Why "Why" Matters
http://www.100md.com
《新英格兰医药杂志》
In this Journal feature, information about a real patient is presented in stages (boldface type) to an expert clinician, who responds to the information, sharing his or her reasoning with the reader (regular type). The authors' commentary follows.
A 38-year-old woman presented to the emergency room for evaluation of shortness of breath and jaundice. The previous day, she had attended a wedding, where she felt well. Several hours after the wedding, a headache developed, she had mild dizziness, and she noticed that her urine was dark brown. The day after the wedding, she awoke with mild shortness of breath and yellow discoloration of her eyes and skin.
Jaundice is usually caused by hepatobiliary disorders or hemolysis. Biliary obstruction, acute hepatitis, or brisk intravascular hemolysis could explain the jaundice and dark urine. Wilson's disease — with hepatocellular disease and associated hemolysis — could cause jaundice and dark urine.
Dyspnea and jaundice can be related in several ways. Sepsis can cause hyperbilirubinemia and hyperventilation. Hemolysis can lead to dyspnea from anemia and to indirect hyperbilirubinemia from red-cell destruction. Some infections, such as amebiasis, can affect both the liver and the lung. In addition, in the setting of chronic liver disease with associated jaundice — which would be unlikely in this patient, since her illness was acute — dyspnea can develop as a result of the hepatopulmonary syndrome, portopulmonary hypertension, hepatic hydrothorax, or ascites.
Since the patient's illness occurred in the context of her recent attendance at a wedding, I am interested to know what she ate and drank there.
At the wedding, the patient consumed Chinese dumplings containing salt-cured meat and a glass of wine. No one else who attended the wedding became sick. The patient had no abdominal pain, nausea, vomiting, diarrhea, or fever.
I am uncertain whether the patient's illness can be attributed to events at the wedding. Acute infection with hepatitis A virus can occur with ingestion of contaminated food, but the latency period is weeks, not hours. A fatty meal might lead to cholecystitis or choledocholithiasis, but the absence of abdominal pain argues against these disorders. Enterohemorrhagic Escherichia coli, specifically strain O157:H7, can cause the hemolytic–uremic syndrome, although the latency period is at least 24 hours, and diarrhea and abdominal pain are typical symptoms.
In the absence of gastrointestinal symptoms, I favor the consideration of an acute hemolytic episode. A potential ingestion at the wedding, followed by hemolysis, raises the possibility of a deficiency of glucose-6-phosphate dehydrogenase (G6PD). The only food I am aware of that is associated with episodic hemolysis from G6PD deficiency is the fava bean.
The patient had a history of mild anemia. She had no history of liver disease. She took no medications or herbal supplements, had no drug allergies, and did not smoke or use illicit drugs. She drank alcohol occasionally. She was Vietnamese and had lived in the United States for seven years. She had been adopted and was unsure of her family history. She worked in a wire-manufacturing plant.
The most likely explanation for her anemia would be iron deficiency or, given her ethnic background, thalassemia. However, many of the chronic hemolytic anemias, which result from defects in the red-cell membrane, enzymes, or hemoglobin, still warrant consideration. The family history, unavailable here, is often helpful in diagnosing these disorders. Long-standing hemolytic anemia can sometimes be manifested as cholecystitis or choledocholithiasis caused by pigment gallstones. Although ingestion of a hepatotoxin might explain the patient's presentation, she reports no history of the use of potentially hepatotoxic medications or herbs and no substantial alcohol intake. Her history of having lived in Vietnam supports the consideration of hepatitis B infection as a cause of her illness.
In general, the combination of jaundice and dark urine is more likely to be a result of hepatobiliary disease than of hemolysis. However, chronic anemia, new dyspnea, and the absence of hepatotoxins, known liver disease, and gastrointestinal symptoms lead me to favor the consideration of hemolysis over an acute hepatobiliary disorder.
The patient was afebrile, with a pulse of 98 beats per minute and a blood pressure of 112/64 mm Hg. Her respiratory rate was 24 breaths per minute, and her oxygen saturation 85 percent. Her skin was jaundiced, her sclerae were icteric, and her conjunctivae were pale. An early systolic murmur was present at the right upper sternal border, and the apical impulse was hyperkinetic. Her lungs were clear on auscultation, and there was no abdominal tenderness or hepatosplenomegaly. The remainder of her physical examination was normal.
The pale conjunctivae, tachycardia, systolic murmur, and hyperkinetic apical impulse are all consistent with anemia, and the jaundice — with no evidence of hepatic or abdominal disease — suggests hemolysis as the cause. Splenomegaly may be seen in many forms of hemolysis but is not an invariable feature and can be challenging to detect by physical examination. Therefore, the absence of palpable splenomegaly is relatively uninformative.
