internet that the US Food and Drug Administration recently banned baby bottles and “sippy” cups that contain bisphenol A (BPA), as recommended by the American Medical Association. She wonders if there are other sources of BPA and if they pose a health risk to her, her son, and possible future children.

Environmental toxins have been linked to pregnancy complications and poor birth outcomes. The past several decades have seen a significant rise in reproductive disorders such as early onset of puberty, low sperm count, and birth defects such as cryptorchidism and hypospadias.^1–4 These changes may be partly explained by other trends over time, such as older maternal age, rising incidence of obesity and maternal diabetes, and demographic changes that exacerbate health disparities. However, these well-recognized factors explain only some of the wide range of reproductive health problems that have been identified.

This review focuses on emerging evidence of the adverse reproductive effects of human-produced endocrine-disrupting chemicals (EDCs).

ENDOCRINE-DISRUPTING CHEMICALS ARE UBIQUITOUS

EDCs are exogenous substances that alter normal functioning of the endocrine system and consequently have the potential to cause adverse health effects in an intact organism and its progeny.^2,3,5 Substances classified as EDCs are divided into two groups:

• Human-produced compounds, such as pesticides, industrial solvents and lubricants, plastics (eg, BPA), and plasticizers (eg, phthalates)
• Phytoestrogens, which are naturally occurring plant compounds that bind to estrogen receptors. The major dietary source of phytoestrogens is soy.⁶

Many human-produced EDCs are lipid-soluble. Some bioconcentrate in fat. The dietary route of exposure is the most common, primarily from fat-containing foods such as meats and seafood. Common predatory fish such as tuna and swordfish, as well as other large fish, are more likely to have high levels of EDCs accumulated from contaminated water.⁷ Water, air, soil, and dust in communities, schools, and workplaces may also carry EDCs.⁸

Some human-produced EDCs have a very long half-life and can remain in the environment for years or even decades. Because of their long half-life, some EDCs can “travel” and become part of the food chain, even in “pristine” areas where manufactured substances are not normally found.²

Nearly all pregnant women in the United States have detectable seum levels of EDCs, including BPA

Women of childbearing age and pregnant women come in contact with EDCs on a daily basis. According to the Fourth National Report on Human Exposure to Environmental Chemicals, conducted by the US Centers for Disease Control and Prevention (CDC) in 2009, nearly all pregnant women in the United States have detectable serum levels of EDCs, including BPA, perchlorates, phthalates, polybrominated diphenyl ethers, and pesticides.⁹ Furthermore, BPA was found in more than 90% of the urine samples tested from random subsamples of 2,500 participants in the 2009 National Health and Nutrition Examination Survey report.¹⁰

EDCs AFFECT MULTIPLE PATHWAYS

Laboratory and animal research suggests that both manufactured and naturally occurring EDCs primarily affect sex steroid hormone pathways. However, some EDCs can affect adrenal, thyroid, and other endocrine pathways.

EDCs can potentially alter normal endocrine functioning by several mechanisms. They can bind to nuclear hormone receptors such as estrogen receptors, androgen receptors, progesterone receptors, thyroid hormone receptors, and peroxisome proliferator-activated receptor gamma. They can also bind to nonnuclear steroid hormone receptors (membrane estrogen receptors), nonsteroid receptors such as those for dopamine, serotonin, and norepinephrine, and orphan receptors such as aryl hydrocarbon receptor. In addition, some affect enzymatic pathways involved in steroid biosynthesis and metabolism and disrupt centralized endocrine pathways through positive and negative feedback.² There is also evidence that EDCs have epigenetic and transgenerational effects, as they seem to be able to influence DNA methylation and histone acetylation.¹¹

EVIDENCE OF HARM, LIMITATIONS OF EVIDENCE

In 2009, the Endocrine Society found convincing evidence that EDCs affect the male and female reproductive systems, breast health, oncogenesis, neuroendocrine function, thyroid function, metabolism, and cardiovascular health.² The first evidence of harm from EDCs came from long-term follow-up of randomized trials of diethylstilbestrol, a synthetic estrogen used as a pharmacologic agent between 1938 and 1971 to prevent miscarriage.¹²

In 2013, the American College of Obstetricians and Gynecologists and the American Society for Reproductive Medicine issued a joint opinion report on exposure to toxic environmental agents. The report acknowledged that environmental toxins are ubiquitous and that they can negatively affect health from preconception through adult life. “[We] join leading scientists and other clinical practitioners in calling for timely action to identify and reduce exposure to toxic environmental agents while addressing the consequences of such exposure.”¹³

Because the effects of endocrine disruption can occur during fetal and embryonic development, EDC exposure before and during pregnancy is of special concern.¹⁴ Long-term effects of exposure to hazardous chemicals such as cadmium, lead, and mercury in breast milk can lead to negative consequences in adulthood.^15,16

Animal studies have pointed to the negative health effects of human-produced EDCs. For example, animals exposed to BPA have aberrant breast and genitourinary development and exhibit abnormal endometrial stimulation and early puberty.²

Despite these suggestive findings, the health effects of environmental toxins on humans are difficult to determine definitively. Evidence comes mainly from epidemiologic studies showing distribution of particular medical conditions based on different levels of exposure. More research is needed on human exposure to environmental toxicants and their effects.^2,17,18 The National Children’s Study, a cohort collaborative study being conducted by the Environmental Protection Agency, CDC, and National Institutes of Health, is examining the effects of environmental exposures on more than 100,000 children across the United States, following them from before birth until age 21.¹⁹

THE PRECAUTIONARY PRINCIPLE

The precautionary principle is a fairly new concept in environmental science, although clinicians have been adhering to it for a long time. As proposed in a meeting of scientists, lawyers, policy makers, and environmentalists in 1998, “When an activity raises threats of harm to human health or the environment, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically.”²⁰

For example, studies show that breast milk in the United States commonly contains many chemicals that do not meet US Food and Drug Administration standards for baby food.^15,16 Although medical ethics prohibits randomized clinical trials in humans to examine the potential harms of these chemicals, proponents of the precautionary principle advocate that it is reasonable to be concerned that these chemicals may be harmful.^21–23

Applying the precautionary principle: Recommendations for patients

Without categorizing all chemicals as toxicants, nor all toxicants as worrisome for reproductive health, physicians can provide access to a growing body of information about reasonable strategies to reduce risk.

Educating patients about potential risks of exposure needs to be tempered with an appreciation of the economic and social barriers to avoiding or reducing risks. Often, people do not make conscious decisions about exposure. Rather, they are unknowingly exposed or have no access to safe, equivalent alternatives. Substituting suspected or known harmful products with potentially safer ones can be financially prohibitive. Where a patient lives and works may result in unavoidable exposure to environmental toxins.

Nevertheless, there are simple, economical alternatives that reduce or eliminate exposure to EDCs (Table 1).^24–26 These include avoiding plastic and metal containers that are not marked “BPA-free,” eating locally grown fruits and vegetables, controlling pests through frequent cleaning and trapping, and avoiding pesticides.

Additionally, the flame retardant tris (1,3-dichloro-2-propyl) phosphate may be an endocrine disrupter. It is found in numerous products for infants and young children, as well as in dust, automobiles, and furniture. Flame retardants are no longer used in infant clothing but may still be found in foam products such as upholstered furniture, automobiles, and children’s nap pads.^27,28

National organizations that provide resources for counseling women about human-produced EDCs as well as patient education materials are listed in Table 2.

CLINICAL CASE RESOLUTION

You counsel the patient that, although more research in humans is needed, because of safety concerns, avoiding exposing herself and her baby to BPA when possible is a reasonable approach. You review recommendations including avoiding BPA-containing plastics for food preparation (recycle code #7) and BPA-containing toys that might be mouthed by the baby, as well as minimizing the consumption of canned foods when possible. You further advise her to avoid products with flame retardants.

TAKE THE POST-TEST AND COMPLETE THE CME PROCESS

A 28-year-old woman presents for routine follow-up of asthma. She has a 1-year-old son and is considering a second pregnancy. She says she read on the Internet that the US Food and Drug Administration recently banned baby bottles and “sippy” cups that contain bisphenol A (BPA), as recommended by the American Medical Association. She wonders if there are other sources of BPA and if they pose a health risk to her, her son, and possible future children.

Environmental toxins have been linked to pregnancy complications and poor birth outcomes. The past several decades have seen a significant rise in reproductive disorders such as early onset of puberty, low sperm count, and birth defects such as cryptorchidism and hypospadias.^1–4 These changes may be partly explained by other trends over time, such as older maternal age, rising incidence of obesity and maternal diabetes, and demographic changes that exacerbate health disparities. However, these well-recognized factors explain only some of the wide range of reproductive health problems that have been identified.

This review focuses on emerging evidence of the adverse reproductive effects of human-produced endocrine-disrupting chemicals (EDCs).

ENDOCRINE-DISRUPTING CHEMICALS ARE UBIQUITOUS

EDCs are exogenous substances that alter normal functioning of the endocrine system and consequently have the potential to cause adverse health effects in an intact organism and its progeny.^2,3,5 Substances classified as EDCs are divided into two groups:

• Human-produced compounds, such as pesticides, industrial solvents and lubricants, plastics (eg, BPA), and plasticizers (eg, phthalates)
• Phytoestrogens, which are naturally occurring plant compounds that bind to estrogen receptors. The major dietary source of phytoestrogens is soy.⁶

Many human-produced EDCs are lipid-soluble. Some bioconcentrate in fat. The dietary route of exposure is the most common, primarily from fat-containing foods such as meats and seafood. Common predatory fish such as tuna and swordfish, as well as other large fish, are more likely to have high levels of EDCs accumulated from contaminated water.⁷ Water, air, soil, and dust in communities, schools, and workplaces may also carry EDCs.⁸

Some human-produced EDCs have a very long half-life and can remain in the environment for years or even decades. Because of their long half-life, some EDCs can “travel” and become part of the food chain, even in “pristine” areas where manufactured substances are not normally found.²

Nearly all pregnant women in the United States have detectable seum levels of EDCs, including BPA

Women of childbearing age and pregnant women come in contact with EDCs on a daily basis. According to the Fourth National Report on Human Exposure to Environmental Chemicals, conducted by the US Centers for Disease Control and Prevention (CDC) in 2009, nearly all pregnant women in the United States have detectable serum levels of EDCs, including BPA, perchlorates, phthalates, polybrominated diphenyl ethers, and pesticides.⁹ Furthermore, BPA was found in more than 90% of the urine samples tested from random subsamples of 2,500 participants in the 2009 National Health and Nutrition Examination Survey report.¹⁰

EDCs AFFECT MULTIPLE PATHWAYS

Laboratory and animal research suggests that both manufactured and naturally occurring EDCs primarily affect sex steroid hormone pathways. However, some EDCs can affect adrenal, thyroid, and other endocrine pathways.

EDCs can potentially alter normal endocrine functioning by several mechanisms. They can bind to nuclear hormone receptors such as estrogen receptors, androgen receptors, progesterone receptors, thyroid hormone receptors, and peroxisome proliferator-activated receptor gamma. They can also bind to nonnuclear steroid hormone receptors (membrane estrogen receptors), nonsteroid receptors such as those for dopamine, serotonin, and norepinephrine, and orphan receptors such as aryl hydrocarbon receptor. In addition, some affect enzymatic pathways involved in steroid biosynthesis and metabolism and disrupt centralized endocrine pathways through positive and negative feedback.² There is also evidence that EDCs have epigenetic and transgenerational effects, as they seem to be able to influence DNA methylation and histone acetylation.¹¹

EVIDENCE OF HARM, LIMITATIONS OF EVIDENCE

In 2009, the Endocrine Society found convincing evidence that EDCs affect the male and female reproductive systems, breast health, oncogenesis, neuroendocrine function, thyroid function, metabolism, and cardiovascular health.² The first evidence of harm from EDCs came from long-term follow-up of randomized trials of diethylstilbestrol, a synthetic estrogen used as a pharmacologic agent between 1938 and 1971 to prevent miscarriage.¹²

In 2013, the American College of Obstetricians and Gynecologists and the American Society for Reproductive Medicine issued a joint opinion report on exposure to toxic environmental agents. The report acknowledged that environmental toxins are ubiquitous and that they can negatively affect health from preconception through adult life. “[We] join leading scientists and other clinical practitioners in calling for timely action to identify and reduce exposure to toxic environmental agents while addressing the consequences of such exposure.”¹³

Because the effects of endocrine disruption can occur during fetal and embryonic development, EDC exposure before and during pregnancy is of special concern.¹⁴ Long-term effects of exposure to hazardous chemicals such as cadmium, lead, and mercury in breast milk can lead to negative consequences in adulthood.^15,16

Animal studies have pointed to the negative health effects of human-produced EDCs. For example, animals exposed to BPA have aberrant breast and genitourinary development and exhibit abnormal endometrial stimulation and early puberty.²

Despite these suggestive findings, the health effects of environmental toxins on humans are difficult to determine definitively. Evidence comes mainly from epidemiologic studies showing distribution of particular medical conditions based on different levels of exposure. More research is needed on human exposure to environmental toxicants and their effects.^2,17,18 The National Children’s Study, a cohort collaborative study being conducted by the Environmental Protection Agency, CDC, and National Institutes of Health, is examining the effects of environmental exposures on more than 100,000 children across the United States, following them from before birth until age 21.¹⁹

THE PRECAUTIONARY PRINCIPLE

The precautionary principle is a fairly new concept in environmental science, although clinicians have been adhering to it for a long time. As proposed in a meeting of scientists, lawyers, policy makers, and environmentalists in 1998, “When an activity raises threats of harm to human health or the environment, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically.”²⁰

For example, studies show that breast milk in the United States commonly contains many chemicals that do not meet US Food and Drug Administration standards for baby food.^15,16 Although medical ethics prohibits randomized clinical trials in humans to examine the potential harms of these chemicals, proponents of the precautionary principle advocate that it is reasonable to be concerned that these chemicals may be harmful.^21–23

Applying the precautionary principle: Recommendations for patients

Without categorizing all chemicals as toxicants, nor all toxicants as worrisome for reproductive health, physicians can provide access to a growing body of information about reasonable strategies to reduce risk.

Educating patients about potential risks of exposure needs to be tempered with an appreciation of the economic and social barriers to avoiding or reducing risks. Often, people do not make conscious decisions about exposure. Rather, they are unknowingly exposed or have no access to safe, equivalent alternatives. Substituting suspected or known harmful products with potentially safer ones can be financially prohibitive. Where a patient lives and works may result in unavoidable exposure to environmental toxins.

Nevertheless, there are simple, economical alternatives that reduce or eliminate exposure to EDCs (Table 1).^24–26 These include avoiding plastic and metal containers that are not marked “BPA-free,” eating locally grown fruits and vegetables, controlling pests through frequent cleaning and trapping, and avoiding pesticides.

Additionally, the flame retardant tris (1,3-dichloro-2-propyl) phosphate may be an endocrine disrupter. It is found in numerous products for infants and young children, as well as in dust, automobiles, and furniture. Flame retardants are no longer used in infant clothing but may still be found in foam products such as upholstered furniture, automobiles, and children’s nap pads.^27,28

National organizations that provide resources for counseling women about human-produced EDCs as well as patient education materials are listed in Table 2.

CLINICAL CASE RESOLUTION

You counsel the patient that, although more research in humans is needed, because of safety concerns, avoiding exposing herself and her baby to BPA when possible is a reasonable approach. You review recommendations including avoiding BPA-containing plastics for food preparation (recycle code #7) and BPA-containing toys that might be mouthed by the baby, as well as minimizing the consumption of canned foods when possible. You further advise her to avoid products with flame retardants.

TAKE THE POST-TEST AND COMPLETE THE CME PROCESS

A 50-year-old man presented to the otolaryngology clinic, complaining of throat discomfort for the past 3 months that worsened in the supine position. The clinical examination revealed one cystic lesion on either side of the epiglottis (Figure 1). The larger cyst measured 23 × 13 × 12 mm. The man’s symptoms resolved completely after endoscopic removal of the cysts under general anesthesia (Figure 2).

CLINICAL IMPLICATIONS

Epiglottic cysts are benign lesions on the lingual or laryngeal aspect of the epiglottis and are often a result of mucus retention. Otolaryngologists, anesthesiologists, and endoscopists are usually the first to discover them. Because the cysts are usually asymptomatic, it is difficult to calculate their prevalence in the general population.

If the cysts are symptomatic, patients usually describe nonspecific complaints such as pharyngeal discomfort and foreign-body sensation. Voluminous cysts discovered incidentally in patients requiring general anesthesia pose a challenge to intubation and increase the risk of airway obstruction.

