Introduction: AIHA is a rare and potentially severe condition in children. To assess the presentation, treatment response and outcome of childhood AIHA, a national multicentric observational study has been started in 2002 in 32 French pediatric hematologic units.Patients and Methods: A descriptive analysis of 171 children under 18 years old, who are still being followed is presented. Inclusion criteria were hemoglobin <10 g/dl, positive Coombs test, and one of the three following hemolysis criteria: reticulocytes >120 000/mm3, haptoglobin <10 mg/dl, bilirubin >10 mg/dl. Etiological investigations were standardised. Treatment procedures were at the appreciation of each physician, according to the SHIP guidelines. Complete response (CR) was defined as Hb >10g/dl and absence of biological hemolysis for more than 3 months. Variables associated with survival in CR without treatment at the last follow up were assessed using a Cox model of multivariate analysis.Results: AIHA was diagnosed between January 1986 and March 2005 in 94 males and 77 females. The median follow-up is 9 months (range 0.2 to 234). Median age at diagnosis is 3,8 years (range 0.1–17.4). Familial or personal dysimmune history was respectively positive in 21 and 28 patients. A concomitant febrile episode was present in 50% of cases, and for half of the patients, the onset was sudden and severe. Median hemoglobin level at diagnosis was 5,6 g/dl (2 to 12,8) with a reticulocyte count below 100.000/mm3 in 43 patients. The direct Coombs test was positive for IgG in 37%, for IgG+C3b in 37%, for isolated C3b in 17%, for cold agglutinin +IgM in 7% and for IgA in 2% of patients. Thrombopenia and neutropenia were associated in 26% and 18% of patients respectively. Evans syndrome was diagnosed in 66/171 patients, among whom idiopathic thrombopenic purpura and AIHA occurred simultaneously in 31/66. AIHA was a component of a dysimmune clinical or biological disease in 50% of patients, of post infectious origin in 25%, associated with malformations in 5%, and isolated in 20%. All but 8 patients received steroid therapy (1 to 5 mg/kg prednisone daily). Resistance to steroids led 91/163 patients to receive a median of 3 (1 to 8) subsequent modalities of treatment, mainly ciclosporine (n=25), anti-CD20 (n=24), splenectomy (n=22), and azathioprine (n=14). Nine children died, 8 of whom had Evans syndrome. Survival in CR without treatment at 5 and 10 years from diagnosis were respectively 42% and 20%. In multivariate analysis, three variables were associated with the absence of CR without treatment at the last follow up: Coombs test of IgG/IgG+C3b class (RR 0,22, 95% IC [0,09–0,5], p = 0,0004), dysgammaglobulinemia (RR 0,28, 95% IC [0,11–0,71], p = 0,007), and age more than 4 years (RR 0,53, 95% IC [0,27–1,02], p = 0,05).Conclusion: The present French cohort is the largest reported study of AIHA in children. Heterogeneous underlying etiological subsets are confirmed. Two prognostic factors are clearly identified: direct Coombs test of IgG or IgG+C3b class, and presence of a dysgammaglobulinemia. Ongoing hemato-immunologic and therapeutic studies will allow better comprehension and treatment of this rare disease
Read full abstract