The oxygen saturation of 85 percent is unexpected and not explained by hemolysis. Anemia is a disorder involving a decreased overall capacity to carry oxygen, but the remaining red cells should be adequately saturated in the presence of normal alveolar gas exchange. This particular oxygen saturation brings to mind methemoglobinemia, an acquired dyshemoglobinemia caused by a number of drugs or chemicals that oxidize the iron in hemoglobin, rendering it incapable of carrying oxygen. Sulfonamide antibiotics, local anesthetics such as benzocaine, and antimalarial drugs are common precipitants. Measurement of arterial blood gases with co-oximetry would provide information on the partial pressure of oxygen in arterial blood and on the fraction of the total hemoglobin that is methemoglobin. If both hemolysis and methemoglobinemia were present, I would search for an agent that could trigger both events, and I would suspect that the patient has either a G6PD deficiency or an unstable hemoglobin variant (such as hemoglobin K?ln). There is no history of potential culprit drugs, such as dapsone or primaquine, but perhaps there was a component of the food or drink, such as a nitrate or sulfate, that triggered both processes.
If arterial blood analysis reveals hypoxemia without methemoglobinemia, the chest radiograph might suggest other diagnoses. Hilar lymphadenopathy might indicate lymphoma with associated autoimmune hemolysis; an infiltrate could suggest cold autoimmune hemolytic anemia with Mycoplasma pneumoniae infection. A normal chest radiograph would raise the possibility of a pulmonary embolism, in which case the rare disorder of paroxysmal nocturnal hemoglobinuria, characterized by hemolysis, pancytopenia, and venous thrombosis, would be worth considering.
The white-cell count was 13,600 per cubic millimeter, the hemoglobin 8.1 g per deciliter, and the platelet count 194,000 per cubic millimeter. The reticulocyte count was 1.8 percent (45,000 reticulocytes per cubic millimeter of blood [normal range, 25,000 to 75,000]). The total bilirubin level was 9.3 mg per deciliter (10.2 μmol per liter), and the lactate dehydrogenase level 2152 U per milliliter (normal range, 91 to 180). The aspartate aminotransferase and alanine aminotransferase levels were 120 U per liter (normal range, 6 to 30) and 41 units per liter (normal range, 10 to 40), respectively. The serum alkaline phosphatase, albumin, and electrolyte levels were normal. A urine dipstick test was positive for blood; no red cells or casts were seen on microscopical examination. Viral serologic tests for hepatitis A, B, and C viruses were negative. The chest radiograph was normal. An electrocardiogram revealed sinus tachycardia.
The low hemoglobin level and elevated lactate dehydrogenase and bilirubin levels suggest ongoing hemolysis. The results of the urine studies are compatible with hemoglobinuria and indicate rapid and severe red-cell destruction; brisk hemolysis can sometimes cause elevations in aminotransferases. The initial reticulocyte count is inappropriately low, but the bone marrow can be delayed in its response to hemolysis. The normal chest radiograph increases my suspicion of methemoglobinemia.
The cause of the patient's hemolysis remains a puzzle. Given its acute onset, the preexisting anemia, the possibility of methemoglobinemia, and no clear evidence of underlying illness, I favor the hypothesis that some type of oxidant exposure at the wedding triggered a hemolytic episode in this patient, who probably has G6PD deficiency or an unstable hemoglobinopathy. Direct examination of the peripheral blood would be helpful. Disorders that render red cells susceptible to oxidant stress may result in bite cells on the peripheral smear. Heinz bodies (cellular inclusions consisting of damaged aggregated hemoglobin) may be detected with supravital staining.
Despite the administration of supplemental oxygen, the oxygen saturation remained at 85 percent. Arterial blood gas values while the patient was breathing pure oxygen by face mask were as follows: pH 7.45, partial pressure of carbon dioxide 29 mm Hg, and partial pressure of oxygen 432 mm Hg. Co-oximetry demonstrated a methemoglobin level of 8.8 percent (normal range, 0.4 to 1.5).
With the acute onset of hemolysis and methemoglobinemia, I suspect that the patient's red cells were exposed to oxidant stress and that she has either a G6PD deficiency or an unstable-hemoglobin disorder that makes her susceptible to oxidant-induced hemolysis.
How should this patient be treated? Two mechanisms — decreased red-cell mass and ineffective hemoglobin (methemoglobin) — have led to insufficient delivery of oxygen to her tissues. The first-line treatment for acquired methemoglobinemia is the administration of methylene blue, which quickly reduces methemoglobin to hemoglobin. However, G6PD deficiency is a contraindication to treatment with methylene blue because patients with this disorder are unable to metabolize methylene blue, an oxidant that can exacerbate hemolysis. With a hemoglobin level that is already quite low, this patient may not tolerate additional hemolysis.
Toxin-mediated or oxidant-mediated hemolysis is usually managed supportively after the exposure has been terminated. With G6PD deficiency, most acute episodes of hemolysis are self-limited, and only the most severe cases require transfusion. Blood transfusion would accomplish the goal of restoring a normal hemoglobin level and oxygen-carrying capacity, but it involves additional risks. I favor blood transfusion over the use of methylene blue because of the suspicion of oxidant-mediated hemolysis due to G6PD deficiency.
