• Open Access

Paroxysmal Nocturnal Hemoglobinuria from Bench to Bedside


RA Brodsky (rbrodsky@jhmi.edu)


Paroxysmal nocturnal hemoglobinuria (PNH) is a rare hematologic disease that presents with protean manifestations. Clinical and laboratory investigation over the past 25 years has uncovered most of the basic science underpinnings of PNH and has led to the development of a highly effective targeted therapy. PNH originates from a multipotent hematopoietic stem cell (HSC) that acquires a somatic mutation in a gene called phosphatidylinositol glycan anchor biosynthesis, class A (PIG-A). The PIG-A gene is required for the first step in glycosylphosphatidylinositol (GPI) anchor biosynthesis. Failure to synthesize GPI anchors leads to an absence of all proteins that utilize GPI to attach to the plasma membrane. Two GPI-anchor proteins, CD55 and CD59, are complement regulatory proteins; their absence on the surface of PNH cells leads to complement-mediated hemolysis. The release of free hemoglobin leads to scavenging of nitric oxide and contributes to many clinical manifestations, including esophageal spasm, fatigue, and possibly thrombosis. Aerolysin is a pore-forming toxin that binds GPI-anchored proteins and kills normal cells, but not PNH cells. A fluorescinated aerolysin variant (FLAER) binds GPI-anchor and serves as a novel reagent diagnosing PNH. Eculizumab, a humanized monoclonal antibody against C5, is the first effective drug therapy for PNH. Clin Trans Sci 2011; Volume 4: 219–224


Paroxysmal nocturnal hemoglobinuria (PNH) is a rare, clonal hematopoietic stem cell (HSC) disease that manifests as bone marrow failure, hemolytic anemia, smooth muscle dystonias, and thrombosis.1,2 The median survival in untreated patients ranges from 10 to 20 years.3–5 PNH originates from a multipotent HSC that acquires a mutation in a gene called phosphatidylinositol glycan anchor biosynthesis, class A (PIG-A).6,7 The PIG-A gene product is required for the biosynthesis of the glycosylphosphatidylinositol (GPI) anchor, a glycolipid moiety that attaches dozens of proteins to the plasma membrane of cells. Consequently, the PNH stem cell and all of its progeny have a reduction or absence of GPI-anchor proteins (GPI-APs). Two of these proteins, CD55 and CD59, are complement regulatory proteins; the absence of these proteins is fundamental to the pathophysiology of PNH. CD55 inhibits C3 convertases and CD59 blocks C9 incorporation into the membrane attack complex (MAC). The absence of CD55 and CD59 makes PNH cells vulnerable to complement mediated intravascular and extravascular hemolysis, although it is the intravascular hemolysis that most contributes to the morbidity of the disease. Intravascular hemolysis releases free hemoglobin into the plasma. Free plasma hemoglobin scavenges nitric oxide (NO) and depletion of NO at the tissue level contributes to numerous PNH manifestations. Recently, eculizumab, a monoclonal antibody that inhibits the terminal stage of the complement cascade, has been shown to decrease hemolysis and thrombosis and to markedly improve the quality of life in patients with PNH.


One of the earliest descriptions of PNH occurred in 1882 by Dr. Paul Strübing.8 He described a 29-year-old Cartwright who presented with intermittent fatigue, abdominal pain, and severe nocturnal paroxysms of hemoglobinuria. In 1925, Enneking formally introduced the term “paroxysmal nocturnal hemoglobinuria”9 and in 1937, Thomas Hale Ham found that erythrocytes collected from PNH patients were hemolyzed when incubated with normal, acidified serum. Interestingly, the hemolytic process could be abrogated by heat inactivation, which implicated complement in the pathophysiology.10 This seminal laboratory discovery was quickly translated to clinical practice as the first diagnostic test for PNH, the acidified serum test (Ham test). In 1954, with the discovery of the alternative pathway of complement activation, Pillemer et al. demonstrated that increased complement sensitivity was the cause of hemolysis in PNH.11 During 1980s, it was discovered that PNH cells display a global deficiency of an entire class of cell surface proteins, known as GPI-APs. Shortly thereafter, the gene defect in PNH, PIG-A, was discovered.12–14 This ushered in a period on intensive laboratory and clinical studies that ultimately led to the development of effective drug therapy for PNH.