In addition, epiglottic cysts have been reported in a proportion of adult patients with acute epiglottitis (25% in one series)¹; They may be discovered either during the acute infection or when inflammation subsides.² These patients have been reported to have a higher  risk of acute airway obstruction, with a sixfold increase in the need for acute airway management in one series.¹ Moreover, epiglottitis was reported to recur in 12.5% and 17% of patients with cysts in two different series,^1,2 corresponding to a recurrence rate 15 times higher than in patients with no cysts.¹ Hence, it would be useful to rule out their presence after any episode of acute epiglottitis.

Once an epiglottic cyst is discovered in a symptomatic patient, it can be managed either by elective endoscopic resection or by marsupialization. Either procedure is safe, with good long-term results. Simple evacuation of the cyst should be considered in cases of acute airway management.

A 50-year-old man presented to the otolaryngology clinic, complaining of throat discomfort for the past 3 months that worsened in the supine position. The clinical examination revealed one cystic lesion on either side of the epiglottis (Figure 1). The larger cyst measured 23 × 13 × 12 mm. The man’s symptoms resolved completely after endoscopic removal of the cysts under general anesthesia (Figure 2).

CLINICAL IMPLICATIONS

Epiglottic cysts are benign lesions on the lingual or laryngeal aspect of the epiglottis and are often a result of mucus retention. Otolaryngologists, anesthesiologists, and endoscopists are usually the first to discover them. Because the cysts are usually asymptomatic, it is difficult to calculate their prevalence in the general population.

If the cysts are symptomatic, patients usually describe nonspecific complaints such as pharyngeal discomfort and foreign-body sensation. Voluminous cysts discovered incidentally in patients requiring general anesthesia pose a challenge to intubation and increase the risk of airway obstruction.

In addition, epiglottic cysts have been reported in a proportion of adult patients with acute epiglottitis (25% in one series)¹; They may be discovered either during the acute infection or when inflammation subsides.² These patients have been reported to have a higher  risk of acute airway obstruction, with a sixfold increase in the need for acute airway management in one series.¹ Moreover, epiglottitis was reported to recur in 12.5% and 17% of patients with cysts in two different series,^1,2 corresponding to a recurrence rate 15 times higher than in patients with no cysts.¹ Hence, it would be useful to rule out their presence after any episode of acute epiglottitis.

Once an epiglottic cyst is discovered in a symptomatic patient, it can be managed either by elective endoscopic resection or by marsupialization. Either procedure is safe, with good long-term results. Simple evacuation of the cyst should be considered in cases of acute airway management.

A 50-year-old man presented to the otolaryngology clinic, complaining of throat discomfort for the past 3 months that worsened in the supine position. The clinical examination revealed one cystic lesion on either side of the epiglottis (Figure 1). The larger cyst measured 23 × 13 × 12 mm. The man’s symptoms resolved completely after endoscopic removal of the cysts under general anesthesia (Figure 2).

CLINICAL IMPLICATIONS

Epiglottic cysts are benign lesions on the lingual or laryngeal aspect of the epiglottis and are often a result of mucus retention. Otolaryngologists, anesthesiologists, and endoscopists are usually the first to discover them. Because the cysts are usually asymptomatic, it is difficult to calculate their prevalence in the general population.

If the cysts are symptomatic, patients usually describe nonspecific complaints such as pharyngeal discomfort and foreign-body sensation. Voluminous cysts discovered incidentally in patients requiring general anesthesia pose a challenge to intubation and increase the risk of airway obstruction.

In addition, epiglottic cysts have been reported in a proportion of adult patients with acute epiglottitis (25% in one series)¹; They may be discovered either during the acute infection or when inflammation subsides.² These patients have been reported to have a higher  risk of acute airway obstruction, with a sixfold increase in the need for acute airway management in one series.¹ Moreover, epiglottitis was reported to recur in 12.5% and 17% of patients with cysts in two different series,^1,2 corresponding to a recurrence rate 15 times higher than in patients with no cysts.¹ Hence, it would be useful to rule out their presence after any episode of acute epiglottitis.

Once an epiglottic cyst is discovered in a symptomatic patient, it can be managed either by elective endoscopic resection or by marsupialization. Either procedure is safe, with good long-term results. Simple evacuation of the cyst should be considered in cases of acute airway management.

Measles continues to rear its head in the United States. Because it is so contagious, even the few cases introduced by travelers quickly spread to susceptible contacts. Life-threatening and severely disabling complications can occur, although this is rare. Widespread immunization and prompt recognition and isolation of contacts are key to controlling outbreaks.

This article reviews the epidemiology of measles, describes its distinctive clinical picture, and provides recommendations for infection control and prevention, including in immunosuppressed populations.

MEASLES IS SERIOUS AND HIGHLY CONTAGIOUS

Up to 90% of susceptible people develop measles after exposure, making it one of the most contagious of infections. The virus is transmitted by airborne spread when an infected person coughs or sneezes, or by direct contact with infectious droplets. The virus can remain infectious in the air or on a surface for up to 2 hours.¹

Worldwide, an estimated 20 million people are infected with measles each year, and 146,000 die of complications. In 1980, before widespread vaccination, 2.6 million deaths were attributable to measles annually. In the United States before the introduction of measles vaccine in 1963, measles was a significant cause of disease and death: an estimated 3 to 4 million people were infected annually, although only about 549,000 were reported. There were 48,000 hospitalizations, 1,000 cases of permanent brain damage from measles encephalitis, and 495 deaths annually.²

Outbreaks still occur regularly

In 2000, measles was declared eliminated from the United States,³ but annual outbreaks have occurred since then as a result of cases imported from other countries and their subsequent transmission to unvaccinated people. From 2001 to 2012, a median of four outbreaks and 60 cases were reported annually to the US Centers for Disease Control and Prevention.⁴

In January 2015, a multistate measles outbreak originating in Disneyland in California was recognized. As of April 17, when the outbreak was declared over, 111 measles cases from seven states had been linked to this outbreak.⁵ Of the evaluable cases, 44% were in unvaccinated people and 38% were in those whose vaccination status was unknown or undocumented. The median age of patients was 21, and 20% required hospitalization.

This outbreak, as well as four other smaller US outbreaks the same year, underscores the transmissibility of the virus in populations containing only a small percentage of unvaccinated people.⁶

DISTINCTIVE CLINICAL PICTURE

The incubation period for measles infection is 7 to 21 days, with most cases becoming apparent 10 to 12 days after exposure. Measles should be suspected in a patient with the following clinical features whose history indicates susceptibility and exposure (ie, an unimmunized person with a history of exposure or travel):

Severe acute respiratory illness. Measles usually presents as an acute respiratory viral illness, which typically lasts 2 to 4 days. The illness involves high fevers, malaise, anorexia, and the “three Cs”: cough, coryza (rhinitis), and conjunctivitis. Patients usually appear sicker than those with more common viral illnesses.

From the US Centers for Disease Control and Prevention.

Koplik spots, which are pathognomonic for measles, are seen in the first few days of illness. They are bluish-white, slightly raised lesions on an erythematous base on the buccal mucosa, usually opposite the first molar (Figure 1). Spots can also be seen on the soft palate, conjunctiva, and vaginal mucosa. Koplik spots usually disappear after a few days and often are not appreciable at the time of evaluation.

From the US Centers for Disease Control and Prevention.

Discrete erythematous patches develop on the face and neck a day after the appearance of Koplik spots. This rash becomes more confluent as it spreads to involve the entire body (Figure 2). It typically lasts for 3 to 7 days, then fades in a similar pattern. The confluent nature of this rash and its spread from the face and neck to the entire body are characteristic of measles. Patients are highly contagious from 4 days before the onset of the rash to 4 days after.

COMPLICATIONS CAN BE SEVERE

Those at highest risk for measles complications are infants, children under age 5, adults over age 20, pregnant women, and immunosuppressed individuals.⁷

Pneumonia—either a primary measles pneumonia or a secondary viral or bacterial pneumonia—is the most common cause of death.^8,9 Viruses complicating measles are typically adenovirus and herpes simplex virus. Bacteria causing secondary infection are usually Staphylococcus aureus and Streptococcus pneumoniae and, less commonly, gram-negative bacteria.

Laryngotracheobronchitis (croup) is the second most common cause of death, with bacteria and viruses similar to those causing measles-related pneumonia.

Otitis media is the most common complication of measles. Other respiratory complications include mastoiditis, pneumothorax, and mediastinal emphysema.

Acute measles encephalitis occurs in 1 measles case per 1,000 and often results in permanent brain damage. During the convalescent phase of the illness, fever again emerges, with the development of headaches, seizures, and altered consciousness.¹⁰

Subacute sclerosing panencephalitis is a rare fatal degenerative disease of the central nervous system caused by a persistent infection with a defective measles virus. The precise pathophysiology is unclear, but it is thought that mutations of the viral genome lead to altered cellular immunity.¹¹ The condition typically occurs 7 to 10 years after the initial measles infection, particularly in those who developed measles before age 2. Clinical manifestations include behavioral disturbances, intellectual deterioration, and myoclonic seizures, slowly progressing to a vegetative state and death.¹²

Other complications of measles include diarrhea and stomatitis, which are associated with malnutrition in developing countries, and subclinical hepatitis, thrombocytopenia, appendicitis, ileocolitis, hypokalemia, and myocarditis.

During pregnancy, measles infection can be complicated by primary measles pneumonia and is associated with an increased risk of miscarriage and premature birth.¹³

Patients with a cell-mediated immunodeficiency who develop measles are particularly susceptible to fatal measles pneumonia and acute progressive encephalitis.¹⁴

ATYPICAL MEASLES IN THOSE WHO RECEIVED KILLED VACCINE

From 1963 to 1967, a killed measles vaccine was available in the United States. Those who received this vaccine are susceptible to an atypical form of measles when exposed to the virus,¹⁵ characterized by a 1- to 2-day prodrome, followed by the appearance of a maculopapular or petechial rash on the distal extremities that spreads centripetally. Patients develop high fever and edema of the hands and feet, and have a more prolonged course than with classic measles. It is believed not to be contagious.¹⁶

LABORATORY CONFIRMATION

Laboratory confirmation of measles is recommended for suspected cases. Because viral isolation is technically difficult and is not readily available in most laboratories, measles-specific immunoglobulin M antibody serologic testing is most commonly used. It is almost 100% sensitive when done 2 to 3 days after the onset of the rash.¹⁷

Measles RNA testing by real-time polymerase chain reaction to detect measles virus in the blood, throat, or urine is more specific and if available may be preferred over serologic testing.¹⁸

SUPPORTIVE MANAGEMENT AND VITAMIN A SUPPLEMENTATION

No specific antiviral therapy for measles is available. Management involves supportive measures and monitoring for secondary bacterial complications.

The World Health Organization and the American Academy of Pediatrics recommend vitamin A supplementation for all children with acute measles.¹⁹ In developing countries, it has been shown to reduce rates of morbidity and death in measles-infected children.²⁰ In the United States, children with measles have been found to have low serum levels of vitamin A, with lower levels associated with more severe disease.

VACCINATION RECOMMENDATIONS

The only measles vaccine available in the United States is a live further-attenuated strain prepared in chick embryo cell culture and combined with mumps and rubella vaccine (MMR) or with measles, mumps, rubella, and varicella vaccine (MMRV).

Healthy children. Two doses of measles vaccine are recommended, as a single dose is associated with a 5% failure rate. The recommended schedule is:

• First dose at age 12 to 15 months
• Second dose at the time of school entry (ages 4 to 6), or at any time at least 28 days after the first dose.¹⁹

More than 99% of children who receive two doses of vaccine according to this schedule develop serologic evidence of measles immunity. Vaccination provides long-term immunity, and many epidemiologic studies have documented that waning immunity after vaccination occurs only very rarely.²¹

All school-age children, including elementary, middle, and high school students, who received only one dose of measles vaccine should receive the second dose.

Adults born in 1957 or later should receive at least one dose of measles vaccine unless they have other acceptable evidence of immunity, such as:⁴

• Documentation of age-appropriate live measles vaccine, ie, one dose of vaccine for adults not at high risk, or two doses for those at high risk (see below)
• Laboratory evidence of immunity (ie, measles immunoglobulin G in serum)
• Laboratory confirmation of disease.

Adults born before 1957 can be considered to be immune to measles, although MMR vaccine can be administered in those without contraindications.

Vitamin A supplementation is recommended for acute measles

Adults at increased risk of exposure or transmission of measles and who do not have evidence of immunity should receive two doses of MMR vaccine, given at least 28 days apart. This high-risk group includes:

• Students attending college or other post-high school educational institution
• Healthcare personnel
• International travelers.

During measles outbreaks, every effort should be made to ensure that those at high risk are vaccinated with two doses of MMR or have other acceptable evidence of immunity.

LIVE VACCINE IS SAFE FOR MOST PEOPLE

Mild side effects. A transient fever, which may be accompanied by a discrete or confluent rash, occurs in 5% to 15% of recipients 5 to 12 days after vaccination.

Transmission does not occur. People who have been newly vaccinated do not transmit the virus to susceptible contacts, even if they develop a vaccine-associated rash. The vaccine can safely be given to close contacts of immunocompromised and other susceptible people.

Egg allergy not a concern. Measles vaccine is produced in chick embryo cell culture but has been shown to be safe for people with egg allergy and is recommended without the need for egg allergy testing.¹⁹

Autism link debunked. No scientific evidence shows that the risk of autism is higher in children who receive MMR vaccine than in those who do not. In 2001, an Institute of Medicine report rejected a causal relationship between MMR vaccine and autism spectrum disorders.²²

CONTRAINDICATIONS

Measles vaccine is contraindicated for:

• Patients who have cell-mediated immune deficiencies, except human immunodeficiency virus (HIV) infection
• Pregnant women
• Those who have had a severe allergic reaction to a vaccine component in the past
• Those with moderate or severe acute illness
• Those who have recently received immunoglobulin products.

People with HIV infection who are severely immunosuppressed should not receive live measles vaccine. However, because of the risk of severe measles in HIV-infected patients and because the vaccine has been shown to be safe for patients with HIV without severe immunosuppression, the vaccine is recommended for those with asymptomatic or mildly symptomatic HIV infection who do not have evidence of severe immunosuppression (ie, CD4 lymphocytes < 15% or < 200 cells/µL).^3,4

INFECTION CONTROL AND PREVENTION

All school-age children who received only one dose of measles vaccine should receive the second dose

Healthcare workers should maintain a high index of suspicion for measles and implement isolation procedures promptly in patients with a febrile illness, rash, and a history of travel abroad or contact with travelers from abroad.²³ Suspected cases should be reported promptly to local health agencies to help limit spread.

Patients with measles should be placed in airborne isolation (eg, use of an N95 or higher level respirator and an airborne infection isolation room) for 4 days after the onset of the rash in a normal host and for the duration of the illness in an immunocompromised patient. Healthcare staff, regardless of their immunity status, should adhere to these precautions when entering the room of infected patients.

Immunization programs should be established to ensure that everyone who works or volunteers in healthcare facilities is protected against measles.⁴

Postexposure prophylaxis. Measles vaccination given to susceptible contacts within 72 hours of exposure may provide protection against infection and induces protection against subsequent measles exposures.^24,25

Vaccination is the best intervention for susceptible contacts older than 12 months who do not have a contraindication to measles vaccination, and for those who have received only one dose of measles vaccine.

Passive immunization. Active immunization is the best strategy for controlling measles outbreaks. Passive immunization with intramuscularly or intravenously administered immunoglobulin given within 6 days of exposure can be used to prevent transmission or modify the clinical course of infection for susceptible contacts at high risk of developing severe or fatal measles. This includes people who are being treated with immunosuppressive agents, HIV-infected, pregnant, or younger than 1 year of age.

Measles continues to rear its head in the United States. Because it is so contagious, even the few cases introduced by travelers quickly spread to susceptible contacts. Life-threatening and severely disabling complications can occur, although this is rare. Widespread immunization and prompt recognition and isolation of contacts are key to controlling outbreaks.

This article reviews the epidemiology of measles, describes its distinctive clinical picture, and provides recommendations for infection control and prevention, including in immunosuppressed populations.

MEASLES IS SERIOUS AND HIGHLY CONTAGIOUS

Up to 90% of susceptible people develop measles after exposure, making it one of the most contagious of infections. The virus is transmitted by airborne spread when an infected person coughs or sneezes, or by direct contact with infectious droplets. The virus can remain infectious in the air or on a surface for up to 2 hours.¹

Worldwide, an estimated 20 million people are infected with measles each year, and 146,000 die of complications. In 1980, before widespread vaccination, 2.6 million deaths were attributable to measles annually. In the United States before the introduction of measles vaccine in 1963, measles was a significant cause of disease and death: an estimated 3 to 4 million people were infected annually, although only about 549,000 were reported. There were 48,000 hospitalizations, 1,000 cases of permanent brain damage from measles encephalitis, and 495 deaths annually.²

Outbreaks still occur regularly

In 2000, measles was declared eliminated from the United States,³ but annual outbreaks have occurred since then as a result of cases imported from other countries and their subsequent transmission to unvaccinated people. From 2001 to 2012, a median of four outbreaks and 60 cases were reported annually to the US Centers for Disease Control and Prevention.⁴

In January 2015, a multistate measles outbreak originating in Disneyland in California was recognized. As of April 17, when the outbreak was declared over, 111 measles cases from seven states had been linked to this outbreak.⁵ Of the evaluable cases, 44% were in unvaccinated people and 38% were in those whose vaccination status was unknown or undocumented. The median age of patients was 21, and 20% required hospitalization.