Intravenous methylene blue was administered. Subsequent testing demonstrated worsening anemia, with a hemoglobin value of 6.5 g per deciliter. The patient received 4 units of packed red cells, and her red-cell count was stabilized. The activity of G6PD, as measured in a blood sample obtained before transfusion, was 3.9 U per gram of hemoglobin (normal range, 4.6 to 13.5).
This measured G6PD level may actually understate the patient's degree of G6PD deficiency, since the G6PD level measured during a hemolytic episode reflects only the surviving red-cell population; these surviving red cells evade hemolysis owing to G6PD levels that are higher than average. The gene for G6PD resides on the X chromosome. In a female patient, this raises the interesting possibility that the surviving red-cell population represents cells whose precursors have selectively deactivated their deficient X chromosome.
The patient was seen in the clinic several months later and was doing well. Her hemoglobin measured 12.6 g per deciliter, and a peripheral-blood smear demonstrated only mild reticulocytosis.
Commentary
A fundamental component in the care of any patient is an understanding and appreciation of the pathophysiology underlying his or her disease. Furthermore, it is essential that clinicians continue to evaluate available clinical data — even after a particular diagnosis is suspected — to ensure that there are no discordant findings. As this case illustrates, failure in these two areas can lead to inappropriate management and iatrogenic morbidity. Although the treating physicians correctly identified the patient's methemoglobinemia, the hemolytic process was not recognized as a potential contraindication to treatment. Understanding why manifestations of disease occur and incorporating this knowledge into diagnostic decision making remain important parts of the practice of medicine.
What was the pathophysiology underlying this patient's condition? Both the methemoglobinemia and the hemolytic anemia were triggered by oxidative stress. Methemoglobin (oxidized hemoglobin) is constantly produced in erythrocytes as a result of the close interaction between hemoglobin and oxygen. Under normal circumstances, methemoglobin is converted back into hemoglobin by the NADH-dependent cytochrome-b5 methemoglobin reductase system (Figure 1), thereby maintaining an equilibrium in which less than 1 percent of hemoglobin exists as methemoglobin.1 This balance can be upset by exposure to agents such as local anesthetics, sulfa antibiotics, and nitrite-containing compounds (including nitroglycerin and nitroprusside) that increase the levels of reactive oxygen species.1,2
Figure 1. Reduction of Hemoglobin to Methemoglobin under Normal Circumstances and after the Administration of Methylene Blue.
Under normal circumstances, the NADH-dependent–cytochrome-b5 methemoglobin reductase system efficiently reduces methemoglobin to hemoglobin. Methylene blue provides an alternative means by which methemoglobin can be reduced. In patients with G6PD deficiency, the production of NADPH is impaired, thereby limiting the conversion of methylene blue to its active metabolite, leukomethylene blue.
Oxidative stress also contributes to the hemolytic anemia seen in patients with G6PD deficiency. G6PD is important for the elimination of reactive oxygen species (Figure 2). In the absence of functional G6PD, accumulated oxidant exposure leads to denaturation of hemoglobin, impaired integrity of cell membranes, and hemolysis. Common precipitants include infection, drug exposure (especially exposure to dapsone, sulfamethoxazole, and primaquine), and, in susceptible persons, ingestion of foods such as fava beans.3,4
Figure 2. Role of G6PD in the Elimination of Reactive Oxygen Species in the Erythrocyte.
G6PD catalyzes the first step in the hexose monophosphate shunt by converting glucose-6-phosphate to 6-phosphogluconate. The reaction produces NADPH, which donates electrons to glutathione, thereby providing a means by which reactive oxygen species can be reduced to hydrogen peroxide and, ultimately, to water.
Two concepts are critical for understanding how such severe hemolysis developed in this patient: inactivation of the X chromosome and activity of the G6PD enzyme. As the discussant points out, G6PD deficiency is transmitted on the X chromosome. Therefore, all men with the disease are hemizygous, and most women are heterozygous. Early in the process of female embryonic development, one X chromosome is inactivated in each somatic cell, which results in two cell populations, one with normal G6PD and one with deficient G6PD (Figure 3).4,5 The clinical phenotype is determined both by the percentage of G6PD-deficient red cells and by the degree of G6PD activity in the enzyme-deficient cells. Many of the Asian variants of G6PD deficiency are characterized by enzyme levels that are barely detectable. In persons with these variants, hemolysis can be severe and life-threatening.6
Figure 3. Inactivation of the X Chromosome.
Females who are heterozygous for G6PD deficiency inherit one normal X chromosome (shown in black) and one abnormal one (red). Early in embryonic life, an X chromosome is inactivated in each somatic cell to form a Barr body (condensed ovals). The process results in two populations of cells: one with normal G6PD activity (cells with black chromosomes) and one with a deficiency of G6PD (cells with red chromosomes). In most cases, equal numbers of normal cells and G6PD-deficient cells are produced; however, it is possible to have an imbalance between the two cell populations.