GPI-anchor biosynthesis and PIG-A

GPI is synthesized in the endoplasmic reticulum and is transferred en bloc to the carboxyl terminus of a protein that has a GPI-attachment signal peptide. The mature GPI-AP is then transported to the plasma membrane and resides in 50–350 nm microdomains known as lipid rafts. Biosynthesis of GPI anchors involves at least 10 reactions and more than 20 different genes.15 (Figure 1) The first step in GPI-anchor biosynthesis is the transfer of N-acetylglucosamine (GlcNAc) from UDP-GlcNAc to phosphatidylinositol (PI) to yield GlcNAc-PI. This step is catalyzed by GlcNAc:PI a1-6 GlcNAc transferase, an enzyme whose subunits are encoded by seven different genes: PIGA, PIGC, PIGH, GPI1, and, PIGY, PIGP, and DPM2. In the second step, GlcNAc-PI is deacetylated by the gene product of PIG-L to form GlcN-PI. GPI anchor assembly continues in the endoplasmic reticulum with acylation of the inositol and stepwise addition of mannosyl and phosphoethanolamine residues. The common core structure of GPI consists of a molecule of PI, and a glycan core that contains glucosamine, three mannoses, and an ethanolamine phosphate (Figure 2). Given the numerous gene products involved in GPI-anchor assembly, it seemed improbable that PNH would be the consequence of a single genetic mutation. However, after intense investigation of this pathway it became apparent that in all PNH patients the defect could be attributed to mutations in the PIG-A gene.12,13 Later, it was determined that the PIG-A gene resides on the X chromosome and that its product is part of a complex that transfers N-aceytlglucosamine (GlcNAc) to phosphatidylinositol (PI) to form GlcNAc-PI, the first step in GPI-anchor biosynthesis.16 Thus, a single “hit” will generate a PNH phenotype since males have only one X chromosome and in females one X chromosome is inactivated through lyonization. Conceivably a mutation in any one of the genes in this pathway would cause the disease; however, other genes involved in GPI-anchor biosynthesis are located on autosomes. Inactivating mutations in these genes would have to occur on both alleles to produce the PNH phenotype.

Figure 1.

Biosynthesis of GPI-anchored proteins. GPI-anchor biosynthesis takes place in the endoplasmic reticulum. PIG-A is one of seven genes required for the first step, which transfers N-acetylglucosamine (GlcNAc) from uridine 5′-diphospho-N-acetylglucosamide (UDP-GlcNAc) to phosphatidylinositol (PI) to yield GlcNAc-PI. After synthesis of the mature GPI precursor, the cognate protein is attached and then transported to the plasma membrane where the GPI-anchored protein resides in membrane rafts. PIG-A mutations lead to a defect in the first step in GPI-anchor biosynthesis resulting in intracellular degradation of the cognate protein and a lack of cell surface GPI-anchored proteins. (Image cited from Brodsky RA in Blood Reviews 2008:22;65-74. “Advances in the diagnosis and therapy of paroxysmal nocturnal hemoglobinuria” with permission.) 254 × 190 mm (96 × 96 DPI)

Figure 2.

Core structure of the GPI anchor. GPI-AP consisits of Phosphatidylinositol (PI), Glycan core, and Ethanolamine (EtN). PI is inserted into the lipid bilayer of the plasma membrane. The glycan core, which contains the proaerolysin binding site, consists of a molecule of N-glucosamine (GlcNH2), three manose molecules (Man), and a molecule of ethanolamine (EtN). Ethanolamine on the terminal mannose covalently attaches to protein through an amide bond. (Image modified from Brodsky RA & Hu R. Leukemia & Lymphoma 2006; 47:1215-1221 “PIG-A mutations in paroxysmal nocturnal hemoglobinuria and in normal hematopoiesis” with permission.)

