Introduction: Paroxysmal nocturnal hemoglobinuria (PNH) is a nonmalignant clonal disease of hematopoietic stem cells resulting from a somatic mutation in the PIGA gene. PNH frequently manifests in association with aplastic anemia (AA), in which PIGA mutations are believed to enable escape from the immune-mediated destruction by pathogenic T cells. Recent studies using next-generation sequencing have revealed that frequent somatic PIGA mutationsin AA patients are associated with a better response to IST and prognosis (Yoshizato et al N Engl J Med. 2015; 373: 35-47). However, clinical PNH is a progressive and life-threatening disease driven by chronic hemolysis that leads to thrombosis, renal impairment, poor quality of life, and death. Large studies in adults have reported that clinical PNH developed in 10%-25% of AA patients; however; the frequency of clinical PNH in children with AA has rarely been described. Here we aimed to elucidate the pathological link between PNH and AA in children.

Methods: In total, 57 children (35 boys and 22 girls) diagnosed with acquired AA at our hospital between 1992 and 2010 were retrospectively studied. Patients who underwent hematopoietic stem cell transplantation as first-line treatment within 1 year after AA diagnosis and those with clinical PNH at AA diagnosis were excluded. Flow cytometry (FCM) was used to detect PNH CD13+/CD55/CD59 granulocytes and PNH glycophorin A+/CD55/CD59 red blood cells (RBCs). Clinical PNH was defined as the presence of intravascular hemolysis and ≥5% PNH granulocytes or PNH RBCs. Minor PNH clones were defined as those with >0.005% PNH granulocytes or >0.010% PNH RBCs. We performed targeted sequencing of bone marrow samples from patients with clinical PNH that were obtained at 2 time points: at AA diagnosis and after PNH development. The panel of 184 genes for targeted sequencing included most of the genes known to be mutated in inherited bone marrow failure syndromes and myeloid cancers, as well as PIGA.

Results: The median patient age at AA diagnosis was 9.3 (1.2-17.8) years, and the median follow-up period was 123 (2-228) months. A total of 43 patients were screened for PNH clones by FCM after AA diagnosis, and 21 of these with minor PNH clones were identified. The median percentages of PNH granulocytes and PNH RBCs were 0.001% (0.000%-4.785%) and 0.000% (0.000%-3.829%), respectively. During follow-up, 5 patients developed clinical PNH after adolescence (15-22 years of age). The median time between AA diagnosis and PNH development was 4.9 (3.3-7.9) years. All clinical PNH patients were treated with IST for AA, and complete and partial response after 6 months were achieved in 1 and 4 patients, respectively. Gross hemoglobinuria was present in all clinical PNH patients, but thrombosis was not observed. The size of PNH clones varied greatly among patients: PNH granulocytes and PNH RBCs were 42.96% (10.04%-59.50%) and 48.87% (15.02%-90.80%), respectively. Oral cyclosporine A and intravenous eculizumab were administered to 3 and 1 patients, respectively; all patients showed sustained response as indicated by improvement in gross hemoglobinuria and normal blood counts after treatment. The remaining 1 patient underwent bone marrow transplantation from the HLA-identical mother and was alive without any complications. Overall, the 10-year probability of developing clinical PNH was 10.2% (95%CI, 3.6-20.7). Among 43 patients screened for PNH clones at AA diagnosis, the 10-year cumulative clinical PNH incidence was significantly higher in patients with minor PNH clones than in those without minor PNH clones at AA diagnosis [29% (95% CI, 10%-51%) vs. 0% (95% CI, 0%-0%); p = 0.015]. Among all clinical PNH patients, a total of 8 somatic PIGA mutations were detected (missense, 2; splice site, 2; and frameshift, 4). However, PIGA mutations were not detected at AA diagnosis even in patients who subsequently developed clinical PNH.

Conclusion: In our cohort, the percentage of patients who eventually developed clinical PNH was comparable to that reported in adults in a previous study. Furthermore, the current study showed that the presence of minor PNH clones at AA diagnosis was a risk factor for the subsequent development of clinical PNH, although the clones were not detected by targeted sequencing. Thus, pediatric AA patients with PNH clones at AA diagnosis should undergo long-term periodic monitoring for potential clinical PNH development.


Kojima:SANOFI: Honoraria, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.