This case was made especially challenging by the absence of a classically recognized precipitant. Could it have been the Chinese dumplings? No localizing symptoms were present to suggest infection, which is the most common precipitant of hemolysis in patients with G6PD deficiency3; despite repeated review of the patient's history, no other gustatory culprit emerged. The patient specifically denied having ingested fava beans, which are uncommon in Chinese cuisine. Nitrates are converted to nitrites by intestinal bacteria, and the ingestion of foods, including dumplings, that contain high quantities of nitrates have been reported in rare cases to cause methemoglobinemia.7,8,9 Sodium nitrite, a preservative commonly used to cure meats and fish, has also been implicated in outbreaks of methemoglobinemia that is related to such foods.7,8 Nitrites can also cause brisk hemolysis in patients with G6PD deficiency.10,11,12 We suspect that the dumplings were indeed the precipitating agent.
The association between methemoglobinemia and G6PD deficiency carries important implications for treatment. Although methylene blue rapidly reverses methemoglobinemia in most patients, persons with G6PD deficiency are a notable exception. Methylene blue is largely ineffective in these patients because NADPH, which is required to reduce methylene blue to its active metabolite, leukomethylene blue, is not available (Figure 1).13 Furthermore, methylene blue is a powerful oxidant and can precipitate or worsen hemolytic anemia associated with G6PD deficiency, as in this patient.13,14
The failure of the treating physicians to consider the concomitant diagnosis of G6PD deficiency in this patient and to connect the underlying pathophysiology of her hemolysis with methemoglobinemia left them at risk for making a common error in diagnostic reasoning, known as "premature closure."15 Premature closure occurs when a diagnosis is applied to a patient's condition despite incomplete or contradictory supporting data. The danger of rendering a premature diagnosis is that diagnostic evaluation is often discontinued, and associated conditions may be overlooked or ignored. It is critical that clinicians completely analyze all available data, consider the pathophysiology of suspected diagnoses, and keep their minds open to the possibility of alternative or concomitant diagnoses. As this case illustrates, appropriate diagnosis and treatment require that clinicians understand not only what is happening but also why it is happening.
Supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and by a grant (P20-HS11540) from the Patient Safety Developmental Center of the Agency for Healthcare Research and Quality (both to Dr. Saint).
We are indebted to Dr. Ernest Beutler for his kind review of the manuscript.
Source Information
From the Department of Medicine, Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado Health Sciences Center, Denver (W.J.J., H.R.C.); the Department of Medicine, San Francisco Veterans Affairs Medical Center and University of California, San Francisco, School of Medicine, San Francisco (G.D.); and Ann Arbor Veterans Affairs Health Services Research and Development Center of Excellence, the Patient Safety Enhancement Program, Ann Arbor Veterans Affairs Medical Center, and the University of Michigan Health System and the Department of Internal Medicine — all in Ann Arbor, Mich. (S.S.).
Address reprint requests to Dr. Janssen at the University of Colorado Health Sciences Center, 4200 E. 9th Ave., Box C-272, Denver, CO 80262, or at william.janssen@uchsc.edu.
References
Wright RO, Lewander WJ, Woolf AD. Methemoglobinemia: etiology, pharmacology, and clinical management. Ann Emerg Med 1999;34:646-656.
Mansouri A, Lurie AA. Concise review: methemoglobinemia. Am J Hematol 1993;42:7-12.
Beutler E. Glucose-6-phosphate dehydrogenase deficiency. N Engl J Med 1991;324:169-174.
Beutler E. G6PD deficiency. Blood 1994;84:3613-3636.
Beutler E. The genetics of glucose-6-phosphate dehydrogenase deficiency. Semin Hematol 1990;27:137-164.
Verle P, Nhan DH, Tinh TT, et al. Glucose-6-phosphate dehydrogenase deficiency in northern Vietnam. Trop Med Int Health 2000;5:203-206.
Chan TY. Food-borne nitrates and nitrites as a cause of methemoglobinemia. Southeast Asian J Trop Med Public Health 1996;27:189-192.
Methemoglobinemia following unintentional ingestion of sodium nitrite -- New York, 2002. MMWR Morb Mortal Wkly Rep 2002;51:639-642.
Walley T, Flanagan M. Nitrite-induced methaemoglobinaemia. Postgrad Med J 1987;63:643-644.
Brandes JC, Bufill JA, Pisciotta AV. Amyl nitrite-induced hemolytic anemia. Am J Med 1989;86:252-254.
Costello C, Pourgourides E, Youle M. Amyl nitrite induced acute haemolytic anaemia in HIV-antibody positive man. Int J STD AIDS 2000;11:334-335.
Neuberger A, Fishman S, Golik A. Hemolytic anemia in a G6PD-deficient man after inhalation of amyl nitrite ("poppers"). Isr Med Assoc J 2002;4:1085-1086.
Rosen PJ, Johnson C, McGehee WG, Beutler E. Failure of methylene blue treatment in toxic methemoglobinemia: association with glucose-6-phosphate dehydrogenase deficiency. Ann Intern Med 1971;75:83-86.
Liao YP, Hung DZ, Yang DY. Hemolytic anemia after methylene blue therapy for aniline-induced methemoglobinemia. Vet Hum Toxicol 2002;44:19-21.
