Abstract

BackgroundHematopoietic stem cell transplantation remains the only treatment for most patients with severe combined immunodeficiencies (SCIDs) or other primary immunodeficiencies (non-SCID PIDs).ObjectiveTo analyze the long-term outcome of patients with SCID and non-SCID PID from European centers treated between 1968 and 2005.MethodsThe product-limit method estimated cumulative survival; the log-rank test compared survival between groups. A Cox proportional-hazard model evaluated the impact of independent predictors on patient survival.ResultsIn patients with SCID, survival with genoidentical donors (n = 25) from 2000 to 2005 was 90%. Survival using a mismatched relative (n = 96) has improved (66%), similar to that using an unrelated donor (n = 46; 69%; P = .005). Transplantation after year 1995, a younger age, B+ phenotype, genoidentical and phenoidentical donors, absence of respiratory impairment, or viral infection before transplantation were associated with better prognosis on multivariate analysis. For non-SCID PID, in contrast with patients with SCID, we confirm that, in the 2000 to 2005 period, using an unrelated donor (n = 124) gave a 3-year survival rate similar to a genoidentical donor (n = 73), 79% for both. Survival was 76% in phenoidentical transplants (n = 23) and worse in mismatched related donor transplants (n = 47; 46%; P = .016).ConclusionThis is the largest cohort study of such patients with the longest follow-up. Specific issues arise for different patient groups. Patients with B-SCID have worse survival than other patients with SCID, despite improvements in each group. For non-SCID PID, survival is worse than SCID, although more conditions are now treated. Individual disease categories now need to be analyzed so that disease-specific prognosis may be better understood and the best treatments planned. Hematopoietic stem cell transplantation remains the only treatment for most patients with severe combined immunodeficiencies (SCIDs) or other primary immunodeficiencies (non-SCID PIDs). To analyze the long-term outcome of patients with SCID and non-SCID PID from European centers treated between 1968 and 2005. The product-limit method estimated cumulative survival; the log-rank test compared survival between groups. A Cox proportional-hazard model evaluated the impact of independent predictors on patient survival. In patients with SCID, survival with genoidentical donors (n = 25) from 2000 to 2005 was 90%. Survival using a mismatched relative (n = 96) has improved (66%), similar to that using an unrelated donor (n = 46; 69%; P = .005). Transplantation after year 1995, a younger age, B+ phenotype, genoidentical and phenoidentical donors, absence of respiratory impairment, or viral infection before transplantation were associated with better prognosis on multivariate analysis. For non-SCID PID, in contrast with patients with SCID, we confirm that, in the 2000 to 2005 period, using an unrelated donor (n = 124) gave a 3-year survival rate similar to a genoidentical donor (n = 73), 79% for both. Survival was 76% in phenoidentical transplants (n = 23) and worse in mismatched related donor transplants (n = 47; 46%; P = .016). This is the largest cohort study of such patients with the longest follow-up. Specific issues arise for different patient groups. Patients with B-SCID have worse survival than other patients with SCID, despite improvements in each group. For non-SCID PID, survival is worse than SCID, although more conditions are now treated. Individual disease categories now need to be analyzed so that disease-specific prognosis may be better understood and the best treatments planned.