This outbreak, as well as four other smaller US outbreaks the same year, underscores the transmissibility of the virus in populations containing only a small percentage of unvaccinated people.⁶

DISTINCTIVE CLINICAL PICTURE

The incubation period for measles infection is 7 to 21 days, with most cases becoming apparent 10 to 12 days after exposure. Measles should be suspected in a patient with the following clinical features whose history indicates susceptibility and exposure (ie, an unimmunized person with a history of exposure or travel):

Severe acute respiratory illness. Measles usually presents as an acute respiratory viral illness, which typically lasts 2 to 4 days. The illness involves high fevers, malaise, anorexia, and the “three Cs”: cough, coryza (rhinitis), and conjunctivitis. Patients usually appear sicker than those with more common viral illnesses.

From the US Centers for Disease Control and Prevention.

Koplik spots, which are pathognomonic for measles, are seen in the first few days of illness. They are bluish-white, slightly raised lesions on an erythematous base on the buccal mucosa, usually opposite the first molar (Figure 1). Spots can also be seen on the soft palate, conjunctiva, and vaginal mucosa. Koplik spots usually disappear after a few days and often are not appreciable at the time of evaluation.

From the US Centers for Disease Control and Prevention.

Discrete erythematous patches develop on the face and neck a day after the appearance of Koplik spots. This rash becomes more confluent as it spreads to involve the entire body (Figure 2). It typically lasts for 3 to 7 days, then fades in a similar pattern. The confluent nature of this rash and its spread from the face and neck to the entire body are characteristic of measles. Patients are highly contagious from 4 days before the onset of the rash to 4 days after.

COMPLICATIONS CAN BE SEVERE

Those at highest risk for measles complications are infants, children under age 5, adults over age 20, pregnant women, and immunosuppressed individuals.⁷

Pneumonia—either a primary measles pneumonia or a secondary viral or bacterial pneumonia—is the most common cause of death.^8,9 Viruses complicating measles are typically adenovirus and herpes simplex virus. Bacteria causing secondary infection are usually Staphylococcus aureus and Streptococcus pneumoniae and, less commonly, gram-negative bacteria.

Laryngotracheobronchitis (croup) is the second most common cause of death, with bacteria and viruses similar to those causing measles-related pneumonia.

Otitis media is the most common complication of measles. Other respiratory complications include mastoiditis, pneumothorax, and mediastinal emphysema.

Acute measles encephalitis occurs in 1 measles case per 1,000 and often results in permanent brain damage. During the convalescent phase of the illness, fever again emerges, with the development of headaches, seizures, and altered consciousness.¹⁰

Subacute sclerosing panencephalitis is a rare fatal degenerative disease of the central nervous system caused by a persistent infection with a defective measles virus. The precise pathophysiology is unclear, but it is thought that mutations of the viral genome lead to altered cellular immunity.¹¹ The condition typically occurs 7 to 10 years after the initial measles infection, particularly in those who developed measles before age 2. Clinical manifestations include behavioral disturbances, intellectual deterioration, and myoclonic seizures, slowly progressing to a vegetative state and death.¹²

Other complications of measles include diarrhea and stomatitis, which are associated with malnutrition in developing countries, and subclinical hepatitis, thrombocytopenia, appendicitis, ileocolitis, hypokalemia, and myocarditis.

During pregnancy, measles infection can be complicated by primary measles pneumonia and is associated with an increased risk of miscarriage and premature birth.¹³

Patients with a cell-mediated immunodeficiency who develop measles are particularly susceptible to fatal measles pneumonia and acute progressive encephalitis.¹⁴

ATYPICAL MEASLES IN THOSE WHO RECEIVED KILLED VACCINE

From 1963 to 1967, a killed measles vaccine was available in the United States. Those who received this vaccine are susceptible to an atypical form of measles when exposed to the virus,¹⁵ characterized by a 1- to 2-day prodrome, followed by the appearance of a maculopapular or petechial rash on the distal extremities that spreads centripetally. Patients develop high fever and edema of the hands and feet, and have a more prolonged course than with classic measles. It is believed not to be contagious.¹⁶

LABORATORY CONFIRMATION

Laboratory confirmation of measles is recommended for suspected cases. Because viral isolation is technically difficult and is not readily available in most laboratories, measles-specific immunoglobulin M antibody serologic testing is most commonly used. It is almost 100% sensitive when done 2 to 3 days after the onset of the rash.¹⁷

Measles RNA testing by real-time polymerase chain reaction to detect measles virus in the blood, throat, or urine is more specific and if available may be preferred over serologic testing.¹⁸

SUPPORTIVE MANAGEMENT AND VITAMIN A SUPPLEMENTATION

No specific antiviral therapy for measles is available. Management involves supportive measures and monitoring for secondary bacterial complications.

The World Health Organization and the American Academy of Pediatrics recommend vitamin A supplementation for all children with acute measles.¹⁹ In developing countries, it has been shown to reduce rates of morbidity and death in measles-infected children.²⁰ In the United States, children with measles have been found to have low serum levels of vitamin A, with lower levels associated with more severe disease.

VACCINATION RECOMMENDATIONS

The only measles vaccine available in the United States is a live further-attenuated strain prepared in chick embryo cell culture and combined with mumps and rubella vaccine (MMR) or with measles, mumps, rubella, and varicella vaccine (MMRV).

Healthy children. Two doses of measles vaccine are recommended, as a single dose is associated with a 5% failure rate. The recommended schedule is:

• First dose at age 12 to 15 months
• Second dose at the time of school entry (ages 4 to 6), or at any time at least 28 days after the first dose.¹⁹

More than 99% of children who receive two doses of vaccine according to this schedule develop serologic evidence of measles immunity. Vaccination provides long-term immunity, and many epidemiologic studies have documented that waning immunity after vaccination occurs only very rarely.²¹

All school-age children, including elementary, middle, and high school students, who received only one dose of measles vaccine should receive the second dose.

Adults born in 1957 or later should receive at least one dose of measles vaccine unless they have other acceptable evidence of immunity, such as:⁴

• Documentation of age-appropriate live measles vaccine, ie, one dose of vaccine for adults not at high risk, or two doses for those at high risk (see below)
• Laboratory evidence of immunity (ie, measles immunoglobulin G in serum)
• Laboratory confirmation of disease.

Adults born before 1957 can be considered to be immune to measles, although MMR vaccine can be administered in those without contraindications.

Vitamin A supplementation is recommended for acute measles

Adults at increased risk of exposure or transmission of measles and who do not have evidence of immunity should receive two doses of MMR vaccine, given at least 28 days apart. This high-risk group includes:

• Students attending college or other post-high school educational institution
• Healthcare personnel
• International travelers.

During measles outbreaks, every effort should be made to ensure that those at high risk are vaccinated with two doses of MMR or have other acceptable evidence of immunity.

LIVE VACCINE IS SAFE FOR MOST PEOPLE

Mild side effects. A transient fever, which may be accompanied by a discrete or confluent rash, occurs in 5% to 15% of recipients 5 to 12 days after vaccination.

Transmission does not occur. People who have been newly vaccinated do not transmit the virus to susceptible contacts, even if they develop a vaccine-associated rash. The vaccine can safely be given to close contacts of immunocompromised and other susceptible people.

Egg allergy not a concern. Measles vaccine is produced in chick embryo cell culture but has been shown to be safe for people with egg allergy and is recommended without the need for egg allergy testing.¹⁹

Autism link debunked. No scientific evidence shows that the risk of autism is higher in children who receive MMR vaccine than in those who do not. In 2001, an Institute of Medicine report rejected a causal relationship between MMR vaccine and autism spectrum disorders.²²

CONTRAINDICATIONS

Measles vaccine is contraindicated for:

• Patients who have cell-mediated immune deficiencies, except human immunodeficiency virus (HIV) infection
• Pregnant women
• Those who have had a severe allergic reaction to a vaccine component in the past
• Those with moderate or severe acute illness
• Those who have recently received immunoglobulin products.

People with HIV infection who are severely immunosuppressed should not receive live measles vaccine. However, because of the risk of severe measles in HIV-infected patients and because the vaccine has been shown to be safe for patients with HIV without severe immunosuppression, the vaccine is recommended for those with asymptomatic or mildly symptomatic HIV infection who do not have evidence of severe immunosuppression (ie, CD4 lymphocytes < 15% or < 200 cells/µL).^3,4

INFECTION CONTROL AND PREVENTION

All school-age children who received only one dose of measles vaccine should receive the second dose

Healthcare workers should maintain a high index of suspicion for measles and implement isolation procedures promptly in patients with a febrile illness, rash, and a history of travel abroad or contact with travelers from abroad.²³ Suspected cases should be reported promptly to local health agencies to help limit spread.

Patients with measles should be placed in airborne isolation (eg, use of an N95 or higher level respirator and an airborne infection isolation room) for 4 days after the onset of the rash in a normal host and for the duration of the illness in an immunocompromised patient. Healthcare staff, regardless of their immunity status, should adhere to these precautions when entering the room of infected patients.

Immunization programs should be established to ensure that everyone who works or volunteers in healthcare facilities is protected against measles.⁴

Postexposure prophylaxis. Measles vaccination given to susceptible contacts within 72 hours of exposure may provide protection against infection and induces protection against subsequent measles exposures.^24,25

Vaccination is the best intervention for susceptible contacts older than 12 months who do not have a contraindication to measles vaccination, and for those who have received only one dose of measles vaccine.

Passive immunization. Active immunization is the best strategy for controlling measles outbreaks. Passive immunization with intramuscularly or intravenously administered immunoglobulin given within 6 days of exposure can be used to prevent transmission or modify the clinical course of infection for susceptible contacts at high risk of developing severe or fatal measles. This includes people who are being treated with immunosuppressive agents, HIV-infected, pregnant, or younger than 1 year of age.

Measles continues to rear its head in the United States. Because it is so contagious, even the few cases introduced by travelers quickly spread to susceptible contacts. Life-threatening and severely disabling complications can occur, although this is rare. Widespread immunization and prompt recognition and isolation of contacts are key to controlling outbreaks.

This article reviews the epidemiology of measles, describes its distinctive clinical picture, and provides recommendations for infection control and prevention, including in immunosuppressed populations.

MEASLES IS SERIOUS AND HIGHLY CONTAGIOUS

Up to 90% of susceptible people develop measles after exposure, making it one of the most contagious of infections. The virus is transmitted by airborne spread when an infected person coughs or sneezes, or by direct contact with infectious droplets. The virus can remain infectious in the air or on a surface for up to 2 hours.¹

Worldwide, an estimated 20 million people are infected with measles each year, and 146,000 die of complications. In 1980, before widespread vaccination, 2.6 million deaths were attributable to measles annually. In the United States before the introduction of measles vaccine in 1963, measles was a significant cause of disease and death: an estimated 3 to 4 million people were infected annually, although only about 549,000 were reported. There were 48,000 hospitalizations, 1,000 cases of permanent brain damage from measles encephalitis, and 495 deaths annually.²

Outbreaks still occur regularly

In 2000, measles was declared eliminated from the United States,³ but annual outbreaks have occurred since then as a result of cases imported from other countries and their subsequent transmission to unvaccinated people. From 2001 to 2012, a median of four outbreaks and 60 cases were reported annually to the US Centers for Disease Control and Prevention.⁴

In January 2015, a multistate measles outbreak originating in Disneyland in California was recognized. As of April 17, when the outbreak was declared over, 111 measles cases from seven states had been linked to this outbreak.⁵ Of the evaluable cases, 44% were in unvaccinated people and 38% were in those whose vaccination status was unknown or undocumented. The median age of patients was 21, and 20% required hospitalization.

This outbreak, as well as four other smaller US outbreaks the same year, underscores the transmissibility of the virus in populations containing only a small percentage of unvaccinated people.⁶

DISTINCTIVE CLINICAL PICTURE

The incubation period for measles infection is 7 to 21 days, with most cases becoming apparent 10 to 12 days after exposure. Measles should be suspected in a patient with the following clinical features whose history indicates susceptibility and exposure (ie, an unimmunized person with a history of exposure or travel):

Severe acute respiratory illness. Measles usually presents as an acute respiratory viral illness, which typically lasts 2 to 4 days. The illness involves high fevers, malaise, anorexia, and the “three Cs”: cough, coryza (rhinitis), and conjunctivitis. Patients usually appear sicker than those with more common viral illnesses.

From the US Centers for Disease Control and Prevention.

Koplik spots, which are pathognomonic for measles, are seen in the first few days of illness. They are bluish-white, slightly raised lesions on an erythematous base on the buccal mucosa, usually opposite the first molar (Figure 1). Spots can also be seen on the soft palate, conjunctiva, and vaginal mucosa. Koplik spots usually disappear after a few days and often are not appreciable at the time of evaluation.

From the US Centers for Disease Control and Prevention.

Discrete erythematous patches develop on the face and neck a day after the appearance of Koplik spots. This rash becomes more confluent as it spreads to involve the entire body (Figure 2). It typically lasts for 3 to 7 days, then fades in a similar pattern. The confluent nature of this rash and its spread from the face and neck to the entire body are characteristic of measles. Patients are highly contagious from 4 days before the onset of the rash to 4 days after.

COMPLICATIONS CAN BE SEVERE

Those at highest risk for measles complications are infants, children under age 5, adults over age 20, pregnant women, and immunosuppressed individuals.⁷

Pneumonia—either a primary measles pneumonia or a secondary viral or bacterial pneumonia—is the most common cause of death.^8,9 Viruses complicating measles are typically adenovirus and herpes simplex virus. Bacteria causing secondary infection are usually Staphylococcus aureus and Streptococcus pneumoniae and, less commonly, gram-negative bacteria.

Laryngotracheobronchitis (croup) is the second most common cause of death, with bacteria and viruses similar to those causing measles-related pneumonia.

Otitis media is the most common complication of measles. Other respiratory complications include mastoiditis, pneumothorax, and mediastinal emphysema.

Acute measles encephalitis occurs in 1 measles case per 1,000 and often results in permanent brain damage. During the convalescent phase of the illness, fever again emerges, with the development of headaches, seizures, and altered consciousness.¹⁰

Subacute sclerosing panencephalitis is a rare fatal degenerative disease of the central nervous system caused by a persistent infection with a defective measles virus. The precise pathophysiology is unclear, but it is thought that mutations of the viral genome lead to altered cellular immunity.¹¹ The condition typically occurs 7 to 10 years after the initial measles infection, particularly in those who developed measles before age 2. Clinical manifestations include behavioral disturbances, intellectual deterioration, and myoclonic seizures, slowly progressing to a vegetative state and death.¹²

Other complications of measles include diarrhea and stomatitis, which are associated with malnutrition in developing countries, and subclinical hepatitis, thrombocytopenia, appendicitis, ileocolitis, hypokalemia, and myocarditis.

During pregnancy, measles infection can be complicated by primary measles pneumonia and is associated with an increased risk of miscarriage and premature birth.¹³

Patients with a cell-mediated immunodeficiency who develop measles are particularly susceptible to fatal measles pneumonia and acute progressive encephalitis.¹⁴

ATYPICAL MEASLES IN THOSE WHO RECEIVED KILLED VACCINE

From 1963 to 1967, a killed measles vaccine was available in the United States. Those who received this vaccine are susceptible to an atypical form of measles when exposed to the virus,¹⁵ characterized by a 1- to 2-day prodrome, followed by the appearance of a maculopapular or petechial rash on the distal extremities that spreads centripetally. Patients develop high fever and edema of the hands and feet, and have a more prolonged course than with classic measles. It is believed not to be contagious.¹⁶

LABORATORY CONFIRMATION

Laboratory confirmation of measles is recommended for suspected cases. Because viral isolation is technically difficult and is not readily available in most laboratories, measles-specific immunoglobulin M antibody serologic testing is most commonly used. It is almost 100% sensitive when done 2 to 3 days after the onset of the rash.¹⁷

Measles RNA testing by real-time polymerase chain reaction to detect measles virus in the blood, throat, or urine is more specific and if available may be preferred over serologic testing.¹⁸

SUPPORTIVE MANAGEMENT AND VITAMIN A SUPPLEMENTATION

No specific antiviral therapy for measles is available. Management involves supportive measures and monitoring for secondary bacterial complications.