Voytovich AE, Rippey RM, Suffredini A. Premature conclusions in diagnostic reasoning. J Med Educ 1985;60:302-307.(William J. Janssen, M.D.,)
A 38-year-old woman presented to the emergency room for evaluation of shortness of breath and jaundice. The previous day, she had attended a wedding, where she felt well. Several hours after the wedding, a headache developed, she had mild dizziness, and she noticed that her urine was dark brown. The day after the wedding, she awoke with mild shortness of breath and yellow discoloration of her eyes and skin.
Jaundice is usually caused by hepatobiliary disorders or hemolysis. Biliary obstruction, acute hepatitis, or brisk intravascular hemolysis could explain the jaundice and dark urine. Wilson's disease — with hepatocellular disease and associated hemolysis — could cause jaundice and dark urine.
Dyspnea and jaundice can be related in several ways. Sepsis can cause hyperbilirubinemia and hyperventilation. Hemolysis can lead to dyspnea from anemia and to indirect hyperbilirubinemia from red-cell destruction. Some infections, such as amebiasis, can affect both the liver and the lung. In addition, in the setting of chronic liver disease with associated jaundice — which would be unlikely in this patient, since her illness was acute — dyspnea can develop as a result of the hepatopulmonary syndrome, portopulmonary hypertension, hepatic hydrothorax, or ascites.
Since the patient's illness occurred in the context of her recent attendance at a wedding, I am interested to know what she ate and drank there.
At the wedding, the patient consumed Chinese dumplings containing salt-cured meat and a glass of wine. No one else who attended the wedding became sick. The patient had no abdominal pain, nausea, vomiting, diarrhea, or fever.
I am uncertain whether the patient's illness can be attributed to events at the wedding. Acute infection with hepatitis A virus can occur with ingestion of contaminated food, but the latency period is weeks, not hours. A fatty meal might lead to cholecystitis or choledocholithiasis, but the absence of abdominal pain argues against these disorders. Enterohemorrhagic Escherichia coli, specifically strain O157:H7, can cause the hemolytic–uremic syndrome, although the latency period is at least 24 hours, and diarrhea and abdominal pain are typical symptoms.
In the absence of gastrointestinal symptoms, I favor the consideration of an acute hemolytic episode. A potential ingestion at the wedding, followed by hemolysis, raises the possibility of a deficiency of glucose-6-phosphate dehydrogenase (G6PD). The only food I am aware of that is associated with episodic hemolysis from G6PD deficiency is the fava bean.
The patient had a history of mild anemia. She had no history of liver disease. She took no medications or herbal supplements, had no drug allergies, and did not smoke or use illicit drugs. She drank alcohol occasionally. She was Vietnamese and had lived in the United States for seven years. She had been adopted and was unsure of her family history. She worked in a wire-manufacturing plant.
The most likely explanation for her anemia would be iron deficiency or, given her ethnic background, thalassemia. However, many of the chronic hemolytic anemias, which result from defects in the red-cell membrane, enzymes, or hemoglobin, still warrant consideration. The family history, unavailable here, is often helpful in diagnosing these disorders. Long-standing hemolytic anemia can sometimes be manifested as cholecystitis or choledocholithiasis caused by pigment gallstones. Although ingestion of a hepatotoxin might explain the patient's presentation, she reports no history of the use of potentially hepatotoxic medications or herbs and no substantial alcohol intake. Her history of having lived in Vietnam supports the consideration of hepatitis B infection as a cause of her illness.
In general, the combination of jaundice and dark urine is more likely to be a result of hepatobiliary disease than of hemolysis. However, chronic anemia, new dyspnea, and the absence of hepatotoxins, known liver disease, and gastrointestinal symptoms lead me to favor the consideration of hemolysis over an acute hepatobiliary disorder.
The patient was afebrile, with a pulse of 98 beats per minute and a blood pressure of 112/64 mm Hg. Her respiratory rate was 24 breaths per minute, and her oxygen saturation 85 percent. Her skin was jaundiced, her sclerae were icteric, and her conjunctivae were pale. An early systolic murmur was present at the right upper sternal border, and the apical impulse was hyperkinetic. Her lungs were clear on auscultation, and there was no abdominal tenderness or hepatosplenomegaly. The remainder of her physical examination was normal.
The pale conjunctivae, tachycardia, systolic murmur, and hyperkinetic apical impulse are all consistent with anemia, and the jaundice — with no evidence of hepatic or abdominal disease — suggests hemolysis as the cause. Splenomegaly may be seen in many forms of hemolysis but is not an invariable feature and can be challenging to detect by physical examination. Therefore, the absence of palpable splenomegaly is relatively uninformative.