Other mechanisms of GPI-AP deficiency

Alternative mechanisms of GPI-AP deficiency in human cells have recently been discovered. Two rare, inherited forms of GPI-AP deficiency have recently been described. In the first case, two unrelated kindreds were found to harbor a promoter mutation in PIGM that disrupts binding of the transcription factor Sp1 to its cognate promoter motif.17 This mutation reduces the transcription of PIGM (a mannosyltransferase-encoding gene involved GPI-anchor biosynthesis) leading to severe GPI-AP deficiency. The affected patients presented with portal and hepatic vein thrombosis and intractable absence seizures by age two. Interestingly, the GPI-AP deficiency varied between different cells and tissues: GPI-AP expression was almost completely deficient in granulocytes, but near normal in red cells, hence the absence of anemia and hemolysis. Treatment of the patient with sodium butyrate effectively restored GPI-AP expression and terminated the seizure activity.18 In the second case, patients with hyperphosphatasia mental retardation (HPMR) syndrome, an autosomal recessive form of mental retardation with distinct facial features and elevated serum alkaline phosphatase, were found to harbor PIGV mutations.19 Acquired forms of GPI-AP deficiency that do not involve PIG-A has also been described.20 Three different Burkitt lymphoma cell lines and primary cells from Burkitt lymphoma patients were found to have marked GPI-AP deficiency due to transcriptional silencing of PIGY and/or PIGL. The GPI-AP-deficient Burkett cells also overexpressed Runx-1 and had a greater percentage of cells in G0/G1, suggesting that they may mark of population of cells with stem cell properties. Similar to the PIGM promoter defects, treatment with a demethylating agent was able to restore cell surface GPI-AP expression.

Intravascular hemolysis in PNH

The absence of CD59 from the surface of PNH erythrocytes leads to intravascular hemolysis. CD59 is a 19,000 molecular weight GPI-AP that blocks the aggregation of C9 into the C5b-8 complex; thereby, effectively preventing the formation of the intact MAC.21 Hemolysis in PNH is chronic due to a continuous state of complement activation (tick-over), but brisk periods of hemolysis (paroxysms) may result from increases in complement activation due to infections, surgery, strenuous activity, and excessive alcohol intake (Figure 3).

Figure 3.

The complement cascade, hemolyisis, and eculizumab. CD55 and CD59 are complement regulatory proteins, CD 55 downregulates C3 convertase activity and CD59 blocks MAC formation. Deficiency of CD59 increases MAC formation and induces intravascular hemolysis, which is the primary cause of hemoglobinuria in PNH patients; deficiency of CD55 leads to increased C3 convertase activity and C3d-associated extravascular hemolysis. Eculizumab blocks C5 activation, therefore prevents MAC formation and intravascular hemolysis. In the alternative pathway, C3 tickover and C3 convertase amplification generate membrane bound C3d, which may lead to significant extravascular hemolysis in some PNH patients receiving eculizumab treatment.

Extravascular hemolysis in PNH

PNH patients also experience extravascular hemolysis due to absence of CD55 on the surface of red cells. CD55,22 is a 68,000 molecular weight GPI-AP that functions to accelerate the rate of destruction of membrane-bound C3 convertase. Extravascular hemolysis tends to be less evident in most PNH patients since CD59 deficiency leads more rapid red cell destruction.23 However, treatment of PNH patients with the terminal complement inhibitor, eculizumab, can make the extravascular hemolysis of PNH more conspicuous.24 (Figure 3).

NO scavenging

Many clinical manifestations of PNH are the consequence of NO depletion at the tissue level.25 Normally, NO is synthesized in the endothelium and functions to maintain proper vascular and muscle tone and to limit platelet activation. In PNH, extensive intravascular hemolysis generates large amount of free hemoglobin in serum that binds NO and serves as a potent scavenger. Furthermore, the release of erythrocyte arginase into serum during intravascular hemolysis decreases the substrate (arginine) for NO synthesis (Figure 4). The resulting deprivation of NO from tissues may manifest as fatigue, pain, esophageal spasm, male erectile dysfunction, renal impairment, and possibly thrombosis.

Figure 4.

