Black Mornings, Yellow Sunsets — A Day with Paroxysmal Nocturnal Hemoglobinuria
http://www.100md.com
《新英格兰医药杂志》
The first known description of hemoglobinuria — urine the color of a cola beverage — appears in a 13th-century text, the De Urinis of Johannes Zacharias (Actuarius), physician to the court of Byzantium. Zacharias, the first person to use a graduated cylinder for urinalysis, had probably recorded a rare complication of falciparum malaria, one known as blackwater fever. In the centuries after the conquest of Constantinople, other causes of black urine were uncovered, but none were more fascinating than paroxysmal nocturnal hemoglobinuria (PNH).
PNH has a particular resonance in Boston for two reasons. One is that the first diagnostic test for the disease — Ham's test — was developed in 1938 by Thomas Hale Ham and his colleagues at the Thorndike Laboratory of Boston City Hospital a year after Ham had reported in the Journal that the erythrocytes in PNH are highly susceptible to lysis in an acidic environment. The second is the 1967 editorial in Blood by William Dameshek of the Tufts–New England Medical Center, in which he proposed that PNH, aplastic anemia, and acute leukemia are related disorders that arise in the wake of bone marrow injury. The mechanism of the susceptibility of PNH erythrocytes to acid lysis, as observed by Ham, is now known, but the biologic basis of Dameshek's conjecture remains a puzzle.
PNH has three attributes: complement-dependent intravascular hemolytic anemia, thrombosis, and bone marrow failure. Hemolysis often occurs during sleep, when the decline in the blood pH triggers an activation of complement components. The urine is black when the patient awakens, because the amount of free hemoglobin passing through the glomeruli exceeds the absorptive capacity of the renal tubules. The hemoglobin-drenched tubules lock the iron of hemoglobin into hemosiderin, and the continuous sloughing of these cells into the urine culminates in iron deficiency. Thrombosis is frequent in PNH for reasons that are unclear; thrombosis of the hepatic vein, a much-feared complication, occurs in up to one third of patients with PNH.
Ham suspected that the lysis of PNH erythrocytes was actually mediated by complement, but a search for a complement-fixing antibody was fruitless, because the alternative, antibody-independent pathway of complement activation was unknown in his time. The alternative and antibody-dependent pathways converge at the point of the cleavage of C3 into C3a and C3b, both of which participate in cleaving C5. The cleavage of C5 into C5a and C5b initiates the formation of the C5b–C9 membrane-attack complex and ultimately the polymerization of C9 into molecular tubes that puncture the cell membrane.
It is likely that the alternative pathway continuously assembles membrane-attack complexes, because it is in a constant state of low-level activation by microbial and plant substances. Erythrocytes (and other blood cells) defend themselves from these complexes with two membrane-bound proteins, CD55 (decay-accelerating factor), which binds C3b, and CD59, which inhibits the insertion of C9 into the membrane. CD59 is probably the dominant protector. In all bona fide cases of PNH, blood cells are deficient not only in CD55 and CD59, but also in a large variety of other membrane proteins that have only one obvious connection to one another: all are anchored to the membrane by a phospholipid, glycosylphosphatidylinositol (GPI).
The susceptibility of erythrocytes to complement in PNH is due to an acquired (somatic) mutation of the phatidylinositol glycan class A gene (PIG-A) in a pluripotent hematopoietic stem cell. The PIG-A protein is required for the synthesis of the GPI anchor — hence the lack of numerous anchor-dependent proteins from the surface of blood cells in patients with PNH. Almost 200 different mutations of PIG-A have been found in PNH; some of them completely inactivate the protein, and others cause partial inactivation. These variants, PNH type III and PNH type II, respectively, can coexist as distinct clones in the same patient. Indeed, all cases of PNH are erythrocyte chimeras, with one or more mutant clones within a population of normal red cells (called PNH type I). The lack of CD59 permits the precise quantitation of PNH erythrocytes and the estimation of the size of the PNH clone by means of flow cytometry.
In his groundbreaking 1967 editorial, Dameshek proposed that erythropoiesis in PNH was "ecologically advantageous." This insight emphasized the overlapping relationship Dameshek saw between PNH and aplastic anemia — in about 10 percent of patients, aplastic anemia evolves into overt PNH. More recently, leukocytes with the PNH phenotype have been detected in about one third of patients with aplastic anemia. In such cases, it is not unusual to find multiple mutations of PIG-A arising in different stem cells. Equally interesting is that impaired hematopoiesis is a regular feature not only of aplastic anemia, but also of PNH. Moreover, in PNH, the growth of blood-forming stem cells without the PIG-A mutation is impaired. Unlike a tumor cell, however, the mutant PNH clone does not have an inherent proliferative advantage; it cannot expand, and thereby become clinically evident, unless there is a simultaneous element of bone marrow failure.
