Late Mortality After Hematopoietic Stem-Cell Transplantation for a Childhood Malignancy

Abstract 4458

Hematopoietic stem cell transplantation (HSCT) is used increasingly as a curative therapy in children with malignancies. However, survivors of HSCT are at risk for both disease recurrence and life-threatening complications beyond the acute period of transplantation. The risk for late mortality (> 2 years) after autologous and allogeneic HSCT has not been well reported in pediatric patients. The aim of this population based retrospective cohort study was to assess late mortality in children (0–18 years of age) who underwent a HSCT in the province of Ontario, Canada. All but a small minority of the province’s pediatric transplantations are performed at a single center (The Hospital for Sick Children [SickKids], Toronto). Indication for HSCT, type of transplant, donor source and survival were determined via a record linkage between the SickKids’ clinical transplant database and POGONIS, the provincial pediatric cancer registry. Mortality information contained in POGONIS is captured at the time of death in hospital, or by record linkage to the provincial mortality file. Records of all 842 children with a malignancy who underwent HSCT at SickKids between 1985 and 2009 were retrieved. Underlying diagnoses were leukemia (N=464), lymphoma (N=88), central nervous system tumor (N=60), neuroblastoma (N=167), sarcoma (N=44) and other (N=19). Four-hundred and nineteen children underwent allogeneic HSCT and 423 underwent autologous HSCT. Among the allogeneic HSCT group, 252 received stem cells from a related donor and 167 from an unrelated donor. Overall, 537 children (67.8%) survived for at least 2 years after HSCT. The median follow-up of these 537 survivors was 10.2 years, with 134 (25%) of the patients being followed for ≥15 years. Of these 537 survivors, 103 (19.2%) subsequently suffered a late death. A greater proportion of patients who underwent autologous HSCT had a late death (67/279; 24.0%) than patients who underwent allogeneic HSCT (36/258; 14.0%). In a multivariate analysis of patients who underwent autologous HSCT, males had increased risk of death compared to females (HR=2.3, 95% CI: 1.29–4.0), and patients treated for neuroblastoma had a greater risk of death than those treated for leukemia (HR=10.9, 95% CI: 3.0–40.0). Among patients treated with allogeneic HSCT, none of gender, donor source (related vs. unrelated), age at transplant or underlying diagnosis was associated with increased late mortality. In conclusion, children with cancer who are treated with a HSCT continue to be at risk for premature death even if they survive for two or more years after their transplant. Further investigation will focus on examining the contributions of treatment toxicity and disease recurrence to late mortality.