Basic nitric oxide (NO) metabolism. A. NO synthase (NOS) combines oxygen and arginine to form NO and citrulline. B. Oxygen bound Fe2+ from hemoglobin is released into plasma by intravascular hemolysis, and NO is converted to inert nitrite along with the oxidation of hemoglobin to methemoglobin. In addition, hemolysis releases erythrocyte arginase, which depletes substrate for NOS. (Image modified from Brodsky RA in Blood Reviews 2008:22; 65-74 “Advances in the diagnosis and therapy of paroxysmal nocturnal hemoglobinuria” with permission.)


Thrombosis is a notorious complication of PNH and the leading cause of death from this disease.3–5,26 It occurs in about 40% of PNH patients and more commonly involves the venous system. Patients with severe hemolytic anemia and PNH granulocyte clones of greater than 60% appear to be at greatest risk for thrombosis.5,27 The exact mechanism of thrombosis in PNH is still not clear. NO depletion has been associated with increased platelet aggregation, adhesion, and accelerated clot formation. Furthermore, in an attempt to repair complement-mediated damage, PNH platelets undergo exocytosis of the complement attack complex that results in the formation of microvesicles with phosphatidylserine externalization, a potent in vitro procoagulant. Activation of complement cascade and subsequent generation of C5a significantly upregulates expression of tissue factor (TF), a key initiating component of coagulation cascade.28 Fibrinolysis may also be perturbed in PNH given that PNH blood cells lack the GPI-anchored urokinase receptor. In addition, tissue factor pathway inhibitor (TFPI) has been shown to require a GPI-anchored chaperone protein for trafficking to the endothelial cell surface.29

The stem cell and clonal expansion

The mechanism(s) by which the PNH stem cell achieves dominance is still not entirely clear. Several hypotheses have been proposed. One of these hypotheses is a “two-step model,” which proposes that PIG-A mutations are benign events in HSCs and that these cells only undergo clonal expansion in the setting of immune selection that targets normal stem cells but spares PNH stem cells. Others have proposed that a second mutation is required to give the PNH clone a survival advantage or that PNH cells may undergo less apoptosis, especially in the setting of an immunologic attack on the bone marrow. Lastly, it has been hypothesized that clonal dominance in PNH arises from “neutral evolution.”30 This hypothesis proposes that PNH stem cells have neither an absolute nor conditional advantage, but that stochastic expansion of a PNH stem cell may be favored when there are few, but highly replicating, HSCs (e.g., recovery from aplastic anemia). There is indirect evidence to support parts of all of these hypotheses, but the lack of reliable disease models of aplastic anemia and PNH makes this an area of continued investigation.

Diagnosis of PNH

Flow cytometric assays can accurately detect GPI-AP deficiency in human blood cells, but PNH remains a clinical diagnosis.31–33 The diagnosis of PNH requires the presence of hemolysis (usually demonstrated by elevated levels of lactate dehydrogenase, an elevated reticulocyte count, and anemia) and documented deficiency of GPI-AP on two or more blood cell lineages. GPI-AP deficiency in blood cells is usually documented by flow cytometry using fluoresceinated monoclonal antibodies directed against individual GPI-AP. A fluorescein-labeled proaerolysin variant (FLAER) is increasingly being used as a flow cytometric assay to diagnose PNH.32 Aerolysin, the principal virulence factor of the bacterium Aeromonas hydrophila,34,35 is secreted as an inert protoxin, proaerolysin, that binds selectively and with high affinity to the GPI anchor.36 FLAER binds to the GPI anchor without forming channels and gives a highly accurate assessment of the GPI anchor deficit in PNH.32 Since the GPI anchor is the major determinant for binding FLAER, it allows for the direct assessment of GPI anchor expression on most cell lineages. Red cells are a notable exception; this may be because both normal and PNH red cells express large amounts of glycophorin, a protein shown to bind aerolysin weakly. Guidelines for the diagnosis and monitoring of PNH have recently been published.33

It is important to recognize that small populations of GPI-AP-deficient blood cells that can be found in up to 60% of patients with acquired aplastic anemia (usually 0.1–20% GPI-AP deficient granulocytes),37,38 20% of patients with myelodysplastic syndrome (usually 0.01–2% GPI-AP-deficient granulocytes), and even healthy controls (less than 0.01% GPI-AP-deficient granulocytes).39,40 In contrast to PNH patients, the PIG-A mutations in healthy controls arise from colony-forming cells rather than multipotent HSCs.40 Colony-forming cells lack self-renewal capacity (Figure 5). This probably explains why PIG-A mutations in healthy controls are transient and do not result in disease.

