Пароксизмальная ночная гемоглобинурия у детей и взрослых: сравнительный клинический профиль и долгосрочный прогноз

Abstract

Paroxysmal nocturnal hemoglobinuria (PNH) is an extremely rare and still poorly understood disease in children. It is generally accepted that PNH in children is characterized by severe bone marrow failure, less intensive hemolysis, and less often complicated by thrombosis and organ dysfunction. Aim: to compare the clinical and laboratory characteristics and to determine the long-term prognosis of PNH with disease onset in children and adults. Retrospective and prospective study included patients with active hemolytic PNH, defined by flow cytometry results (clone more than 10% in granulocytes), clinical and laboratory criteria (lactate dehydrogenase (LDH) level above 1.5 of the upper limit of normal (ULN)). The classical PNH and aplastic anemia/PNH (AA/PNH) were studied. A detailed analysis of clinical manifestations, primary diagnoses, complications and overall survival was performed depending on the age of PNH onset (before and after age 18 years). As of August 2018, 355 patients were included in the study (56% female, 44% male) with classic form (49%) and AA/PNH (51%). The median age at the PNH onset and the diagnosis was 27 (5–80) and 30 years (7–80), respectively. The median PNH clone size in granulocytes and red blood cells was 92% and 33%, respectively, and the LDH level of 5.5 ULN. Hemolytic PNH onset during childhood was in 51 patients (14%) with the same frequency of classical form (n = 27; 16%) and AA/PNH (n = 24, 13%) ( p = 0.549). The diagnosis of classical form and AA/PNH was confirmed in childhood in 9 (5%) and 21 (12%) patients respectively. In 81% of classical PNH presenting in children, a misdiagnosis was initially established, and the median time before precise diagnosis was 52 months (4–426). The PNH clone size (median 93% vs 91%, p = 0.321) and hemolysis intensity according to LDH level (median 4.95 vs. 5.76 ULN, p = 0.690) did not differ in children and adults. The frequency of hemolysis symptoms in the onset of the disease correlated with the clone size and was higher in classical PNH compared with AA/PNH, but did not depend on onset age: weakness (98% in children vs. 96% in adults), hemoglobinuria (51% vs 55%), pain (39% vs 45%), jaundice (46% vs 52%), dysphagia (21% vs 28%). The age of PNH onset did not influence the frequency of early (before the diagnosis of PNH) thrombosis (8% vs 13%, p = 0.482) and episodes of acute kidney injury (AKI, 10% vs 9%, p = 0.788). With prolonged follow-up, there is a tendency towards a higher cumulative incidence of thromboses in adults: 5 years 21% vs 12%, 10 years 33% vs 19%, however in the more distant period these differences become insignificant ( p = 0.151). The cumulative incidence of AKI was similar in children and adults: 5 years 16% vs 20%, 10 years 23% vs 22%, respectively ( p = 0.500). The frequency of transfusion dependence, the development of chronic kidney disease and pulmonary hypertension and overall survival were also independent of the PNH onset age. The results of the study in the large cohort of patients indicate the similar clinical profile and a long-term prognosis of active hemolytic PNH in children and adults. Diagnostic errors and late recognition of classical PNH lead to the relative prevalence of AA/PNH in children.