The oxygen saturation of 85 percent is unexpected and not explained by hemolysis. Anemia is a disorder involving a decreased overall capacity to carry oxygen, but the remaining red cells should be adequately saturated in the presence of normal alveolar gas exchange. This particular oxygen saturation brings to mind methemoglobinemia, an acquired dyshemoglobinemia caused by a number of drugs or chemicals that oxidize the iron in hemoglobin, rendering it incapable of carrying oxygen. Sulfonamide antibiotics, local anesthetics such as benzocaine, and antimalarial drugs are common precipitants. Measurement of arterial blood gases with co-oximetry would provide information on the partial pressure of oxygen in arterial blood and on the fraction of the total hemoglobin that is methemoglobin. If both hemolysis and methemoglobinemia were present, I would search for an agent that could trigger both events, and I would suspect that the patient has either a G6PD deficiency or an unstable hemoglobin variant (such as hemoglobin K?ln). There is no history of potential culprit drugs, such as dapsone or primaquine, but perhaps there was a component of the food or drink, such as a nitrate or sulfate, that triggered both processes.
If arterial blood analysis reveals hypoxemia without methemoglobinemia, the chest radiograph might suggest other diagnoses. Hilar lymphadenopathy might indicate lymphoma with associated autoimmune hemolysis; an infiltrate could suggest cold autoimmune hemolytic anemia with Mycoplasma pneumoniae infection. A normal chest radiograph would raise the possibility of a pulmonary embolism, in which case the rare disorder of paroxysmal nocturnal hemoglobinuria, characterized by hemolysis, pancytopenia, and venous thrombosis, would be worth considering.
The white-cell count was 13,600 per cubic millimeter, the hemoglobin 8.1 g per deciliter, and the platelet count 194,000 per cubic millimeter. The reticulocyte count was 1.8 percent (45,000 reticulocytes per cubic millimeter of blood [normal range, 25,000 to 75,000]). The total bilirubin level was 9.3 mg per deciliter (10.2 μmol per liter), and the lactate dehydrogenase level 2152 U per milliliter (normal range, 91 to 180). The aspartate aminotransferase and alanine aminotransferase levels were 120 U per liter (normal range, 6 to 30) and 41 units per liter (normal range, 10 to 40), respectively. The serum alkaline phosphatase, albumin, and electrolyte levels were normal. A urine dipstick test was positive for blood; no red cells or casts were seen on microscopical examination. Viral serologic tests for hepatitis A, B, and C viruses were negative. The chest radiograph was normal. An electrocardiogram revealed sinus tachycardia.
The low hemoglobin level and elevated lactate dehydrogenase and bilirubin levels suggest ongoing hemolysis. The results of the urine studies are compatible with hemoglobinuria and indicate rapid and severe red-cell destruction; brisk hemolysis can sometimes cause elevations in aminotransferases. The initial reticulocyte count is inappropriately low, but the bone marrow can be delayed in its response to hemolysis. The normal chest radiograph increases my suspicion of methemoglobinemia.
The cause of the patient's hemolysis remains a puzzle. Given its acute onset, the preexisting anemia, the possibility of methemoglobinemia, and no clear evidence of underlying illness, I favor the hypothesis that some type of oxidant exposure at the wedding triggered a hemolytic episode in this patient, who probably has G6PD deficiency or an unstable hemoglobinopathy. Direct examination of the peripheral blood would be helpful. Disorders that render red cells susceptible to oxidant stress may result in bite cells on the peripheral smear. Heinz bodies (cellular inclusions consisting of damaged aggregated hemoglobin) may be detected with supravital staining.
Despite the administration of supplemental oxygen, the oxygen saturation remained at 85 percent. Arterial blood gas values while the patient was breathing pure oxygen by face mask were as follows: pH 7.45, partial pressure of carbon dioxide 29 mm Hg, and partial pressure of oxygen 432 mm Hg. Co-oximetry demonstrated a methemoglobin level of 8.8 percent (normal range, 0.4 to 1.5).
With the acute onset of hemolysis and methemoglobinemia, I suspect that the patient's red cells were exposed to oxidant stress and that she has either a G6PD deficiency or an unstable-hemoglobin disorder that makes her susceptible to oxidant-induced hemolysis.
How should this patient be treated? Two mechanisms — decreased red-cell mass and ineffective hemoglobin (methemoglobin) — have led to insufficient delivery of oxygen to her tissues. The first-line treatment for acquired methemoglobinemia is the administration of methylene blue, which quickly reduces methemoglobin to hemoglobin. However, G6PD deficiency is a contraindication to treatment with methylene blue because patients with this disorder are unable to metabolize methylene blue, an oxidant that can exacerbate hemolysis. With a hemoglobin level that is already quite low, this patient may not tolerate additional hemolysis.
Toxin-mediated or oxidant-mediated hemolysis is usually managed supportively after the exposure has been terminated. With G6PD deficiency, most acute episodes of hemolysis are self-limited, and only the most severe cases require transfusion. Blood transfusion would accomplish the goal of restoring a normal hemoglobin level and oxygen-carrying capacity, but it involves additional risks. I favor blood transfusion over the use of methylene blue because of the suspicion of oxidant-mediated hemolysis due to G6PD deficiency.
Intravenous methylene blue was administered. Subsequent testing demonstrated worsening anemia, with a hemoglobin value of 6.5 g per deciliter. The patient received 4 units of packed red cells, and her red-cell count was stabilized. The activity of G6PD, as measured in a blood sample obtained before transfusion, was 3.9 U per gram of hemoglobin (normal range, 4.6 to 13.5).