The World Health Organization and the American Academy of Pediatrics recommend vitamin A supplementation for all children with acute measles.¹⁹ In developing countries, it has been shown to reduce rates of morbidity and death in measles-infected children.²⁰ In the United States, children with measles have been found to have low serum levels of vitamin A, with lower levels associated with more severe disease.

VACCINATION RECOMMENDATIONS

The only measles vaccine available in the United States is a live further-attenuated strain prepared in chick embryo cell culture and combined with mumps and rubella vaccine (MMR) or with measles, mumps, rubella, and varicella vaccine (MMRV).

Healthy children. Two doses of measles vaccine are recommended, as a single dose is associated with a 5% failure rate. The recommended schedule is:

• First dose at age 12 to 15 months
• Second dose at the time of school entry (ages 4 to 6), or at any time at least 28 days after the first dose.¹⁹

More than 99% of children who receive two doses of vaccine according to this schedule develop serologic evidence of measles immunity. Vaccination provides long-term immunity, and many epidemiologic studies have documented that waning immunity after vaccination occurs only very rarely.²¹

All school-age children, including elementary, middle, and high school students, who received only one dose of measles vaccine should receive the second dose.

Adults born in 1957 or later should receive at least one dose of measles vaccine unless they have other acceptable evidence of immunity, such as:⁴

• Documentation of age-appropriate live measles vaccine, ie, one dose of vaccine for adults not at high risk, or two doses for those at high risk (see below)
• Laboratory evidence of immunity (ie, measles immunoglobulin G in serum)
• Laboratory confirmation of disease.

Adults born before 1957 can be considered to be immune to measles, although MMR vaccine can be administered in those without contraindications.

Vitamin A supplementation is recommended for acute measles

Adults at increased risk of exposure or transmission of measles and who do not have evidence of immunity should receive two doses of MMR vaccine, given at least 28 days apart. This high-risk group includes:

• Students attending college or other post-high school educational institution
• Healthcare personnel
• International travelers.

During measles outbreaks, every effort should be made to ensure that those at high risk are vaccinated with two doses of MMR or have other acceptable evidence of immunity.

LIVE VACCINE IS SAFE FOR MOST PEOPLE

Mild side effects. A transient fever, which may be accompanied by a discrete or confluent rash, occurs in 5% to 15% of recipients 5 to 12 days after vaccination.

Transmission does not occur. People who have been newly vaccinated do not transmit the virus to susceptible contacts, even if they develop a vaccine-associated rash. The vaccine can safely be given to close contacts of immunocompromised and other susceptible people.

Egg allergy not a concern. Measles vaccine is produced in chick embryo cell culture but has been shown to be safe for people with egg allergy and is recommended without the need for egg allergy testing.¹⁹

Autism link debunked. No scientific evidence shows that the risk of autism is higher in children who receive MMR vaccine than in those who do not. In 2001, an Institute of Medicine report rejected a causal relationship between MMR vaccine and autism spectrum disorders.²²

CONTRAINDICATIONS

Measles vaccine is contraindicated for:

• Patients who have cell-mediated immune deficiencies, except human immunodeficiency virus (HIV) infection
• Pregnant women
• Those who have had a severe allergic reaction to a vaccine component in the past
• Those with moderate or severe acute illness
• Those who have recently received immunoglobulin products.

People with HIV infection who are severely immunosuppressed should not receive live measles vaccine. However, because of the risk of severe measles in HIV-infected patients and because the vaccine has been shown to be safe for patients with HIV without severe immunosuppression, the vaccine is recommended for those with asymptomatic or mildly symptomatic HIV infection who do not have evidence of severe immunosuppression (ie, CD4 lymphocytes < 15% or < 200 cells/µL).^3,4

INFECTION CONTROL AND PREVENTION

All school-age children who received only one dose of measles vaccine should receive the second dose

Healthcare workers should maintain a high index of suspicion for measles and implement isolation procedures promptly in patients with a febrile illness, rash, and a history of travel abroad or contact with travelers from abroad.²³ Suspected cases should be reported promptly to local health agencies to help limit spread.

Patients with measles should be placed in airborne isolation (eg, use of an N95 or higher level respirator and an airborne infection isolation room) for 4 days after the onset of the rash in a normal host and for the duration of the illness in an immunocompromised patient. Healthcare staff, regardless of their immunity status, should adhere to these precautions when entering the room of infected patients.

Immunization programs should be established to ensure that everyone who works or volunteers in healthcare facilities is protected against measles.⁴

Postexposure prophylaxis. Measles vaccination given to susceptible contacts within 72 hours of exposure may provide protection against infection and induces protection against subsequent measles exposures.^24,25

Vaccination is the best intervention for susceptible contacts older than 12 months who do not have a contraindication to measles vaccination, and for those who have received only one dose of measles vaccine.

Passive immunization. Active immunization is the best strategy for controlling measles outbreaks. Passive immunization with intramuscularly or intravenously administered immunoglobulin given within 6 days of exposure can be used to prevent transmission or modify the clinical course of infection for susceptible contacts at high risk of developing severe or fatal measles. This includes people who are being treated with immunosuppressive agents, HIV-infected, pregnant, or younger than 1 year of age.

A 72-year-old woman presented to the emergency room with persistent substernal chest discomfort. The initial electrocardiogram (ECG) showed 2.0-mm ST-segment elevation in leads II, III, and aVF, and 1.0-mm ST-segment elevation in lead V₆ (Figure 1). The index troponin T level was 1.5 ng/mL (reference range < 0.01). ST-elevation myocardial infarction protocols were activated, and she was taken for urgent catheterization.

Coronary angiography showed normal coronary arteries. However, intraprocedural left ventriculography identified circumferential midventricular and apical akinesis with compensatory basal hyperkinesis (Figure 2).

Further inquiry into the patient’s medical history revealed that she had been experiencing psychological distress brought on by the failure of businesses she owned.

Transthoracic echocardiography subsequently verified a depressed ejection fraction (30%) with prominent apical and midventricular wall akinesis, thus confirming the diagnosis of stress cardiomyopathy. She was discharged  home on a low-dose beta-blocker and an angiotensin-converting enzyme inhibitor, and within 6 weeks her systolic function had completely returned to normal.

STRESS CARDIOMYOPATHY: CAUSES, DIAGNOSIS, PROGNOSIS

Stress cardiomyopathy—also called broken heart syndrome, stress-induced cardiomyopathy, takotsubo cardiomyopathy, and apical ballooning syndrome—is an increasingly recognized acquired cardiomyopathy that typically affects older postmenopausal women exposed to a triggering stressor such as severe medical illness, major surgery, or a psychologically stressful life event.^1,2

Our patient’s acute presentation is a classic example of how stress cardiomyopathy can be indistinguishable from acute coronary syndrome. It has been estimated that 1% to 2% of patients presenting with an initial diagnosis of acute coronary syndrome actually have stress cardiomyopathy.²

Diagnostic criteria

The diagnosis of stress cardiomyopathy can be established when coronary angiography reveals nonobstructive coronary arteries in patients with abnormal ventricular wall motion identified on echocardiography or ventriculography, or both. These findings are part of the proposed diagnostic criteria²:

• Transient hypokinesis, akinesis, or dyskinesis of the left ventricle midsegments, with or without apical involvement; regional wall motion abnormalities extending beyond a single epicardial vascular distribution; usually, a psychological or physiologic stressor is present
• No obstructive coronary disease or no angiographic evidence of acute plaque rupture
• New abnormalities on ECG, or modest elevation in cardiac enzymes
• No evidence of pheochromocytoma or myocarditis.²

Other characteristics that help to differentiate stress cardiomyopathy from acute coronary syndrome include a prolonged QTc interval, attenuation of the QRS amplitude, and a decreased troponin-ejection fraction product.^3–5

Prognosis

The prognosis is generally excellent, with most patients achieving full recovery of myocardial function within several weeks.² However, in the acute setting, there are relatively high rates of acute heart failure (44% to 46%), left ventricular outflow tract obstruction (19%), and unstable ventricular arrhythmias (3.4%), including torsades de pointes.^1,2,6,7

Stress cardiomyopathy recurs in approximately 11% of patients within 4 years.⁸ Death is considered a rare complication but has occurred in as many as 8% of reported cases.¹

A 72-year-old woman presented to the emergency room with persistent substernal chest discomfort. The initial electrocardiogram (ECG) showed 2.0-mm ST-segment elevation in leads II, III, and aVF, and 1.0-mm ST-segment elevation in lead V₆ (Figure 1). The index troponin T level was 1.5 ng/mL (reference range < 0.01). ST-elevation myocardial infarction protocols were activated, and she was taken for urgent catheterization.

Coronary angiography showed normal coronary arteries. However, intraprocedural left ventriculography identified circumferential midventricular and apical akinesis with compensatory basal hyperkinesis (Figure 2).

Further inquiry into the patient’s medical history revealed that she had been experiencing psychological distress brought on by the failure of businesses she owned.

Transthoracic echocardiography subsequently verified a depressed ejection fraction (30%) with prominent apical and midventricular wall akinesis, thus confirming the diagnosis of stress cardiomyopathy. She was discharged  home on a low-dose beta-blocker and an angiotensin-converting enzyme inhibitor, and within 6 weeks her systolic function had completely returned to normal.

STRESS CARDIOMYOPATHY: CAUSES, DIAGNOSIS, PROGNOSIS

Stress cardiomyopathy—also called broken heart syndrome, stress-induced cardiomyopathy, takotsubo cardiomyopathy, and apical ballooning syndrome—is an increasingly recognized acquired cardiomyopathy that typically affects older postmenopausal women exposed to a triggering stressor such as severe medical illness, major surgery, or a psychologically stressful life event.^1,2

Our patient’s acute presentation is a classic example of how stress cardiomyopathy can be indistinguishable from acute coronary syndrome. It has been estimated that 1% to 2% of patients presenting with an initial diagnosis of acute coronary syndrome actually have stress cardiomyopathy.²

Diagnostic criteria

The diagnosis of stress cardiomyopathy can be established when coronary angiography reveals nonobstructive coronary arteries in patients with abnormal ventricular wall motion identified on echocardiography or ventriculography, or both. These findings are part of the proposed diagnostic criteria²:

• Transient hypokinesis, akinesis, or dyskinesis of the left ventricle midsegments, with or without apical involvement; regional wall motion abnormalities extending beyond a single epicardial vascular distribution; usually, a psychological or physiologic stressor is present
• No obstructive coronary disease or no angiographic evidence of acute plaque rupture
• New abnormalities on ECG, or modest elevation in cardiac enzymes
• No evidence of pheochromocytoma or myocarditis.²

Other characteristics that help to differentiate stress cardiomyopathy from acute coronary syndrome include a prolonged QTc interval, attenuation of the QRS amplitude, and a decreased troponin-ejection fraction product.^3–5

Prognosis

The prognosis is generally excellent, with most patients achieving full recovery of myocardial function within several weeks.² However, in the acute setting, there are relatively high rates of acute heart failure (44% to 46%), left ventricular outflow tract obstruction (19%), and unstable ventricular arrhythmias (3.4%), including torsades de pointes.^1,2,6,7

Stress cardiomyopathy recurs in approximately 11% of patients within 4 years.⁸ Death is considered a rare complication but has occurred in as many as 8% of reported cases.¹

A 72-year-old woman presented to the emergency room with persistent substernal chest discomfort. The initial electrocardiogram (ECG) showed 2.0-mm ST-segment elevation in leads II, III, and aVF, and 1.0-mm ST-segment elevation in lead V₆ (Figure 1). The index troponin T level was 1.5 ng/mL (reference range < 0.01). ST-elevation myocardial infarction protocols were activated, and she was taken for urgent catheterization.

Coronary angiography showed normal coronary arteries. However, intraprocedural left ventriculography identified circumferential midventricular and apical akinesis with compensatory basal hyperkinesis (Figure 2).

Further inquiry into the patient’s medical history revealed that she had been experiencing psychological distress brought on by the failure of businesses she owned.

Transthoracic echocardiography subsequently verified a depressed ejection fraction (30%) with prominent apical and midventricular wall akinesis, thus confirming the diagnosis of stress cardiomyopathy. She was discharged  home on a low-dose beta-blocker and an angiotensin-converting enzyme inhibitor, and within 6 weeks her systolic function had completely returned to normal.

STRESS CARDIOMYOPATHY: CAUSES, DIAGNOSIS, PROGNOSIS

Stress cardiomyopathy—also called broken heart syndrome, stress-induced cardiomyopathy, takotsubo cardiomyopathy, and apical ballooning syndrome—is an increasingly recognized acquired cardiomyopathy that typically affects older postmenopausal women exposed to a triggering stressor such as severe medical illness, major surgery, or a psychologically stressful life event.^1,2

Our patient’s acute presentation is a classic example of how stress cardiomyopathy can be indistinguishable from acute coronary syndrome. It has been estimated that 1% to 2% of patients presenting with an initial diagnosis of acute coronary syndrome actually have stress cardiomyopathy.²

Diagnostic criteria

The diagnosis of stress cardiomyopathy can be established when coronary angiography reveals nonobstructive coronary arteries in patients with abnormal ventricular wall motion identified on echocardiography or ventriculography, or both. These findings are part of the proposed diagnostic criteria²:

• Transient hypokinesis, akinesis, or dyskinesis of the left ventricle midsegments, with or without apical involvement; regional wall motion abnormalities extending beyond a single epicardial vascular distribution; usually, a psychological or physiologic stressor is present
• No obstructive coronary disease or no angiographic evidence of acute plaque rupture
• New abnormalities on ECG, or modest elevation in cardiac enzymes
• No evidence of pheochromocytoma or myocarditis.²

Other characteristics that help to differentiate stress cardiomyopathy from acute coronary syndrome include a prolonged QTc interval, attenuation of the QRS amplitude, and a decreased troponin-ejection fraction product.^3–5

Prognosis

The prognosis is generally excellent, with most patients achieving full recovery of myocardial function within several weeks.² However, in the acute setting, there are relatively high rates of acute heart failure (44% to 46%), left ventricular outflow tract obstruction (19%), and unstable ventricular arrhythmias (3.4%), including torsades de pointes.^1,2,6,7

Stress cardiomyopathy recurs in approximately 11% of patients within 4 years.⁸ Death is considered a rare complication but has occurred in as many as 8% of reported cases.¹

An 85-year-old woman was brought to our intensive care unit because of worsening hypoxemia over the past day. About 3 weeks earlier she had been diagnosed with acute bilateral pulmonary emboli in the distal branches of the left and right lower lobes and right middle lobe, for which she was receiving anticoagulation therapy.

At presentation she had generalized fatigue and dyspnea at rest that was worse with exertion, but she denied having fever, chest pain, or cough. Her medical history included hypertension, hyperlipidemia, hypothyroidism, stage 1 breast cancer in remission, thromboembolic stroke, and myasthenia gravis. Before her hospital admission, she had been taking rosuvastatin, metoprolol tartrate, pyridostigmine, prednisone, furosemide, levothyroxine, and rivaroxaban. She did not smoke, she was retired, and she had not traveled recently.

Her blood pressure was 135/66 mm Hg, pulse 73 beats per minute, respiratory rate 16, temperature 35.4ºC (95.7ºF), and oxygen saturation 88% while receiving oxygen at 6 L/min via nasal cannula. Physical examination revealed mild edema in the lower extremities and basilar decreased breath sounds. She had no finger clubbing or cyanosis and was not using accessory muscles to breathe. Of note, her oxygen saturation remained more than 93% when she was recumbent but sharply dropped to less than 85% when she was upright.

Laboratory values

Results of initial laboratory testing were as follows:

• Sodium 138 mmol/L (reference range 132–148)
• Potassium 4.2 mmol/L (3.5–5.0)
• Chloride 99 mmol/L (98–111)
• Bicarbonate 29 mmol/L (23–32)
• Creatinine 0.52 mg/dL (0.7–1.4).
• White blood cell count 11.06 × 10⁹/L (3.7–11.0)
• Hemoglobin 12.6 g/dL (12–16)
• Platelet count 211 × 10⁹/L (150–400).
• International normalized ratio 1.4.

Electrocardiography and imaging studies

Standard 12-lead electrocardiography  showed normal sinus rhythm with left axis deviation and left ventricular hypertrophy.

Chest radiography showed bilateral interstitial opacities and small pleural effusions.

Computed tomography (CT) of the chest with contrast, compared with a CT scan done 20 days earlier, showed that the pulmonary emboli had resolved.

Arterial blood gases

In view of her positional hypoxemia, blood for arterial blood gas measurements was drawn in the supine and upright positions.

