A Single Institution’s Experience with Allogeneic Haematopoietic Stem Cell Transplantation for Acute Lymphoblastic Leukaemia: Analysis of Factors Leading to Improved Outcome

Allogeneic haematopoietic stem cell transplantation remains the only curative option for many patients with acute lymphoblastic leukaemia (ALL). The optimal timing for transplant, and donor selection in this context, as well as the differing outcomes in paediatric and adult patients, remain areas of study. We reviewed the results in a single institution treating both children and adults, over a 16-year period from 1991–2007, to address a number of these issues. The study included 196 patients. The median age of the patients was 16 years (range: 2–60). The donor was a sibling in 99, unrelated (UD) in 93 and mismatched relative in 4. The transplant was performed in CR1 (71), CR2 (83), >CR2 (24) or with persistent disease (16). 16% of patients were Philadelphia chromosome positive and 22% had T-cell ALL. 8% of patients had received a prior autograft. The majority of patients had conditioning with a TBI-containing regimen. Most of the patients with an UD had in-vivo T-cell depletion (TCD) with Alemtuzumab (78%), while sibling donor transplants were T-cell replete. The median follow-up was 3.9 years (0.8–14.9 years). The overall survival in the whole group was 48% and 42% at 5 and 10 years respectively. The relapse risk and non-relapse mortality (NRM) were 22% and 29% at 3 years and 25% and 30% at 5 years. Acute GvHD was present in 64% of patients (grade II–IV in 29%). We were able to show a significant improvement in outcome dependent on the time period over which the transplant was performed. The 3-year survival was 33%, 51% and 57% in those transplanted from 1991–1995, 1996–2003 and 2004–2007 respectively (p=0.03). This appeared to be largely due to a decrease in the NRM in more recent years (3 year: 40%, 32% and 18% for these groups, p=0.04). Interestingly, there was no significant difference observed in OS, NRM or relapse risk based on the use of either a sibling or UD. Whilst there was no significant difference in outcome dependent on whether the transplant was performed in CR1/CR2 or beyond CR2 in children, transplants beyond CR2 or with residual disease were associated with a worse outcome in adults (p=0.01). Survival was significantly worse in patients who had received a previous autograft (p=0.0001). Although only a small number of patients (10) received reduced intensity conditioning for their transplant, the OS was not significantly different to those receiving myeloablative conditioning (2 years: 45% vs 53%, p=0.4). We did observe a significant difference in outcome in adult (>16) compared to paediatric patients (16 and younger), 5 year OS: 32% and 62% respectively, p=0.0004. The NRM was significantly higher in adult patients (3 year: 37% compared to 21% in children, p=0.01), however there was no significant difference in disease relapse (3 year: 28% in adults compared to 24% in children). In adult patients only one factor impacted significantly on OS in multivariate analysis: a transplant undertaken in 2004–2007 compared to earlier years resulted in a significant survival benefit (HR: 0.2, 95% CI 0.1; 0.7, p=0.009). In paediatric patients two factors impacted on OS in multivariate analysis: transplants performed after 1995 compared to earlier years resulted in a significant survival benefit (HR: 0.2, 95% CI 0.05; 0.8, p=0.024), and there was a significantly worse survival in patients with a T cell phenotype (HR: 2.7, 95% CI 1.0; 6.7, p=0.031). In conclusion, we report outcomes in a large group of patients transplanted for ALL in a single institution. We show an improvement in patient survival and NRM over time, particularly in adult patients, despite the inclusion of older patients and RIC regimens in recent years. It appears that unrelated and sibling donors can be used to achieve a similar long-term outcome.