Figure 5.

Origin of PIG-A mutations in PNH and normal hematopoiesis. In PNH, PIG-A mutations arise from multipotent HSC, which can self-renew and have the capacity for clonal expansion. In normal hematopoiesis, PIG-A mutations occur in progenitor cells that can differentiate but are unable to self-renew.


Bone marrow transplantation

Bone marrow transplantation (BMT) is the only curative therapy for PNH with success rates ranging from 50% to 70%.41–44 BMT should not be offered as initial therapy for most patients with classical PNH given the high transplant related morbidity and mortality, especially when using unrelated or mismatched donors. It is now clear that a myeloablative conditioning regimen is not required to eradicate the PNH clone.45,46 Allogeneic BMT following nonmyeloablative conditioning regimens can cure PNH. Whether or not there is an advantage to one approach over the other will require further study; however, nonmyeloablative regimens may be preferable in young patients seeking to maintain fertility or patients with moderate organ dysfunction who may not tolerate a myeloablative regimen.

Targeted complement inhibition

Eculizumab is a humanized monoclonal antibody that blocks terminal complement activation at C5; it is the first drug to be approved by the Food and Drug Administration to treat PNH.47 Because C5 is common to all pathways of complement activation, eculizumab stops progression of the complement cascade. Moreover, prevention of C5 cleavage blocks the generation of C5a, a potent proinflammatory molecule, and C5b, the initial substrate for the formation of the MAC. Although blocking the terminal portion of the complement cascade increases the risk for Neisserial infections, C5 blockade preserves the more critical immunoprotective and immunoregulatory functions of the upstream components that result in C3b-mediated opsonization and clearance of immune complexes. The drug is administered intravenously at a dose of 600 mg weekly for the first 4 weeks; then increased to 900 mg biweekly indefinitely.

Eculizumab has been studied in two phase III clinical trials.48,49 Treatment with eculizumab markedly reduces the rate of intravascular hemolysis by preventing the assembly of the MAC on PNH erythrocytes, which decreases or eliminates the need for blood transfusions, improves quality of life, and reduces the risk of thrombosis.48–50 Eculizumab therapy seems to be most appropriate in PNH patients with disabling fatigue, thromboses, transfusion dependence, frequent pain paroxysms, renal insufficiency, or other end organ complications from disease.2 Watchful waiting is appropriate for asymptomatic patients or those with mild symptoms. The response to eculizumab approaches 100% but the quality of response varies from patient to patient; some patients achieve near-normal hemoglobin levels, while others show evidence of a compensated hemolytic anemia with an increased reticulocyte count and elevated levels of unconjugated bilirubin. This heterogenous response is due to variations in the amount of C3d deposition on the PNH red cells.24 Because eculizumab blocks complement at C5, the earlier steps of the complement pathway, including activation, deposition, and proteolytic cleavage of C3 to C3b and further split products, are not affected by eculizumab. Thus, CD55-deficient PNH red cells may become overloaded with C3 fragments (Figure 3).


PNH and its protean manifestations have fascinated clinicians and scientists for decades. Virtually all of the biology and clinical manifestations can be explained by the increased susceptibility of PNH cells to the alternative pathway of complement. The study of PNH has contributed to our understanding of the complement cascade, the red cell membrane, hematopoiesis, thrombosis, and the consequences of intravascular hemolysis. Eculizumab is the first clinically active complement inhibitor. It is currently indicated for the treatment of PNH, but eculizumab is demonstrating activity in a number of other complement-mediated diseases such as cold agglutinin disease,51 atypical hemolytic uremic syndrome,52 and kidney transplantation.53–55 Newer and even more targeted complement inhibitors are in clinical development and will likely be useful in treating a large number of complement mediated conditions.56


This study was supported by the National Institutes of Health (grant P01CA70970 and T32HL007525).