It seems likely that an insult to the marrow underlies PNH. One view is that the damage causes a mutation in PIG-A, resulting in either subtle marrow failure or blatant aplastic anemia. The finding of PNH cells in aplastic anemia and myelodysplastic syndromes and the rare cases of acute myelogenous leukemia that emerge from PNH support this hypothesis (see Figure). Others argue that PNH clones resist the autoimmune attack on hematopoietic stem cells that causes aplastic anemia and thereby gain a selective advantage, but this appealing theory — an echo of Dameshek's notion of the "ecologically advantageous" hematopoiesis — has not been incisively tested. The presence of rare cells with the PNH phenotype in some normal people suggests an alternative mechanism: bone marrow damage does not induce the PIG-A mutation but, rather, allows these minute clones to expand.
Figure. Relationship of PNH to Aplastic Anemia and Myelodysplasia.
An unknown mechanism, probably an environmental toxin, damages hematopoietic stem cells and induces a mutation of PIG-A. The type and extent of damage determine whether aplastic anemia, PNH, or a myelodysplastic syndrome results.
PNH requires supportive treatment, at least at first. Some patients may thrive after receiving a bone marrow transplant from a sibling. Immunosuppressive therapy for the underlying bone marrow failure is another option. In this issue of the Journal, Hillmen and colleagues (pages 552–559) report the treatment of 11 cases of PNH with a humanized antibody that prevents the cleavage of C5 and hence the formation of the membrane-attack complex. Treatment with the antibody sharply decreased the need for transfusions and the rate of bouts of hemoglobinuria and increased the proportion of PNH type III cells in the circulation (presumably because they were protected from destruction). These results not only substantiate the complement-dependent mechanism of hemolytic anemia in PNH, but they also open possibilities for other novel treatments of the disease. It is unclear, however, whether the antibody therapy must be continued for the rest of the patient's life. Also unknown is whether PNH type III clones expand or contract with treatment and whether the antibody improves marrow function generally. Since spontaneous remissions of PNH do occasionally occur, it is possible that the mutant clone could eventually contract, eliminating the need for treatment. The effect, if any, of long-term impairment of the membrane-attack complex is unknown.
PNH is a rare disease, even a curiosity to some. But studies of this disorder have advanced our knowledge on a broad front, and in so doing they have revealed new kinds of treatment that will have considerable relevance to other diseases involving impaired marrow function.(Robert S. Schwartz, M.D.)
PNH has a particular resonance in Boston for two reasons. One is that the first diagnostic test for the disease — Ham's test — was developed in 1938 by Thomas Hale Ham and his colleagues at the Thorndike Laboratory of Boston City Hospital a year after Ham had reported in the Journal that the erythrocytes in PNH are highly susceptible to lysis in an acidic environment. The second is the 1967 editorial in Blood by William Dameshek of the Tufts–New England Medical Center, in which he proposed that PNH, aplastic anemia, and acute leukemia are related disorders that arise in the wake of bone marrow injury. The mechanism of the susceptibility of PNH erythrocytes to acid lysis, as observed by Ham, is now known, but the biologic basis of Dameshek's conjecture remains a puzzle.
PNH has three attributes: complement-dependent intravascular hemolytic anemia, thrombosis, and bone marrow failure. Hemolysis often occurs during sleep, when the decline in the blood pH triggers an activation of complement components. The urine is black when the patient awakens, because the amount of free hemoglobin passing through the glomeruli exceeds the absorptive capacity of the renal tubules. The hemoglobin-drenched tubules lock the iron of hemoglobin into hemosiderin, and the continuous sloughing of these cells into the urine culminates in iron deficiency. Thrombosis is frequent in PNH for reasons that are unclear; thrombosis of the hepatic vein, a much-feared complication, occurs in up to one third of patients with PNH.
Ham suspected that the lysis of PNH erythrocytes was actually mediated by complement, but a search for a complement-fixing antibody was fruitless, because the alternative, antibody-independent pathway of complement activation was unknown in his time. The alternative and antibody-dependent pathways converge at the point of the cleavage of C3 into C3a and C3b, both of which participate in cleaving C5. The cleavage of C5 into C5a and C5b initiates the formation of the C5b–C9 membrane-attack complex and ultimately the polymerization of C9 into molecular tubes that puncture the cell membrane.