Supine, with a fraction of inspired oxygen (Fio₂) via high-flow nasal cannula of 45%, her values were:

• pH 7.45 (reference range 7.35–7.45)
• Pco₂ 34 mm Hg (36–46)
• Po₂ 81 mm Hg (85–95)
• Bicarbonate 23 mmol/L (22–26).

Upright, her hypoxemia was significantly worse:

• pH 7.46
• Pco₂ 33 mm Hg
• Po₂ 57 mm Hg
• Bicarbonate 23 mmol/L.

The methemoglobin level was normal on both measurements.

During her stay in the intensive care unit, she required up to 100% Fio₂ because of persistent hypoxemia.

CAUSES OF HYPOXEMIA

1. So far, the patient’s laboratory tests and imaging studies point to which of the following as the most likely cause of her severe hypoxemia?

• Ventilation-perfusion (V/Q) mismatch
• Diffusion abnormality
• Hypoventilation
• Shunting
• None of the above

The arterial blood gas measurements suggested the possibility of shunting as the cause, although further imaging would be needed to confirm that.

V/Q mismatch can occur in respiratory failure due to pulmonary embolism, pulmonary edema, or shunting. If ventilation is preserved but perfusion is impaired, the V/Q ratio approaches infinity (dead-space ventilation), a situation that can be seen in pulmonary embolism. If perfusion is preserved and ventilation is impaired, the V/Q ratio approaches zero, which is consistent with a physiologic shunt.

Hypoxemia may improve in less severe forms of V/Q mismatch. In our patient, the repeat CT with contrast showed that her pulmonary embolism had resolved, so this is probably not the cause of her severe hypoxemia.

Diffusion abnormalities are due to defects in the lung parenchyma, such as in chronic obstructive pulmonary disease, interstitial lung disease, and lung fibrosis.

Hypoxemia from diffusion defects is usually aggravated by a precipitating factor that increases oxygen demand, and it usually improves with oxygen supplementation. This is unlikely in our patient, as she did not have a history of chronic interstitial lung disease and CT showed no evidence of severe lung parenchymal disease.

Hypoventilation is usually due to drugs that cause respiratory depression, to stroke, or to neuromuscular diseases such as myasthenia gravis that can cause respiratory muscle weakness. It results in elevation of Pco₂ and, if not corrected, respiratory acidosis.

Our patient had a diagnosis of myasthenia gravis, though hypoventilation is unlikely in her case because she had a normal respiratory rate and low Pco₂ values.

Shunting can be physiologic or anatomic and can occur in the heart or the lungs. In physiologic shunting, severe V/Q mismatch can occur when ventilation is affected, as in severe pulmonary edema and pneumonia. In anatomic shunting, a defect such as an atrial septal defect or a pulmonary arteriovenous malformation allows blood to bypass areas of ventilation from the venous to the arterial circulation, preventing it from being oxygenated. In true anatomic shunting, supplemental oxygen with 100% Fio₂ has little effect, whereas in V/Q mismatch it can raise the arterial oxygen saturation.

Our patient’s radiograph did not suggest severe pneumonia or pulmonary edema, which makes these unlikely causes of her hypoxemia. At this point, because of her positional hypoxemia, further evaluation with contrast-enhanced echocardiography was needed to evaluate for anatomic shunting in the heart or lungs.

FURTHER TESTING

Transthoracic echocardiography (TTE) with agitated saline with a Valsalva maneuver was performed. Normally, no bubbles are seen in  the left-sided chambers after intravenous injection of agitated saline contrast, whereas bubbles on the left side suggest an intracardiac or intrapulmonary shunt. In our patient, this test was negative, and her right ventricular systolic pressure was normal.

2. What further testing should be considered to evaluate our patient’s hypoxemia?

• High-resolution chest CT
• Transesophageal echocardiography (TEE)
• Pulmonary function testing
• Repeated arterial blood gas measurement
• Edrophonium testing

Repeat imaging with high-resolution CT would likely not provide additional information and would expose the patient to additional radiation without adding much clinical benefit.

TEE could help further evaluate the intracardiac anatomy and look for shunting, which may be missed on TTE because of suboptimal positioning or image quality.

Pulmonary function testing is useful in establishing the baseline function and impairment in respiratory volumes. If an acute myasthenic crisis is suspected, measuring the negative inspiratory force and the forced vital capacity can be useful in monitoring worsening respiratory muscle weakness and assessing the need for mechanical ventilation.

In our patient, it is unlikely that pulmonary function testing would help, since her acute respiratory failure was probably not caused by neuromuscular weakness.

Repeated arterial blood gas measurement would likely only confirm that the patient still has positional hypoxemia but would not help sort through the differential diagnosis.

Edrophonium testing is useful in diagnosing myasthenia gravis and differentiating it from other neuromuscular diseases, such as Lambert-Eaton syndrome. Edrophonium, a reversible acetylcholinesterase inhibitor, prevents degradation of acetylcholine and prolongs its effect at the synaptic cleft, thus improving muscle weakness.

Our patient has already been diagnosed with myasthenia gravis, so this test is not likely to uncover the cause of her hypoxemia.

Because we still strongly suspected a shunt, TEE was performed with intravenous injection of agitated saline. TEE with the patient upright revealed intracardiac right-to-left shunting through a patent foramen ovale. The midesophageal view after saline injection showed a large interatrial septal aneurysm with total excursion of 2 cm, and right-to-left shunting within the first beat, consistent with an intracardiac shunt (Figure 1). Color Doppler imaging (Figure 2) demonstrated turbulent flow through the patent foramen ovale, consistent with right-to-left shunting, and also showed the patent foramen ovale in a closed position (Figure 3).

3. Which is now most likely the cause of our patient’s hypoxemia?

• Chronic thromboembolic pulmonary hypertension
• Myasthenic crisis
• Platypnea-orthodeoxia syndrome due to the patent foramen ovale
• Methemoglobinemia

Chronic thromboembolic pulmonary hypertension is usually a long-term result of untreated or inadequately treated thromboembolic disease (eg, pulmonary emboli), which causes vascular remodeling and pulmonary arteriopathy, which in turn leads to increased pulmonary vascular resistance and pulmonary hypertension.

This is unlikely the cause of our patient’s acute hypoxemia, as her symptoms did not suggest it. Moreover, an elevated right ventricular systolic pressure on TTE would suggest pulmonary hypertension, but TTE did not show this, and repeat chest CT indicated that her pulmonary embolism had been adequately treated and had resolved. A V/Q scan and right heart catheterization would help rule out chronic thromboembolic pulmonary hypertension, although these were not done in our patient.

Myasthenic crisis is the progressive fatiguing and paralysis of respiratory muscles ultimately requiring mechanical ventilation to sustain life. It is often brought on by infection or drug therapy.

Our patient did not require intubation and she had no signs or symptoms of myasthenic crisis such as ptosis, dysphagia, or dysarthria. She had a negative inspiratory force of −21 cm H₂O, and pulmonary function testing 4 days before her hospital admission had shown a forced vital capacity of 1.84 L, making myasthenic crisis an unlikely cause of her respiratory failure.

Platypnea-orthodeoxia syndrome is a syndrome of dyspnea (platypnea) and hypoxemia (orthodeoxia) that is induced by sitting upright or standing and resolves when lying down. It is a result of right-to-left intracardiac or intrapulmonary shunting in the presence of an anatomic defect and a functional element causing redirection of shunt flow through the anatomic defect in an upright position.¹ It is associated with specific cardiac, pulmonary, and hepatic diseases, such as atrial septal defect, pulmonary arteriovenous malformation, and hepatopulmonary syndrome.² It can occur even if right-sided chamber pressures are normal, and several mechanisms of the underlying pathophysiology have been described.³

Platypnea-orthodeoxia syndrome can be triggered by an event that causes a spontaneous transient elevation of right atrial pressure and pulmonary hypertension, such as our patient’s acute pulmonary embolism. Increased right-to-left shunting occurs in an upright position, causing preferential redirection of flow from the inferior vena cava through the interatrial septum and the patent foramen ovale.⁴

Our patient was elderly and, like one in every four people in the world, she had had a patent foramen ovale since the day she was born. Never causing a problem, it had remained undiagnosed until complicated by platypnea-orthodeoxia syndrome after her recent pulmonary embolism.

Methemoglobinemia. Methemoglobin has a lower affinity for oxygen than normal hemoglobin. Elevations usually occur with medications such as anesthetics and nitrates and can be diagnosed through an elevated level on arterial blood gas testing.

Our patient did not have elevated methemoglobin on her blood gas measurements on admission; therefore, this is unlikely to be the diagnosis.

CASE CONCLUDED

Percutaneous closure of the patent foramen ovale with a 30-mm Amplatzer Cribriform Occluder brought significant improvement in our patient’s functional status and arterial oxygenation saturation, and 2 weeks later at follow-up she no longer needed  supplemental oxygen. TEE 6 months later showed an intact closure device and no interatrial shunting.

WHEN HYPOXEMIA DOES NOT RESPOND TO OXYGEN

In the intensive care unit, time is critical, and when hypoxia is refractory to high Fio₂, shunting should be considered.

In the acute-care setting, platypnea-orthodeoxia syndrome can be identified quickly by pulse oximetry and serial blood gas measurements in the upright and supine positions. A drop in arterial oxygenation in the upright position vs the supine position helps confirm the diagnosis.

Other conditions in the differential diagnosis of this syndrome include recurrent pulmonary embolism, acute respiratory distress syndrome, interstitial pulmonary fibrosis, intrapulmonary shunting due to arteriovenous malformation, and diaphragm paralysis due to neuromuscular disease.

In our patient, positional blood gas measurements demonstrated a significant drop in arterial oxygen saturation from the supine to the upright position, raising our suspicion of shunting. It helped narrow the differential diagnosis and guided our selection of additional diagnostic tests.

The initial chest radiograph in our patient was normal. TTE did not reveal shunting and showed a normal right ventricular systolic pressure. TTE with agitated saline also failed to reveal shunting. Because of suboptimal positioning and image quality, TTE may miss the shunting physiology, and that is why we proceeded to positional TEE, which can better evaluate the hemodynamic effects of positional changes on patent foramen ovale and shunting. 

MORE ABOUT PATENT FORAMEN OVALE

The prevalence of patent foramen ovale is estimated at 27% in the general population, but it is usually not symptomatic. It can be associated with atrial septal aneurysm and Chiari network malformations. When associated with atrial septal aneurysm, it carries a higher risk of stroke.⁵

Our patient had a large atrial septal aneurysm with a septal excursion of 2 cm as well as a history of thromboembolic stroke, which was likely associated with the patent foramen ovale and the atrial septal aneurysm.

Atrial septal aneurysm is rare, with a prevalence of 1% at autopsy and 1.9% by TTE. It is defined by a septal excursion of at least 10 mm and a base diameter of at least 15 mm and is more frequently detected on TEE than on TTE.⁶

Studies have shown that contrast and color Doppler TEE are superior to TTE for detecting patent foramen ovale.⁷ Tilt-table TEE  with contrast enhancement has also been used to better demonstrate the morphology of the interatrial septum and the degree of shunting due to the separation between the septum primum and septum secundum causing the patent foramen ovale.⁸ Contrast-enhanced transcranial Doppler has also been shown comparable to contrast TEE to detect interatrial shunting. However, TEE provides additional anatomic information.⁹

In our patient, atrial septal aneurysm and patent foramen ovale were exaggerated by upright positioning, which opened the aneurysm and increased the shunting through the patent foramen ovale.

The treatment of choice in symptomatic patients with platypnea-orthodeoxia syndrome is directed at the underlying cause, in this case closure of the foramen ovale. This treatment has been shown to be safe and effective in these patients,¹⁰ but caution should be used when considering foramen ovale closure in patients with pulmonary hypertension.¹¹

In patients with irreversible or severe pulmonary hypertension, closure of the patent foramen ovale can exacerbate right heart dysfunction and lead to right heart failure. There are situations when closure of a patent foramen ovale can be considered in pulmonary hypertension; however, each decision is individualized, and caution must be used. A detailed discussion is beyond the scope of this paper.

A thorough history and physical examination are important in differentiating the various causes of hypoxemia. Appropriate diagnostic testing is needed along with prompt treatment of the underlying cause of platyp­nea-orthodeoxia syndrome.

An 85-year-old woman was brought to our intensive care unit because of worsening hypoxemia over the past day. About 3 weeks earlier she had been diagnosed with acute bilateral pulmonary emboli in the distal branches of the left and right lower lobes and right middle lobe, for which she was receiving anticoagulation therapy.

At presentation she had generalized fatigue and dyspnea at rest that was worse with exertion, but she denied having fever, chest pain, or cough. Her medical history included hypertension, hyperlipidemia, hypothyroidism, stage 1 breast cancer in remission, thromboembolic stroke, and myasthenia gravis. Before her hospital admission, she had been taking rosuvastatin, metoprolol tartrate, pyridostigmine, prednisone, furosemide, levothyroxine, and rivaroxaban. She did not smoke, she was retired, and she had not traveled recently.

Her blood pressure was 135/66 mm Hg, pulse 73 beats per minute, respiratory rate 16, temperature 35.4ºC (95.7ºF), and oxygen saturation 88% while receiving oxygen at 6 L/min via nasal cannula. Physical examination revealed mild edema in the lower extremities and basilar decreased breath sounds. She had no finger clubbing or cyanosis and was not using accessory muscles to breathe. Of note, her oxygen saturation remained more than 93% when she was recumbent but sharply dropped to less than 85% when she was upright.

Laboratory values

Results of initial laboratory testing were as follows:

• Sodium 138 mmol/L (reference range 132–148)
• Potassium 4.2 mmol/L (3.5–5.0)
• Chloride 99 mmol/L (98–111)
• Bicarbonate 29 mmol/L (23–32)
• Creatinine 0.52 mg/dL (0.7–1.4).
• White blood cell count 11.06 × 10⁹/L (3.7–11.0)
• Hemoglobin 12.6 g/dL (12–16)
• Platelet count 211 × 10⁹/L (150–400).
• International normalized ratio 1.4.

Electrocardiography and imaging studies

Standard 12-lead electrocardiography  showed normal sinus rhythm with left axis deviation and left ventricular hypertrophy.

Chest radiography showed bilateral interstitial opacities and small pleural effusions.

Computed tomography (CT) of the chest with contrast, compared with a CT scan done 20 days earlier, showed that the pulmonary emboli had resolved.

Arterial blood gases

In view of her positional hypoxemia, blood for arterial blood gas measurements was drawn in the supine and upright positions.

Supine, with a fraction of inspired oxygen (Fio₂) via high-flow nasal cannula of 45%, her values were:

• pH 7.45 (reference range 7.35–7.45)
• Pco₂ 34 mm Hg (36–46)
• Po₂ 81 mm Hg (85–95)
• Bicarbonate 23 mmol/L (22–26).

Upright, her hypoxemia was significantly worse:

• pH 7.46
• Pco₂ 33 mm Hg
• Po₂ 57 mm Hg
• Bicarbonate 23 mmol/L.

The methemoglobin level was normal on both measurements.

During her stay in the intensive care unit, she required up to 100% Fio₂ because of persistent hypoxemia.

CAUSES OF HYPOXEMIA

1. So far, the patient’s laboratory tests and imaging studies point to which of the following as the most likely cause of her severe hypoxemia?

• Ventilation-perfusion (V/Q) mismatch
• Diffusion abnormality
• Hypoventilation
• Shunting
• None of the above

The arterial blood gas measurements suggested the possibility of shunting as the cause, although further imaging would be needed to confirm that.

V/Q mismatch can occur in respiratory failure due to pulmonary embolism, pulmonary edema, or shunting. If ventilation is preserved but perfusion is impaired, the V/Q ratio approaches infinity (dead-space ventilation), a situation that can be seen in pulmonary embolism. If perfusion is preserved and ventilation is impaired, the V/Q ratio approaches zero, which is consistent with a physiologic shunt.

Hypoxemia may improve in less severe forms of V/Q mismatch. In our patient, the repeat CT with contrast showed that her pulmonary embolism had resolved, so this is probably not the cause of her severe hypoxemia.

Diffusion abnormalities are due to defects in the lung parenchyma, such as in chronic obstructive pulmonary disease, interstitial lung disease, and lung fibrosis.

Hypoxemia from diffusion defects is usually aggravated by a precipitating factor that increases oxygen demand, and it usually improves with oxygen supplementation. This is unlikely in our patient, as she did not have a history of chronic interstitial lung disease and CT showed no evidence of severe lung parenchymal disease.

Hypoventilation is usually due to drugs that cause respiratory depression, to stroke, or to neuromuscular diseases such as myasthenia gravis that can cause respiratory muscle weakness. It results in elevation of Pco₂ and, if not corrected, respiratory acidosis.

Our patient had a diagnosis of myasthenia gravis, though hypoventilation is unlikely in her case because she had a normal respiratory rate and low Pco₂ values.