This measured G6PD level may actually understate the patient's degree of G6PD deficiency, since the G6PD level measured during a hemolytic episode reflects only the surviving red-cell population; these surviving red cells evade hemolysis owing to G6PD levels that are higher than average. The gene for G6PD resides on the X chromosome. In a female patient, this raises the interesting possibility that the surviving red-cell population represents cells whose precursors have selectively deactivated their deficient X chromosome.
The patient was seen in the clinic several months later and was doing well. Her hemoglobin measured 12.6 g per deciliter, and a peripheral-blood smear demonstrated only mild reticulocytosis.
Commentary
A fundamental component in the care of any patient is an understanding and appreciation of the pathophysiology underlying his or her disease. Furthermore, it is essential that clinicians continue to evaluate available clinical data — even after a particular diagnosis is suspected — to ensure that there are no discordant findings. As this case illustrates, failure in these two areas can lead to inappropriate management and iatrogenic morbidity. Although the treating physicians correctly identified the patient's methemoglobinemia, the hemolytic process was not recognized as a potential contraindication to treatment. Understanding why manifestations of disease occur and incorporating this knowledge into diagnostic decision making remain important parts of the practice of medicine.
What was the pathophysiology underlying this patient's condition? Both the methemoglobinemia and the hemolytic anemia were triggered by oxidative stress. Methemoglobin (oxidized hemoglobin) is constantly produced in erythrocytes as a result of the close interaction between hemoglobin and oxygen. Under normal circumstances, methemoglobin is converted back into hemoglobin by the NADH-dependent cytochrome-b5 methemoglobin reductase system (Figure 1), thereby maintaining an equilibrium in which less than 1 percent of hemoglobin exists as methemoglobin.1 This balance can be upset by exposure to agents such as local anesthetics, sulfa antibiotics, and nitrite-containing compounds (including nitroglycerin and nitroprusside) that increase the levels of reactive oxygen species.1,2
Figure 1. Reduction of Hemoglobin to Methemoglobin under Normal Circumstances and after the Administration of Methylene Blue.
Under normal circumstances, the NADH-dependent–cytochrome-b5 methemoglobin reductase system efficiently reduces methemoglobin to hemoglobin. Methylene blue provides an alternative means by which methemoglobin can be reduced. In patients with G6PD deficiency, the production of NADPH is impaired, thereby limiting the conversion of methylene blue to its active metabolite, leukomethylene blue.
Oxidative stress also contributes to the hemolytic anemia seen in patients with G6PD deficiency. G6PD is important for the elimination of reactive oxygen species (Figure 2). In the absence of functional G6PD, accumulated oxidant exposure leads to denaturation of hemoglobin, impaired integrity of cell membranes, and hemolysis. Common precipitants include infection, drug exposure (especially exposure to dapsone, sulfamethoxazole, and primaquine), and, in susceptible persons, ingestion of foods such as fava beans.3,4
Figure 2. Role of G6PD in the Elimination of Reactive Oxygen Species in the Erythrocyte.
G6PD catalyzes the first step in the hexose monophosphate shunt by converting glucose-6-phosphate to 6-phosphogluconate. The reaction produces NADPH, which donates electrons to glutathione, thereby providing a means by which reactive oxygen species can be reduced to hydrogen peroxide and, ultimately, to water.
Two concepts are critical for understanding how such severe hemolysis developed in this patient: inactivation of the X chromosome and activity of the G6PD enzyme. As the discussant points out, G6PD deficiency is transmitted on the X chromosome. Therefore, all men with the disease are hemizygous, and most women are heterozygous. Early in the process of female embryonic development, one X chromosome is inactivated in each somatic cell, which results in two cell populations, one with normal G6PD and one with deficient G6PD (Figure 3).4,5 The clinical phenotype is determined both by the percentage of G6PD-deficient red cells and by the degree of G6PD activity in the enzyme-deficient cells. Many of the Asian variants of G6PD deficiency are characterized by enzyme levels that are barely detectable. In persons with these variants, hemolysis can be severe and life-threatening.6
Figure 3. Inactivation of the X Chromosome.
Females who are heterozygous for G6PD deficiency inherit one normal X chromosome (shown in black) and one abnormal one (red). Early in embryonic life, an X chromosome is inactivated in each somatic cell to form a Barr body (condensed ovals). The process results in two populations of cells: one with normal G6PD activity (cells with black chromosomes) and one with a deficiency of G6PD (cells with red chromosomes). In most cases, equal numbers of normal cells and G6PD-deficient cells are produced; however, it is possible to have an imbalance between the two cell populations.