It is likely that the alternative pathway continuously assembles membrane-attack complexes, because it is in a constant state of low-level activation by microbial and plant substances. Erythrocytes (and other blood cells) defend themselves from these complexes with two membrane-bound proteins, CD55 (decay-accelerating factor), which binds C3b, and CD59, which inhibits the insertion of C9 into the membrane. CD59 is probably the dominant protector. In all bona fide cases of PNH, blood cells are deficient not only in CD55 and CD59, but also in a large variety of other membrane proteins that have only one obvious connection to one another: all are anchored to the membrane by a phospholipid, glycosylphosphatidylinositol (GPI).
The susceptibility of erythrocytes to complement in PNH is due to an acquired (somatic) mutation of the phatidylinositol glycan class A gene (PIG-A) in a pluripotent hematopoietic stem cell. The PIG-A protein is required for the synthesis of the GPI anchor — hence the lack of numerous anchor-dependent proteins from the surface of blood cells in patients with PNH. Almost 200 different mutations of PIG-A have been found in PNH; some of them completely inactivate the protein, and others cause partial inactivation. These variants, PNH type III and PNH type II, respectively, can coexist as distinct clones in the same patient. Indeed, all cases of PNH are erythrocyte chimeras, with one or more mutant clones within a population of normal red cells (called PNH type I). The lack of CD59 permits the precise quantitation of PNH erythrocytes and the estimation of the size of the PNH clone by means of flow cytometry.
In his groundbreaking 1967 editorial, Dameshek proposed that erythropoiesis in PNH was "ecologically advantageous." This insight emphasized the overlapping relationship Dameshek saw between PNH and aplastic anemia — in about 10 percent of patients, aplastic anemia evolves into overt PNH. More recently, leukocytes with the PNH phenotype have been detected in about one third of patients with aplastic anemia. In such cases, it is not unusual to find multiple mutations of PIG-A arising in different stem cells. Equally interesting is that impaired hematopoiesis is a regular feature not only of aplastic anemia, but also of PNH. Moreover, in PNH, the growth of blood-forming stem cells without the PIG-A mutation is impaired. Unlike a tumor cell, however, the mutant PNH clone does not have an inherent proliferative advantage; it cannot expand, and thereby become clinically evident, unless there is a simultaneous element of bone marrow failure.
It seems likely that an insult to the marrow underlies PNH. One view is that the damage causes a mutation in PIG-A, resulting in either subtle marrow failure or blatant aplastic anemia. The finding of PNH cells in aplastic anemia and myelodysplastic syndromes and the rare cases of acute myelogenous leukemia that emerge from PNH support this hypothesis (see Figure). Others argue that PNH clones resist the autoimmune attack on hematopoietic stem cells that causes aplastic anemia and thereby gain a selective advantage, but this appealing theory — an echo of Dameshek's notion of the "ecologically advantageous" hematopoiesis — has not been incisively tested. The presence of rare cells with the PNH phenotype in some normal people suggests an alternative mechanism: bone marrow damage does not induce the PIG-A mutation but, rather, allows these minute clones to expand.
Figure. Relationship of PNH to Aplastic Anemia and Myelodysplasia.
An unknown mechanism, probably an environmental toxin, damages hematopoietic stem cells and induces a mutation of PIG-A. The type and extent of damage determine whether aplastic anemia, PNH, or a myelodysplastic syndrome results.
PNH requires supportive treatment, at least at first. Some patients may thrive after receiving a bone marrow transplant from a sibling. Immunosuppressive therapy for the underlying bone marrow failure is another option. In this issue of the Journal, Hillmen and colleagues (pages 552–559) report the treatment of 11 cases of PNH with a humanized antibody that prevents the cleavage of C5 and hence the formation of the membrane-attack complex. Treatment with the antibody sharply decreased the need for transfusions and the rate of bouts of hemoglobinuria and increased the proportion of PNH type III cells in the circulation (presumably because they were protected from destruction). These results not only substantiate the complement-dependent mechanism of hemolytic anemia in PNH, but they also open possibilities for other novel treatments of the disease. It is unclear, however, whether the antibody therapy must be continued for the rest of the patient's life. Also unknown is whether PNH type III clones expand or contract with treatment and whether the antibody improves marrow function generally. Since spontaneous remissions of PNH do occasionally occur, it is possible that the mutant clone could eventually contract, eliminating the need for treatment. The effect, if any, of long-term impairment of the membrane-attack complex is unknown.
PNH is a rare disease, even a curiosity to some. But studies of this disorder have advanced our knowledge on a broad front, and in so doing they have revealed new kinds of treatment that will have considerable relevance to other diseases involving impaired marrow function.(Robert S. Schwartz, M.D.)