Shunting can be physiologic or anatomic and can occur in the heart or the lungs. In physiologic shunting, severe V/Q mismatch can occur when ventilation is affected, as in severe pulmonary edema and pneumonia. In anatomic shunting, a defect such as an atrial septal defect or a pulmonary arteriovenous malformation allows blood to bypass areas of ventilation from the venous to the arterial circulation, preventing it from being oxygenated. In true anatomic shunting, supplemental oxygen with 100% Fio₂ has little effect, whereas in V/Q mismatch it can raise the arterial oxygen saturation.

Our patient’s radiograph did not suggest severe pneumonia or pulmonary edema, which makes these unlikely causes of her hypoxemia. At this point, because of her positional hypoxemia, further evaluation with contrast-enhanced echocardiography was needed to evaluate for anatomic shunting in the heart or lungs.

FURTHER TESTING

Transthoracic echocardiography (TTE) with agitated saline with a Valsalva maneuver was performed. Normally, no bubbles are seen in  the left-sided chambers after intravenous injection of agitated saline contrast, whereas bubbles on the left side suggest an intracardiac or intrapulmonary shunt. In our patient, this test was negative, and her right ventricular systolic pressure was normal.

2. What further testing should be considered to evaluate our patient’s hypoxemia?

• High-resolution chest CT
• Transesophageal echocardiography (TEE)
• Pulmonary function testing
• Repeated arterial blood gas measurement
• Edrophonium testing

Repeat imaging with high-resolution CT would likely not provide additional information and would expose the patient to additional radiation without adding much clinical benefit.

TEE could help further evaluate the intracardiac anatomy and look for shunting, which may be missed on TTE because of suboptimal positioning or image quality.

Pulmonary function testing is useful in establishing the baseline function and impairment in respiratory volumes. If an acute myasthenic crisis is suspected, measuring the negative inspiratory force and the forced vital capacity can be useful in monitoring worsening respiratory muscle weakness and assessing the need for mechanical ventilation.

In our patient, it is unlikely that pulmonary function testing would help, since her acute respiratory failure was probably not caused by neuromuscular weakness.

Repeated arterial blood gas measurement would likely only confirm that the patient still has positional hypoxemia but would not help sort through the differential diagnosis.

Edrophonium testing is useful in diagnosing myasthenia gravis and differentiating it from other neuromuscular diseases, such as Lambert-Eaton syndrome. Edrophonium, a reversible acetylcholinesterase inhibitor, prevents degradation of acetylcholine and prolongs its effect at the synaptic cleft, thus improving muscle weakness.

Our patient has already been diagnosed with myasthenia gravis, so this test is not likely to uncover the cause of her hypoxemia.

Because we still strongly suspected a shunt, TEE was performed with intravenous injection of agitated saline. TEE with the patient upright revealed intracardiac right-to-left shunting through a patent foramen ovale. The midesophageal view after saline injection showed a large interatrial septal aneurysm with total excursion of 2 cm, and right-to-left shunting within the first beat, consistent with an intracardiac shunt (Figure 1). Color Doppler imaging (Figure 2) demonstrated turbulent flow through the patent foramen ovale, consistent with right-to-left shunting, and also showed the patent foramen ovale in a closed position (Figure 3).

3. Which is now most likely the cause of our patient’s hypoxemia?

• Chronic thromboembolic pulmonary hypertension
• Myasthenic crisis
• Platypnea-orthodeoxia syndrome due to the patent foramen ovale
• Methemoglobinemia

Chronic thromboembolic pulmonary hypertension is usually a long-term result of untreated or inadequately treated thromboembolic disease (eg, pulmonary emboli), which causes vascular remodeling and pulmonary arteriopathy, which in turn leads to increased pulmonary vascular resistance and pulmonary hypertension.

This is unlikely the cause of our patient’s acute hypoxemia, as her symptoms did not suggest it. Moreover, an elevated right ventricular systolic pressure on TTE would suggest pulmonary hypertension, but TTE did not show this, and repeat chest CT indicated that her pulmonary embolism had been adequately treated and had resolved. A V/Q scan and right heart catheterization would help rule out chronic thromboembolic pulmonary hypertension, although these were not done in our patient.

Myasthenic crisis is the progressive fatiguing and paralysis of respiratory muscles ultimately requiring mechanical ventilation to sustain life. It is often brought on by infection or drug therapy.

Our patient did not require intubation and she had no signs or symptoms of myasthenic crisis such as ptosis, dysphagia, or dysarthria. She had a negative inspiratory force of −21 cm H₂O, and pulmonary function testing 4 days before her hospital admission had shown a forced vital capacity of 1.84 L, making myasthenic crisis an unlikely cause of her respiratory failure.

Platypnea-orthodeoxia syndrome is a syndrome of dyspnea (platypnea) and hypoxemia (orthodeoxia) that is induced by sitting upright or standing and resolves when lying down. It is a result of right-to-left intracardiac or intrapulmonary shunting in the presence of an anatomic defect and a functional element causing redirection of shunt flow through the anatomic defect in an upright position.¹ It is associated with specific cardiac, pulmonary, and hepatic diseases, such as atrial septal defect, pulmonary arteriovenous malformation, and hepatopulmonary syndrome.² It can occur even if right-sided chamber pressures are normal, and several mechanisms of the underlying pathophysiology have been described.³

Platypnea-orthodeoxia syndrome can be triggered by an event that causes a spontaneous transient elevation of right atrial pressure and pulmonary hypertension, such as our patient’s acute pulmonary embolism. Increased right-to-left shunting occurs in an upright position, causing preferential redirection of flow from the inferior vena cava through the interatrial septum and the patent foramen ovale.⁴

Our patient was elderly and, like one in every four people in the world, she had had a patent foramen ovale since the day she was born. Never causing a problem, it had remained undiagnosed until complicated by platypnea-orthodeoxia syndrome after her recent pulmonary embolism.

Methemoglobinemia. Methemoglobin has a lower affinity for oxygen than normal hemoglobin. Elevations usually occur with medications such as anesthetics and nitrates and can be diagnosed through an elevated level on arterial blood gas testing.

Our patient did not have elevated methemoglobin on her blood gas measurements on admission; therefore, this is unlikely to be the diagnosis.

CASE CONCLUDED

Percutaneous closure of the patent foramen ovale with a 30-mm Amplatzer Cribriform Occluder brought significant improvement in our patient’s functional status and arterial oxygenation saturation, and 2 weeks later at follow-up she no longer needed  supplemental oxygen. TEE 6 months later showed an intact closure device and no interatrial shunting.

WHEN HYPOXEMIA DOES NOT RESPOND TO OXYGEN

In the intensive care unit, time is critical, and when hypoxia is refractory to high Fio₂, shunting should be considered.

In the acute-care setting, platypnea-orthodeoxia syndrome can be identified quickly by pulse oximetry and serial blood gas measurements in the upright and supine positions. A drop in arterial oxygenation in the upright position vs the supine position helps confirm the diagnosis.

Other conditions in the differential diagnosis of this syndrome include recurrent pulmonary embolism, acute respiratory distress syndrome, interstitial pulmonary fibrosis, intrapulmonary shunting due to arteriovenous malformation, and diaphragm paralysis due to neuromuscular disease.

In our patient, positional blood gas measurements demonstrated a significant drop in arterial oxygen saturation from the supine to the upright position, raising our suspicion of shunting. It helped narrow the differential diagnosis and guided our selection of additional diagnostic tests.

The initial chest radiograph in our patient was normal. TTE did not reveal shunting and showed a normal right ventricular systolic pressure. TTE with agitated saline also failed to reveal shunting. Because of suboptimal positioning and image quality, TTE may miss the shunting physiology, and that is why we proceeded to positional TEE, which can better evaluate the hemodynamic effects of positional changes on patent foramen ovale and shunting. 

MORE ABOUT PATENT FORAMEN OVALE

The prevalence of patent foramen ovale is estimated at 27% in the general population, but it is usually not symptomatic. It can be associated with atrial septal aneurysm and Chiari network malformations. When associated with atrial septal aneurysm, it carries a higher risk of stroke.⁵

Our patient had a large atrial septal aneurysm with a septal excursion of 2 cm as well as a history of thromboembolic stroke, which was likely associated with the patent foramen ovale and the atrial septal aneurysm.

Atrial septal aneurysm is rare, with a prevalence of 1% at autopsy and 1.9% by TTE. It is defined by a septal excursion of at least 10 mm and a base diameter of at least 15 mm and is more frequently detected on TEE than on TTE.⁶

Studies have shown that contrast and color Doppler TEE are superior to TTE for detecting patent foramen ovale.⁷ Tilt-table TEE  with contrast enhancement has also been used to better demonstrate the morphology of the interatrial septum and the degree of shunting due to the separation between the septum primum and septum secundum causing the patent foramen ovale.⁸ Contrast-enhanced transcranial Doppler has also been shown comparable to contrast TEE to detect interatrial shunting. However, TEE provides additional anatomic information.⁹

In our patient, atrial septal aneurysm and patent foramen ovale were exaggerated by upright positioning, which opened the aneurysm and increased the shunting through the patent foramen ovale.

The treatment of choice in symptomatic patients with platypnea-orthodeoxia syndrome is directed at the underlying cause, in this case closure of the foramen ovale. This treatment has been shown to be safe and effective in these patients,¹⁰ but caution should be used when considering foramen ovale closure in patients with pulmonary hypertension.¹¹

In patients with irreversible or severe pulmonary hypertension, closure of the patent foramen ovale can exacerbate right heart dysfunction and lead to right heart failure. There are situations when closure of a patent foramen ovale can be considered in pulmonary hypertension; however, each decision is individualized, and caution must be used. A detailed discussion is beyond the scope of this paper.

A thorough history and physical examination are important in differentiating the various causes of hypoxemia. Appropriate diagnostic testing is needed along with prompt treatment of the underlying cause of platyp­nea-orthodeoxia syndrome.

An 85-year-old woman was brought to our intensive care unit because of worsening hypoxemia over the past day. About 3 weeks earlier she had been diagnosed with acute bilateral pulmonary emboli in the distal branches of the left and right lower lobes and right middle lobe, for which she was receiving anticoagulation therapy.

At presentation she had generalized fatigue and dyspnea at rest that was worse with exertion, but she denied having fever, chest pain, or cough. Her medical history included hypertension, hyperlipidemia, hypothyroidism, stage 1 breast cancer in remission, thromboembolic stroke, and myasthenia gravis. Before her hospital admission, she had been taking rosuvastatin, metoprolol tartrate, pyridostigmine, prednisone, furosemide, levothyroxine, and rivaroxaban. She did not smoke, she was retired, and she had not traveled recently.

Her blood pressure was 135/66 mm Hg, pulse 73 beats per minute, respiratory rate 16, temperature 35.4ºC (95.7ºF), and oxygen saturation 88% while receiving oxygen at 6 L/min via nasal cannula. Physical examination revealed mild edema in the lower extremities and basilar decreased breath sounds. She had no finger clubbing or cyanosis and was not using accessory muscles to breathe. Of note, her oxygen saturation remained more than 93% when she was recumbent but sharply dropped to less than 85% when she was upright.

Laboratory values

Results of initial laboratory testing were as follows:

• Sodium 138 mmol/L (reference range 132–148)
• Potassium 4.2 mmol/L (3.5–5.0)
• Chloride 99 mmol/L (98–111)
• Bicarbonate 29 mmol/L (23–32)
• Creatinine 0.52 mg/dL (0.7–1.4).
• White blood cell count 11.06 × 10⁹/L (3.7–11.0)
• Hemoglobin 12.6 g/dL (12–16)
• Platelet count 211 × 10⁹/L (150–400).
• International normalized ratio 1.4.

Electrocardiography and imaging studies

Standard 12-lead electrocardiography  showed normal sinus rhythm with left axis deviation and left ventricular hypertrophy.

Chest radiography showed bilateral interstitial opacities and small pleural effusions.

Computed tomography (CT) of the chest with contrast, compared with a CT scan done 20 days earlier, showed that the pulmonary emboli had resolved.

Arterial blood gases

In view of her positional hypoxemia, blood for arterial blood gas measurements was drawn in the supine and upright positions.

Supine, with a fraction of inspired oxygen (Fio₂) via high-flow nasal cannula of 45%, her values were:

• pH 7.45 (reference range 7.35–7.45)
• Pco₂ 34 mm Hg (36–46)
• Po₂ 81 mm Hg (85–95)
• Bicarbonate 23 mmol/L (22–26).

Upright, her hypoxemia was significantly worse:

• pH 7.46
• Pco₂ 33 mm Hg
• Po₂ 57 mm Hg
• Bicarbonate 23 mmol/L.

The methemoglobin level was normal on both measurements.

During her stay in the intensive care unit, she required up to 100% Fio₂ because of persistent hypoxemia.

CAUSES OF HYPOXEMIA

1. So far, the patient’s laboratory tests and imaging studies point to which of the following as the most likely cause of her severe hypoxemia?

• Ventilation-perfusion (V/Q) mismatch
• Diffusion abnormality
• Hypoventilation
• Shunting
• None of the above

The arterial blood gas measurements suggested the possibility of shunting as the cause, although further imaging would be needed to confirm that.

V/Q mismatch can occur in respiratory failure due to pulmonary embolism, pulmonary edema, or shunting. If ventilation is preserved but perfusion is impaired, the V/Q ratio approaches infinity (dead-space ventilation), a situation that can be seen in pulmonary embolism. If perfusion is preserved and ventilation is impaired, the V/Q ratio approaches zero, which is consistent with a physiologic shunt.

Hypoxemia may improve in less severe forms of V/Q mismatch. In our patient, the repeat CT with contrast showed that her pulmonary embolism had resolved, so this is probably not the cause of her severe hypoxemia.

Diffusion abnormalities are due to defects in the lung parenchyma, such as in chronic obstructive pulmonary disease, interstitial lung disease, and lung fibrosis.

Hypoxemia from diffusion defects is usually aggravated by a precipitating factor that increases oxygen demand, and it usually improves with oxygen supplementation. This is unlikely in our patient, as she did not have a history of chronic interstitial lung disease and CT showed no evidence of severe lung parenchymal disease.

Hypoventilation is usually due to drugs that cause respiratory depression, to stroke, or to neuromuscular diseases such as myasthenia gravis that can cause respiratory muscle weakness. It results in elevation of Pco₂ and, if not corrected, respiratory acidosis.

Our patient had a diagnosis of myasthenia gravis, though hypoventilation is unlikely in her case because she had a normal respiratory rate and low Pco₂ values.

Shunting can be physiologic or anatomic and can occur in the heart or the lungs. In physiologic shunting, severe V/Q mismatch can occur when ventilation is affected, as in severe pulmonary edema and pneumonia. In anatomic shunting, a defect such as an atrial septal defect or a pulmonary arteriovenous malformation allows blood to bypass areas of ventilation from the venous to the arterial circulation, preventing it from being oxygenated. In true anatomic shunting, supplemental oxygen with 100% Fio₂ has little effect, whereas in V/Q mismatch it can raise the arterial oxygen saturation.

Our patient’s radiograph did not suggest severe pneumonia or pulmonary edema, which makes these unlikely causes of her hypoxemia. At this point, because of her positional hypoxemia, further evaluation with contrast-enhanced echocardiography was needed to evaluate for anatomic shunting in the heart or lungs.

FURTHER TESTING

Transthoracic echocardiography (TTE) with agitated saline with a Valsalva maneuver was performed. Normally, no bubbles are seen in  the left-sided chambers after intravenous injection of agitated saline contrast, whereas bubbles on the left side suggest an intracardiac or intrapulmonary shunt. In our patient, this test was negative, and her right ventricular systolic pressure was normal.

2. What further testing should be considered to evaluate our patient’s hypoxemia?

• High-resolution chest CT
• Transesophageal echocardiography (TEE)
• Pulmonary function testing
• Repeated arterial blood gas measurement
• Edrophonium testing

Repeat imaging with high-resolution CT would likely not provide additional information and would expose the patient to additional radiation without adding much clinical benefit.

TEE could help further evaluate the intracardiac anatomy and look for shunting, which may be missed on TTE because of suboptimal positioning or image quality.

Pulmonary function testing is useful in establishing the baseline function and impairment in respiratory volumes. If an acute myasthenic crisis is suspected, measuring the negative inspiratory force and the forced vital capacity can be useful in monitoring worsening respiratory muscle weakness and assessing the need for mechanical ventilation.

In our patient, it is unlikely that pulmonary function testing would help, since her acute respiratory failure was probably not caused by neuromuscular weakness.

Repeated arterial blood gas measurement would likely only confirm that the patient still has positional hypoxemia but would not help sort through the differential diagnosis.

