Diagnosing PNH: how and whom

Issue: February 10, 2010

ByWendell F. Rosse, MD

Paroxysmal nocturnal hemoglobinuria is a rare hematopoietic disorder characterized by intravascular hemolysis, thrombosis and a variable degree of bone marrow hypofunction. It is caused by a somatic mutation in a hematopoietic stem cell that prevents the biosynthesis of the glycosylphosphatidylinositol anchor that tethers many proteins to the outer membrane surface. Two of these, CD55 and CD59, are important in modulating the effect of activation of complement on the cell surface; their absence results in the lysis of red cells and the activation of platelets by the even minimal activation of complement.

The intravascular hemolysis results in hemolytic anemia and hemglobinuria; the hemolytic anemia causes symptoms of nitric oxide (NO) depletion — smooth muscle spasm manifest as esophageal spasm, abdominal pain, erectile dysfunction and perhaps elevation of pulmonary artery pressure — while the hemoglobinuria has long-term deleterious effects on the kidney. Thrombosis occurs in the abdomen — Budd-Chiari syndrome, portal vein thrombosis — in the periphery and in the cerebral venous sinuses. The disease is serious and often fatal. As much as 35% of patients die in the first five years, and the mean survival after diagnosis is about 15 years.

How to test

Although the diagnosis may be suspected from clinical signs, the definitive diagnosis must be made from examination of the cells of the blood.

For many years, this was done by assessing the sensitivity of the red blood cells to the lytic effects of complement (the Ham test and sucrose lysis test). This method has been superseded by tests that use flow cytometry to determine the absence of the glycosylphosphatidylinositol (GPI) anchor or of the proteins normally attached to it. The anchor itself can be identified by a bacterial product, an aerolysin from a Pseudomonas species that, when inactivated and fluoresceinated, is called FLAER. This reagent can be used to detect granulocytes and monocytes with the PNH defect with great precision; as few as 0.03% of such cells may be accurately found.

Wendell Rosse, MD
Wendell Rosse

The absence of a variety of proteins may be detected using appropriately labeled monoclonal antibodies: anti-CD16 (with some reservations), anti-CD24 and anti-CD66b for granulocytes, anti-CD14 for monocytes, and anti-CD59 or anti-CD55 for red cells. Examination of granulocytes is important to determine the proportion of abnormal hematopoiesis in the marrow, and examination of red cells is necessary to determine the degree of abnormality.

In most patients, the abnormal cells will totally lack the anchor and the anchored proteins (PNH III cells). In some patients, the lack may be partial (PNH II cells). A consensus report on the optimal techniques for testing for PNH cells is being prepared by the Clinical Cytometry Society.

Whom to test

Since PNH is such a protean disease, the diagnosis is often missed because the defining tests are not done. Therefore, it is important to consider the clinical settings in which the test may give the diagnosis.

Hemoglobinuria: Although the syndrome was first defined by nocturnal hemoglobinuria, less than 20% of patients present with this symptom. However, any patient with unexplained hemoglobinuria (or hemosiderinuria) should be tested for PNH.
Acquired normocytic hemolytic anemia: The red cells in PNH are normal in shape (not spherocytic or schistocytic), the direct antiglobulin test is negative, and the plasma lactic dehydrogenase (LDH) is elevated. The blood of any patient with these clinical and laboratory characteristics should be tested for PNH.
Unexplained iron deficiency: Patients with PNH tend to lose large quantities of iron in the urine, and the resulting iron deficiency may mask some signs of hemolysis (increased reticulocyte count) but not others (increased serum LDH).
PNH symptoms: Patients with symptoms of PNH — excessive fatigue, esophageal spasm, intermittent abdominal pain, and erectile dysfunction (especially in the young) — should be examined for evidence of intravascular hemolysis (serum LDH); if this is elevated, testing for PNH is indicated.
Thrombosis: All patients with thrombosis in the unusual sites characteristic of PNH (hepatic veins, inferior vena cava, veins of the portal system, cerebral venous sinuses) should be tested for PNH. Patients with other thromboses should be tested for PNH if there are any signs of intravascular hemolysis.
Unexplained cytopenia: As much as 75% of patients with PNH have either thrombocytopenia, granulocytopenia or both, due to an element of bone marrow failure present in all patients with PNH. If such cytopenia is found, the patient should be tested for signs of hemolysis that, if present, should lead to testing for PNH.
Other hematologic disorders: About 20% of patients with PNH have a history of aplastic anemia. As much as 70% of patients with newly diagnosed aplastic anemia may have a small population of cells with the PNH defect, and 15% of patients with aplastic anemia who survive untransplanted develop a population sufficiently large (usually greater than 10%) to result in clinical PNH. Therefore, it is important to test all patients with aplastic anemia at diagnosis using very sensitive techniques. If a small PNH population is found, the patient should be tested at least once a year or before if evidence of hemolysis occurs.
Myelodysplastic syndrome: A small population of PNH cells may also be found in about 25% of patients with myelodysplasia, particularly those with refractory anemia, but also in all the variants of the syndrome. Significant populations (> 10%) are much less common than in aplastic anemia (about 3.4%). It is not known whether the progression to clinical PNH occurs as it clearly does in aplastic anemia.

In aplastic anemia and myelodysplastic syndrome, the presence of the PNH clone has important therapeutic implications. In both syndromes, the PNH cells predict a much more favorable response to immunosuppressive therapy and an overall better prognosis.

In most studies, the diagnosis of PNH has been delayed by two to five years from the onset of symptoms in many, if not most, cases. It is hoped that increased awareness of the need to diagnose the disease, combined with better methods for making the diagnosis, will reduce this delay. The reduction may be important, as there is now available specific treatment for many of the manifestations of the disease, and early treatment may reduce the initial death rate.

Wendell Rosse, MD, is Professor Emeritus at Duke University and Red Cells & Hemoglobinopathies Section Editor of the HemOnc Today Editorial Board.