This case was made especially challenging by the absence of a classically recognized precipitant. Could it have been the Chinese dumplings? No localizing symptoms were present to suggest infection, which is the most common precipitant of hemolysis in patients with G6PD deficiency3; despite repeated review of the patient's history, no other gustatory culprit emerged. The patient specifically denied having ingested fava beans, which are uncommon in Chinese cuisine. Nitrates are converted to nitrites by intestinal bacteria, and the ingestion of foods, including dumplings, that contain high quantities of nitrates have been reported in rare cases to cause methemoglobinemia.7,8,9 Sodium nitrite, a preservative commonly used to cure meats and fish, has also been implicated in outbreaks of methemoglobinemia that is related to such foods.7,8 Nitrites can also cause brisk hemolysis in patients with G6PD deficiency.10,11,12 We suspect that the dumplings were indeed the precipitating agent.
The association between methemoglobinemia and G6PD deficiency carries important implications for treatment. Although methylene blue rapidly reverses methemoglobinemia in most patients, persons with G6PD deficiency are a notable exception. Methylene blue is largely ineffective in these patients because NADPH, which is required to reduce methylene blue to its active metabolite, leukomethylene blue, is not available (Figure 1).13 Furthermore, methylene blue is a powerful oxidant and can precipitate or worsen hemolytic anemia associated with G6PD deficiency, as in this patient.13,14
The failure of the treating physicians to consider the concomitant diagnosis of G6PD deficiency in this patient and to connect the underlying pathophysiology of her hemolysis with methemoglobinemia left them at risk for making a common error in diagnostic reasoning, known as "premature closure."15 Premature closure occurs when a diagnosis is applied to a patient's condition despite incomplete or contradictory supporting data. The danger of rendering a premature diagnosis is that diagnostic evaluation is often discontinued, and associated conditions may be overlooked or ignored. It is critical that clinicians completely analyze all available data, consider the pathophysiology of suspected diagnoses, and keep their minds open to the possibility of alternative or concomitant diagnoses. As this case illustrates, appropriate diagnosis and treatment require that clinicians understand not only what is happening but also why it is happening.
Supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and by a grant (P20-HS11540) from the Patient Safety Developmental Center of the Agency for Healthcare Research and Quality (both to Dr. Saint).
We are indebted to Dr. Ernest Beutler for his kind review of the manuscript.
Source Information
From the Department of Medicine, Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado Health Sciences Center, Denver (W.J.J., H.R.C.); the Department of Medicine, San Francisco Veterans Affairs Medical Center and University of California, San Francisco, School of Medicine, San Francisco (G.D.); and Ann Arbor Veterans Affairs Health Services Research and Development Center of Excellence, the Patient Safety Enhancement Program, Ann Arbor Veterans Affairs Medical Center, and the University of Michigan Health System and the Department of Internal Medicine — all in Ann Arbor, Mich. (S.S.).
Address reprint requests to Dr. Janssen at the University of Colorado Health Sciences Center, 4200 E. 9th Ave., Box C-272, Denver, CO 80262, or at william.janssen@uchsc.edu.
References
Wright RO, Lewander WJ, Woolf AD. Methemoglobinemia: etiology, pharmacology, and clinical management. Ann Emerg Med 1999;34:646-656.
Mansouri A, Lurie AA. Concise review: methemoglobinemia. Am J Hematol 1993;42:7-12.
Beutler E. Glucose-6-phosphate dehydrogenase deficiency. N Engl J Med 1991;324:169-174.
Beutler E. G6PD deficiency. Blood 1994;84:3613-3636.
Beutler E. The genetics of glucose-6-phosphate dehydrogenase deficiency. Semin Hematol 1990;27:137-164.
Verle P, Nhan DH, Tinh TT, et al. Glucose-6-phosphate dehydrogenase deficiency in northern Vietnam. Trop Med Int Health 2000;5:203-206.
Chan TY. Food-borne nitrates and nitrites as a cause of methemoglobinemia. Southeast Asian J Trop Med Public Health 1996;27:189-192.
Methemoglobinemia following unintentional ingestion of sodium nitrite -- New York, 2002. MMWR Morb Mortal Wkly Rep 2002;51:639-642.
Walley T, Flanagan M. Nitrite-induced methaemoglobinaemia. Postgrad Med J 1987;63:643-644.
Brandes JC, Bufill JA, Pisciotta AV. Amyl nitrite-induced hemolytic anemia. Am J Med 1989;86:252-254.
Costello C, Pourgourides E, Youle M. Amyl nitrite induced acute haemolytic anaemia in HIV-antibody positive man. Int J STD AIDS 2000;11:334-335.
Neuberger A, Fishman S, Golik A. Hemolytic anemia in a G6PD-deficient man after inhalation of amyl nitrite ("poppers"). Isr Med Assoc J 2002;4:1085-1086.
Rosen PJ, Johnson C, McGehee WG, Beutler E. Failure of methylene blue treatment in toxic methemoglobinemia: association with glucose-6-phosphate dehydrogenase deficiency. Ann Intern Med 1971;75:83-86.
Liao YP, Hung DZ, Yang DY. Hemolytic anemia after methylene blue therapy for aniline-induced methemoglobinemia. Vet Hum Toxicol 2002;44:19-21.
Voytovich AE, Rippey RM, Suffredini A. Premature conclusions in diagnostic reasoning. J Med Educ 1985;60:302-307.(William J. Janssen, M.D.,)