Edrophonium testing is useful in diagnosing myasthenia gravis and differentiating it from other neuromuscular diseases, such as Lambert-Eaton syndrome. Edrophonium, a reversible acetylcholinesterase inhibitor, prevents degradation of acetylcholine and prolongs its effect at the synaptic cleft, thus improving muscle weakness.

Our patient has already been diagnosed with myasthenia gravis, so this test is not likely to uncover the cause of her hypoxemia.

Because we still strongly suspected a shunt, TEE was performed with intravenous injection of agitated saline. TEE with the patient upright revealed intracardiac right-to-left shunting through a patent foramen ovale. The midesophageal view after saline injection showed a large interatrial septal aneurysm with total excursion of 2 cm, and right-to-left shunting within the first beat, consistent with an intracardiac shunt (Figure 1). Color Doppler imaging (Figure 2) demonstrated turbulent flow through the patent foramen ovale, consistent with right-to-left shunting, and also showed the patent foramen ovale in a closed position (Figure 3).

3. Which is now most likely the cause of our patient’s hypoxemia?

• Chronic thromboembolic pulmonary hypertension
• Myasthenic crisis
• Platypnea-orthodeoxia syndrome due to the patent foramen ovale
• Methemoglobinemia

Chronic thromboembolic pulmonary hypertension is usually a long-term result of untreated or inadequately treated thromboembolic disease (eg, pulmonary emboli), which causes vascular remodeling and pulmonary arteriopathy, which in turn leads to increased pulmonary vascular resistance and pulmonary hypertension.

This is unlikely the cause of our patient’s acute hypoxemia, as her symptoms did not suggest it. Moreover, an elevated right ventricular systolic pressure on TTE would suggest pulmonary hypertension, but TTE did not show this, and repeat chest CT indicated that her pulmonary embolism had been adequately treated and had resolved. A V/Q scan and right heart catheterization would help rule out chronic thromboembolic pulmonary hypertension, although these were not done in our patient.

Myasthenic crisis is the progressive fatiguing and paralysis of respiratory muscles ultimately requiring mechanical ventilation to sustain life. It is often brought on by infection or drug therapy.

Our patient did not require intubation and she had no signs or symptoms of myasthenic crisis such as ptosis, dysphagia, or dysarthria. She had a negative inspiratory force of −21 cm H₂O, and pulmonary function testing 4 days before her hospital admission had shown a forced vital capacity of 1.84 L, making myasthenic crisis an unlikely cause of her respiratory failure.

Platypnea-orthodeoxia syndrome is a syndrome of dyspnea (platypnea) and hypoxemia (orthodeoxia) that is induced by sitting upright or standing and resolves when lying down. It is a result of right-to-left intracardiac or intrapulmonary shunting in the presence of an anatomic defect and a functional element causing redirection of shunt flow through the anatomic defect in an upright position.¹ It is associated with specific cardiac, pulmonary, and hepatic diseases, such as atrial septal defect, pulmonary arteriovenous malformation, and hepatopulmonary syndrome.² It can occur even if right-sided chamber pressures are normal, and several mechanisms of the underlying pathophysiology have been described.³

Platypnea-orthodeoxia syndrome can be triggered by an event that causes a spontaneous transient elevation of right atrial pressure and pulmonary hypertension, such as our patient’s acute pulmonary embolism. Increased right-to-left shunting occurs in an upright position, causing preferential redirection of flow from the inferior vena cava through the interatrial septum and the patent foramen ovale.⁴

Our patient was elderly and, like one in every four people in the world, she had had a patent foramen ovale since the day she was born. Never causing a problem, it had remained undiagnosed until complicated by platypnea-orthodeoxia syndrome after her recent pulmonary embolism.

Methemoglobinemia. Methemoglobin has a lower affinity for oxygen than normal hemoglobin. Elevations usually occur with medications such as anesthetics and nitrates and can be diagnosed through an elevated level on arterial blood gas testing.

Our patient did not have elevated methemoglobin on her blood gas measurements on admission; therefore, this is unlikely to be the diagnosis.

CASE CONCLUDED

Percutaneous closure of the patent foramen ovale with a 30-mm Amplatzer Cribriform Occluder brought significant improvement in our patient’s functional status and arterial oxygenation saturation, and 2 weeks later at follow-up she no longer needed  supplemental oxygen. TEE 6 months later showed an intact closure device and no interatrial shunting.

WHEN HYPOXEMIA DOES NOT RESPOND TO OXYGEN

In the intensive care unit, time is critical, and when hypoxia is refractory to high Fio₂, shunting should be considered.

In the acute-care setting, platypnea-orthodeoxia syndrome can be identified quickly by pulse oximetry and serial blood gas measurements in the upright and supine positions. A drop in arterial oxygenation in the upright position vs the supine position helps confirm the diagnosis.

Other conditions in the differential diagnosis of this syndrome include recurrent pulmonary embolism, acute respiratory distress syndrome, interstitial pulmonary fibrosis, intrapulmonary shunting due to arteriovenous malformation, and diaphragm paralysis due to neuromuscular disease.

In our patient, positional blood gas measurements demonstrated a significant drop in arterial oxygen saturation from the supine to the upright position, raising our suspicion of shunting. It helped narrow the differential diagnosis and guided our selection of additional diagnostic tests.

The initial chest radiograph in our patient was normal. TTE did not reveal shunting and showed a normal right ventricular systolic pressure. TTE with agitated saline also failed to reveal shunting. Because of suboptimal positioning and image quality, TTE may miss the shunting physiology, and that is why we proceeded to positional TEE, which can better evaluate the hemodynamic effects of positional changes on patent foramen ovale and shunting. 

MORE ABOUT PATENT FORAMEN OVALE

The prevalence of patent foramen ovale is estimated at 27% in the general population, but it is usually not symptomatic. It can be associated with atrial septal aneurysm and Chiari network malformations. When associated with atrial septal aneurysm, it carries a higher risk of stroke.⁵

Our patient had a large atrial septal aneurysm with a septal excursion of 2 cm as well as a history of thromboembolic stroke, which was likely associated with the patent foramen ovale and the atrial septal aneurysm.

Atrial septal aneurysm is rare, with a prevalence of 1% at autopsy and 1.9% by TTE. It is defined by a septal excursion of at least 10 mm and a base diameter of at least 15 mm and is more frequently detected on TEE than on TTE.⁶

Studies have shown that contrast and color Doppler TEE are superior to TTE for detecting patent foramen ovale.⁷ Tilt-table TEE  with contrast enhancement has also been used to better demonstrate the morphology of the interatrial septum and the degree of shunting due to the separation between the septum primum and septum secundum causing the patent foramen ovale.⁸ Contrast-enhanced transcranial Doppler has also been shown comparable to contrast TEE to detect interatrial shunting. However, TEE provides additional anatomic information.⁹

In our patient, atrial septal aneurysm and patent foramen ovale were exaggerated by upright positioning, which opened the aneurysm and increased the shunting through the patent foramen ovale.

The treatment of choice in symptomatic patients with platypnea-orthodeoxia syndrome is directed at the underlying cause, in this case closure of the foramen ovale. This treatment has been shown to be safe and effective in these patients,¹⁰ but caution should be used when considering foramen ovale closure in patients with pulmonary hypertension.¹¹

In patients with irreversible or severe pulmonary hypertension, closure of the patent foramen ovale can exacerbate right heart dysfunction and lead to right heart failure. There are situations when closure of a patent foramen ovale can be considered in pulmonary hypertension; however, each decision is individualized, and caution must be used. A detailed discussion is beyond the scope of this paper.

A thorough history and physical examination are important in differentiating the various causes of hypoxemia. Appropriate diagnostic testing is needed along with prompt treatment of the underlying cause of platyp­nea-orthodeoxia syndrome.

An 89-year-old man presented complaining of a tender, painful lump in the right lower eyelid that spontaneously appeared 3 days previously. There was no discharge, bleeding, or reduced vision. He had a history of hypertension and macular degeneration. There was no history of a pre-existing eyelid lesion, ocular malignancy, rosacea, or seborrheic dermatitis. Examination of the right lower lid revealed a roundish raised abscess with surrounding erythema (Figure 1). The raised area was tender on palpation; there was no discharge. The palpebral conjunctiva was normal. A diagnosis of a hordeolum was made, and conservative treatment was prescribed, ie, warm compresses and massage for 10 minutes four times a day. The lesion improved gradually and resolved over 3 weeks.

A hordeolum is an acute abscess within an eyelid gland, usually staphylococcal in origin. When it involves a meibomian gland it is termed an internal hordeolum, and when it involves the gland of Zeis or Moll it is termed an external hordeolum (Figure 2).¹ Hordeola may be associated with diabetes, blepharitis, seborrheic dermatitis, rosacea, and high levels of serum lipids. Treatment is with warm compresses and massage. A hordeolum with preseptal cellulitis, signs of bacteremia, or tender preauricular lymph nodes requires systemic antibiotics (eg, flucloxacillin 250–500 mg four times a day for 1 week).

Preseptal cellulitis is an infection of the subcutaneous tissues anterior to the orbital septum. The orbital septum is a sheet of fibrous tissue that originates in the orbital periosteum and inserts in the palpebral tissues along the tarsal plates of the eyelid. The orbital septum provides a barrier against the spread of periorbital infection into the orbit (orbital cellulitis). The causes of preseptal cellulitis include skin trauma (eg, lacerations, insect bites), spread from local infections (eg, hordeolum, dacryocystitis), or systemic infections (eg, upper respiratory tract, middle ear). Clinical features include malaise, fever, and painful eyelid with periorbital edema. Any sign of proptosis, chemosis, painful restricted eye movements, diplopia, lagophthalmos, or optic nerve dysfunction warrants further investigation. Chronic or large hordeola may require incision and curettage.

A recent Cochrane review concluded that there was no evidence of the effectiveness of nonsurgical interventions (including hot or warm compresses, lid scrubs, antibiotics, and steroids) for hordeolum, and controlled clinical trials would be useful.²

Chalazion and hordeolum are similar in appearance and often confused (Table 1). A chalazion is a chronic lipogranuloma due to leakage of sebum from an obstructed meibomian gland. It may develop from an internal hordeolum. Small chalazia usually resolve with time without any intervention, and hot compresses can be effective at encouraging drainage. Persistent lesions may be surgically removed by incision and curettage. Recurrence warrants biopsy and histologic study to rule out sebaceous gland carcinoma.³

An 89-year-old man presented complaining of a tender, painful lump in the right lower eyelid that spontaneously appeared 3 days previously. There was no discharge, bleeding, or reduced vision. He had a history of hypertension and macular degeneration. There was no history of a pre-existing eyelid lesion, ocular malignancy, rosacea, or seborrheic dermatitis. Examination of the right lower lid revealed a roundish raised abscess with surrounding erythema (Figure 1). The raised area was tender on palpation; there was no discharge. The palpebral conjunctiva was normal. A diagnosis of a hordeolum was made, and conservative treatment was prescribed, ie, warm compresses and massage for 10 minutes four times a day. The lesion improved gradually and resolved over 3 weeks.

A hordeolum is an acute abscess within an eyelid gland, usually staphylococcal in origin. When it involves a meibomian gland it is termed an internal hordeolum, and when it involves the gland of Zeis or Moll it is termed an external hordeolum (Figure 2).¹ Hordeola may be associated with diabetes, blepharitis, seborrheic dermatitis, rosacea, and high levels of serum lipids. Treatment is with warm compresses and massage. A hordeolum with preseptal cellulitis, signs of bacteremia, or tender preauricular lymph nodes requires systemic antibiotics (eg, flucloxacillin 250–500 mg four times a day for 1 week).

Preseptal cellulitis is an infection of the subcutaneous tissues anterior to the orbital septum. The orbital septum is a sheet of fibrous tissue that originates in the orbital periosteum and inserts in the palpebral tissues along the tarsal plates of the eyelid. The orbital septum provides a barrier against the spread of periorbital infection into the orbit (orbital cellulitis). The causes of preseptal cellulitis include skin trauma (eg, lacerations, insect bites), spread from local infections (eg, hordeolum, dacryocystitis), or systemic infections (eg, upper respiratory tract, middle ear). Clinical features include malaise, fever, and painful eyelid with periorbital edema. Any sign of proptosis, chemosis, painful restricted eye movements, diplopia, lagophthalmos, or optic nerve dysfunction warrants further investigation. Chronic or large hordeola may require incision and curettage.

A recent Cochrane review concluded that there was no evidence of the effectiveness of nonsurgical interventions (including hot or warm compresses, lid scrubs, antibiotics, and steroids) for hordeolum, and controlled clinical trials would be useful.²

Chalazion and hordeolum are similar in appearance and often confused (Table 1). A chalazion is a chronic lipogranuloma due to leakage of sebum from an obstructed meibomian gland. It may develop from an internal hordeolum. Small chalazia usually resolve with time without any intervention, and hot compresses can be effective at encouraging drainage. Persistent lesions may be surgically removed by incision and curettage. Recurrence warrants biopsy and histologic study to rule out sebaceous gland carcinoma.³

An 89-year-old man presented complaining of a tender, painful lump in the right lower eyelid that spontaneously appeared 3 days previously. There was no discharge, bleeding, or reduced vision. He had a history of hypertension and macular degeneration. There was no history of a pre-existing eyelid lesion, ocular malignancy, rosacea, or seborrheic dermatitis. Examination of the right lower lid revealed a roundish raised abscess with surrounding erythema (Figure 1). The raised area was tender on palpation; there was no discharge. The palpebral conjunctiva was normal. A diagnosis of a hordeolum was made, and conservative treatment was prescribed, ie, warm compresses and massage for 10 minutes four times a day. The lesion improved gradually and resolved over 3 weeks.

A hordeolum is an acute abscess within an eyelid gland, usually staphylococcal in origin. When it involves a meibomian gland it is termed an internal hordeolum, and when it involves the gland of Zeis or Moll it is termed an external hordeolum (Figure 2).¹ Hordeola may be associated with diabetes, blepharitis, seborrheic dermatitis, rosacea, and high levels of serum lipids. Treatment is with warm compresses and massage. A hordeolum with preseptal cellulitis, signs of bacteremia, or tender preauricular lymph nodes requires systemic antibiotics (eg, flucloxacillin 250–500 mg four times a day for 1 week).

Preseptal cellulitis is an infection of the subcutaneous tissues anterior to the orbital septum. The orbital septum is a sheet of fibrous tissue that originates in the orbital periosteum and inserts in the palpebral tissues along the tarsal plates of the eyelid. The orbital septum provides a barrier against the spread of periorbital infection into the orbit (orbital cellulitis). The causes of preseptal cellulitis include skin trauma (eg, lacerations, insect bites), spread from local infections (eg, hordeolum, dacryocystitis), or systemic infections (eg, upper respiratory tract, middle ear). Clinical features include malaise, fever, and painful eyelid with periorbital edema. Any sign of proptosis, chemosis, painful restricted eye movements, diplopia, lagophthalmos, or optic nerve dysfunction warrants further investigation. Chronic or large hordeola may require incision and curettage.

A recent Cochrane review concluded that there was no evidence of the effectiveness of nonsurgical interventions (including hot or warm compresses, lid scrubs, antibiotics, and steroids) for hordeolum, and controlled clinical trials would be useful.²

Chalazion and hordeolum are similar in appearance and often confused (Table 1). A chalazion is a chronic lipogranuloma due to leakage of sebum from an obstructed meibomian gland. It may develop from an internal hordeolum. Small chalazia usually resolve with time without any intervention, and hot compresses can be effective at encouraging drainage. Persistent lesions may be surgically removed by incision and curettage. Recurrence warrants biopsy and histologic study to rule out sebaceous gland carcinoma.³

We have tools to predict bleeding risk, but their predictive value is modest, and the estimated risk of bleeding is often outweighed by the benefits of anticoagulant therapy.

Anticoagulant therapy is commonly prescribed for conditions that disproportionately affect the elderly, including atrial fibrillation, venous thromboembolism, and valvular heart disease. Though anticoagulants are highly effective in preventing clots, they also significantly increase the risk of bleeding. Since older age is a risk factor for bleeding as well as thrombosis, it is essential to weigh the risks and benefits of anticoagulants for each patient.

WHAT KINDS OF BLEEDING DEVELOP IN PATIENTS ON ANTICOAGULANTS?

Patients taking anticoagulants have roughly double the risk of bleeding compared with patients not on anticoagulants.¹ Bleeding rates tend to be slightly higher in patients taking anticoagulants for venous thromboembolism than in those taking them for atrial fibrillation. The average yearly risk of a “major” anticoagulant-associated bleeding event (eg, requiring transfusion or intervention or occurring in a critical anatomic site) is about 2% to 3%, with most of the bleeding being gastrointestinal.²

Intracranial hemorrhage is by far the most deadly complication of anticoagulant therapy: it causes 90% of deaths and disability from warfarin-associated hemorrhage and is associated with a death rate over 50%; however, it is much less common than gastrointestinal bleeding.³ Anticoagulant therapy increases the risk of intracranial hemorrhage by only 0.2% per year.¹

RISK-PREDICTION TOOLS HAVE LIMITATIONS

Not all patients have the same risk of bleeding when taking anticoagulants. Many factors in addition to advanced age have been associated with increased bleeding risk, including coexisting medical conditions (such as malignancy, prior stroke or bleeding event, and renal insufficiency), medications (particularly aspirin, nonsteroidal anti-inflammatory drugs, and other antiplatelet drugs), and the timing and intensity of anticoagulation therapy.⁴

Scoring tools have been developed to identify patients at higher risk of bleeding (Table 1).^4–9 The various schemes incorporate many of the same variables, such as older age, renal impairment, and history of bleeding, but some include additional risk factors while others are more parsimonious. They also differ in how individual risk factors are weighted to generate a final risk score.

In terms of predictive ability, none of the available risk schemes appears to be vastly superior, and their ability to predict hemorrhage is modest at best. There is also no universal or well-established threshold at which the risk of bleeding is so high that one would not consider anticoagulants. In fact, a “high-risk” patient may have an aggregate bleeding rate of only 4% to 6% per year. Using risk schemes such as ATRIA,⁵ HEMORR₂HAGES,⁶ and HAS-BLED⁷ may be more useful because they provide an estimate of bleeding risk for each point on the scale.

Moreover, the current tools to predict bleeding risk have several other limitations. They were developed in patients already taking anticoagulants and so probably underestimate the actual risk of hemorrhage, as people who could not take anticoagulants were excluded, most likely because they were at high risk of bleeding. Therefore, bleeding risk tools probably apply best to a patient for whom anticoagulation can be considered.

Some clinical variables are necessarily broad. For example, “prior bleeding” is a risk factor included in several risk scores, but does not distinguish between massive variceal bleeding and minor hemorrhoidal bleeding.

Risk scores do not effectively predict intracranial hemorrhage.

Finally, these risk tools were developed in patients taking vitamin K antagonists, and it is not yet established that they can effectively predict hemorrhage related to other, newer anticoagulants.

WHEN DOES BLEEDING RISK OUTWEIGH ANTICOAGULATION BENEFIT?

For patients with atrial fibrillation, the net clinical benefit of anticoagulation (strokes prevented minus bleeding events induced) increases as the risk of stroke rises. Updated guidelines for managing atrial fibrillation now recommend anticoagulation for most patients.¹⁰

For most older patients with atrial fibrillation, the decision to anticoagulate may not change even if a bleeding risk tool indicates a high bleeding risk.¹¹ For example, a patient with a history of ischemic stroke will generally derive more benefit than harm from anticoagulants. The primary exception is in patients with prior lobar intracranial hemorrhage, because of the high risk of rebleeding and the worse outcomes associated with intracranial hemorrhage.¹² As a general rule, most patients with atrial fibrillation and an additional risk factor for stroke should be considered for anticoagulant therapy unless they have a history of lobar intracranial hemorrhage.

Anticoagulation may be deferred if the patient is at the lower end of the stroke risk spectrum and if the bleeding risk is calculated to be high. However, as noted before, current bleeding risk tools probably do not capture the experiences of patients at the extremes of high bleeding risk, so clinical judgment continues to be important. In addition, forgoing anticoagulation could be reasonable even in patients at high risk for recurrent stroke if their life expectancy is limited, if anticoagulation is unacceptably burdensome, or if it is not within their goals and preferences.

WHAT ABOUT FALL RISK?

Fall risk commonly deters clinicians from prescribing anticoagulants because of the fear of causing intracranial hemorrhage. In particular, falls increase the risk for subdural hematoma, which has a death rate comparable to that of ischemic stroke.¹³

Studies have had difficulty quantifying the exact risk associated with falls because these patients are less likely to be prescribed anticoagulants. One decision analysis estimated that a person would have to fall about 300 times per year before the risk of intracranial hemorrhage outweighed the benefits from stroke reduction.¹⁴ Studies have found that patients at high risk of falls have a higher risk of intracranial hemorrhage, but that this risk is counterbalanced by an even greater risk of ischemic stroke.¹⁵

Therefore, if the baseline risk of ischemic stroke is high, anticoagulation is still favored.

WHEN SHOULD I USE A BLEEDING RISK TOOL?

Despite their limitations, bleeding risk tools are useful in clinical practice when estimates of bleeding risk affect clinical behavior. They are most helpful for patients at the lower end of the stroke or thromboembolism risk spectrum, where the decision to anticoagulate is strongly influenced by bleeding risk. Risk tools may also be helpful when counseling patients about their bleeding risk off and on anticoagulants.

Finally, recognizing that a patient is at high bleeding risk may lead the clinician to consider closer monitoring of anticoagulants or to implement strategies to reduce the risk.

We have tools to predict bleeding risk, but their predictive value is modest, and the estimated risk of bleeding is often outweighed by the benefits of anticoagulant therapy.

Anticoagulant therapy is commonly prescribed for conditions that disproportionately affect the elderly, including atrial fibrillation, venous thromboembolism, and valvular heart disease. Though anticoagulants are highly effective in preventing clots, they also significantly increase the risk of bleeding. Since older age is a risk factor for bleeding as well as thrombosis, it is essential to weigh the risks and benefits of anticoagulants for each patient.

WHAT KINDS OF BLEEDING DEVELOP IN PATIENTS ON ANTICOAGULANTS?

Patients taking anticoagulants have roughly double the risk of bleeding compared with patients not on anticoagulants.¹ Bleeding rates tend to be slightly higher in patients taking anticoagulants for venous thromboembolism than in those taking them for atrial fibrillation. The average yearly risk of a “major” anticoagulant-associated bleeding event (eg, requiring transfusion or intervention or occurring in a critical anatomic site) is about 2% to 3%, with most of the bleeding being gastrointestinal.²

Intracranial hemorrhage is by far the most deadly complication of anticoagulant therapy: it causes 90% of deaths and disability from warfarin-associated hemorrhage and is associated with a death rate over 50%; however, it is much less common than gastrointestinal bleeding.³ Anticoagulant therapy increases the risk of intracranial hemorrhage by only 0.2% per year.¹

RISK-PREDICTION TOOLS HAVE LIMITATIONS

Not all patients have the same risk of bleeding when taking anticoagulants. Many factors in addition to advanced age have been associated with increased bleeding risk, including coexisting medical conditions (such as malignancy, prior stroke or bleeding event, and renal insufficiency), medications (particularly aspirin, nonsteroidal anti-inflammatory drugs, and other antiplatelet drugs), and the timing and intensity of anticoagulation therapy.⁴

Scoring tools have been developed to identify patients at higher risk of bleeding (Table 1).^4–9 The various schemes incorporate many of the same variables, such as older age, renal impairment, and history of bleeding, but some include additional risk factors while others are more parsimonious. They also differ in how individual risk factors are weighted to generate a final risk score.

In terms of predictive ability, none of the available risk schemes appears to be vastly superior, and their ability to predict hemorrhage is modest at best. There is also no universal or well-established threshold at which the risk of bleeding is so high that one would not consider anticoagulants. In fact, a “high-risk” patient may have an aggregate bleeding rate of only 4% to 6% per year. Using risk schemes such as ATRIA,⁵ HEMORR₂HAGES,⁶ and HAS-BLED⁷ may be more useful because they provide an estimate of bleeding risk for each point on the scale.

Moreover, the current tools to predict bleeding risk have several other limitations. They were developed in patients already taking anticoagulants and so probably underestimate the actual risk of hemorrhage, as people who could not take anticoagulants were excluded, most likely because they were at high risk of bleeding. Therefore, bleeding risk tools probably apply best to a patient for whom anticoagulation can be considered.

Some clinical variables are necessarily broad. For example, “prior bleeding” is a risk factor included in several risk scores, but does not distinguish between massive variceal bleeding and minor hemorrhoidal bleeding.

Risk scores do not effectively predict intracranial hemorrhage.

Finally, these risk tools were developed in patients taking vitamin K antagonists, and it is not yet established that they can effectively predict hemorrhage related to other, newer anticoagulants.

WHEN DOES BLEEDING RISK OUTWEIGH ANTICOAGULATION BENEFIT?

For patients with atrial fibrillation, the net clinical benefit of anticoagulation (strokes prevented minus bleeding events induced) increases as the risk of stroke rises. Updated guidelines for managing atrial fibrillation now recommend anticoagulation for most patients.¹⁰

For most older patients with atrial fibrillation, the decision to anticoagulate may not change even if a bleeding risk tool indicates a high bleeding risk.¹¹ For example, a patient with a history of ischemic stroke will generally derive more benefit than harm from anticoagulants. The primary exception is in patients with prior lobar intracranial hemorrhage, because of the high risk of rebleeding and the worse outcomes associated with intracranial hemorrhage.¹² As a general rule, most patients with atrial fibrillation and an additional risk factor for stroke should be considered for anticoagulant therapy unless they have a history of lobar intracranial hemorrhage.

Anticoagulation may be deferred if the patient is at the lower end of the stroke risk spectrum and if the bleeding risk is calculated to be high. However, as noted before, current bleeding risk tools probably do not capture the experiences of patients at the extremes of high bleeding risk, so clinical judgment continues to be important. In addition, forgoing anticoagulation could be reasonable even in patients at high risk for recurrent stroke if their life expectancy is limited, if anticoagulation is unacceptably burdensome, or if it is not within their goals and preferences.

WHAT ABOUT FALL RISK?

Fall risk commonly deters clinicians from prescribing anticoagulants because of the fear of causing intracranial hemorrhage. In particular, falls increase the risk for subdural hematoma, which has a death rate comparable to that of ischemic stroke.¹³

Studies have had difficulty quantifying the exact risk associated with falls because these patients are less likely to be prescribed anticoagulants. One decision analysis estimated that a person would have to fall about 300 times per year before the risk of intracranial hemorrhage outweighed the benefits from stroke reduction.¹⁴ Studies have found that patients at high risk of falls have a higher risk of intracranial hemorrhage, but that this risk is counterbalanced by an even greater risk of ischemic stroke.¹⁵

Therefore, if the baseline risk of ischemic stroke is high, anticoagulation is still favored.

WHEN SHOULD I USE A BLEEDING RISK TOOL?

Despite their limitations, bleeding risk tools are useful in clinical practice when estimates of bleeding risk affect clinical behavior. They are most helpful for patients at the lower end of the stroke or thromboembolism risk spectrum, where the decision to anticoagulate is strongly influenced by bleeding risk. Risk tools may also be helpful when counseling patients about their bleeding risk off and on anticoagulants.

Finally, recognizing that a patient is at high bleeding risk may lead the clinician to consider closer monitoring of anticoagulants or to implement strategies to reduce the risk.

We have tools to predict bleeding risk, but their predictive value is modest, and the estimated risk of bleeding is often outweighed by the benefits of anticoagulant therapy.

Anticoagulant therapy is commonly prescribed for conditions that disproportionately affect the elderly, including atrial fibrillation, venous thromboembolism, and valvular heart disease. Though anticoagulants are highly effective in preventing clots, they also significantly increase the risk of bleeding. Since older age is a risk factor for bleeding as well as thrombosis, it is essential to weigh the risks and benefits of anticoagulants for each patient.

WHAT KINDS OF BLEEDING DEVELOP IN PATIENTS ON ANTICOAGULANTS?

Patients taking anticoagulants have roughly double the risk of bleeding compared with patients not on anticoagulants.¹ Bleeding rates tend to be slightly higher in patients taking anticoagulants for venous thromboembolism than in those taking them for atrial fibrillation. The average yearly risk of a “major” anticoagulant-associated bleeding event (eg, requiring transfusion or intervention or occurring in a critical anatomic site) is about 2% to 3%, with most of the bleeding being gastrointestinal.²

Intracranial hemorrhage is by far the most deadly complication of anticoagulant therapy: it causes 90% of deaths and disability from warfarin-associated hemorrhage and is associated with a death rate over 50%; however, it is much less common than gastrointestinal bleeding.³ Anticoagulant therapy increases the risk of intracranial hemorrhage by only 0.2% per year.¹

RISK-PREDICTION TOOLS HAVE LIMITATIONS

Not all patients have the same risk of bleeding when taking anticoagulants. Many factors in addition to advanced age have been associated with increased bleeding risk, including coexisting medical conditions (such as malignancy, prior stroke or bleeding event, and renal insufficiency), medications (particularly aspirin, nonsteroidal anti-inflammatory drugs, and other antiplatelet drugs), and the timing and intensity of anticoagulation therapy.⁴

Scoring tools have been developed to identify patients at higher risk of bleeding (Table 1).^4–9 The various schemes incorporate many of the same variables, such as older age, renal impairment, and history of bleeding, but some include additional risk factors while others are more parsimonious. They also differ in how individual risk factors are weighted to generate a final risk score.

In terms of predictive ability, none of the available risk schemes appears to be vastly superior, and their ability to predict hemorrhage is modest at best. There is also no universal or well-established threshold at which the risk of bleeding is so high that one would not consider anticoagulants. In fact, a “high-risk” patient may have an aggregate bleeding rate of only 4% to 6% per year. Using risk schemes such as ATRIA,⁵ HEMORR₂HAGES,⁶ and HAS-BLED⁷ may be more useful because they provide an estimate of bleeding risk for each point on the scale.

Moreover, the current tools to predict bleeding risk have several other limitations. They were developed in patients already taking anticoagulants and so probably underestimate the actual risk of hemorrhage, as people who could not take anticoagulants were excluded, most likely because they were at high risk of bleeding. Therefore, bleeding risk tools probably apply best to a patient for whom anticoagulation can be considered.

Some clinical variables are necessarily broad. For example, “prior bleeding” is a risk factor included in several risk scores, but does not distinguish between massive variceal bleeding and minor hemorrhoidal bleeding.

Risk scores do not effectively predict intracranial hemorrhage.

Finally, these risk tools were developed in patients taking vitamin K antagonists, and it is not yet established that they can effectively predict hemorrhage related to other, newer anticoagulants.

WHEN DOES BLEEDING RISK OUTWEIGH ANTICOAGULATION BENEFIT?

For patients with atrial fibrillation, the net clinical benefit of anticoagulation (strokes prevented minus bleeding events induced) increases as the risk of stroke rises. Updated guidelines for managing atrial fibrillation now recommend anticoagulation for most patients.¹⁰

For most older patients with atrial fibrillation, the decision to anticoagulate may not change even if a bleeding risk tool indicates a high bleeding risk.¹¹ For example, a patient with a history of ischemic stroke will generally derive more benefit than harm from anticoagulants. The primary exception is in patients with prior lobar intracranial hemorrhage, because of the high risk of rebleeding and the worse outcomes associated with intracranial hemorrhage.¹² As a general rule, most patients with atrial fibrillation and an additional risk factor for stroke should be considered for anticoagulant therapy unless they have a history of lobar intracranial hemorrhage.

Anticoagulation may be deferred if the patient is at the lower end of the stroke risk spectrum and if the bleeding risk is calculated to be high. However, as noted before, current bleeding risk tools probably do not capture the experiences of patients at the extremes of high bleeding risk, so clinical judgment continues to be important. In addition, forgoing anticoagulation could be reasonable even in patients at high risk for recurrent stroke if their life expectancy is limited, if anticoagulation is unacceptably burdensome, or if it is not within their goals and preferences.

WHAT ABOUT FALL RISK?

Fall risk commonly deters clinicians from prescribing anticoagulants because of the fear of causing intracranial hemorrhage. In particular, falls increase the risk for subdural hematoma, which has a death rate comparable to that of ischemic stroke.¹³

Studies have had difficulty quantifying the exact risk associated with falls because these patients are less likely to be prescribed anticoagulants. One decision analysis estimated that a person would have to fall about 300 times per year before the risk of intracranial hemorrhage outweighed the benefits from stroke reduction.¹⁴ Studies have found that patients at high risk of falls have a higher risk of intracranial hemorrhage, but that this risk is counterbalanced by an even greater risk of ischemic stroke.¹⁵

Therefore, if the baseline risk of ischemic stroke is high, anticoagulation is still favored.

WHEN SHOULD I USE A BLEEDING RISK TOOL?

Despite their limitations, bleeding risk tools are useful in clinical practice when estimates of bleeding risk affect clinical behavior. They are most helpful for patients at the lower end of the stroke or thromboembolism risk spectrum, where the decision to anticoagulate is strongly influenced by bleeding risk. Risk tools may also be helpful when counseling patients about their bleeding risk off and on anticoagulants.

Finally, recognizing that a patient is at high bleeding risk may lead the clinician to consider closer monitoring of anticoagulants or to implement strategies to reduce the risk.