ABSTRACT
OBJECTIVE: Information about the long-term survival impact of hematopoietic stem cell transplant (HSCT) in adults with relapsed/refractory B-cell acute lymphoblastic leukaemia is limited. The objective was to conduct a systematic review identifying studies reporting survival in HSCT-receiving patients and apply parametric analyses to predict long-term survival. MATERIALS AND METHODS: Twenty-five relevant studies were identified. Analyses were conducted in 10 studies (n=503; "global" analysis) reporting overall survival (OS) data as Kaplan-Meier curves or at patient level. Four studies (n=217; "subgroup" analysis) measured OS from the point of HSCT. Patient-level data were recreated from Kaplan-Meier curves and pooled, with six models tested for longer-term extrapolation. Additionally, a sensitivity analysis was undertaken involving removal of data from the oldest study cohort (recruited between 1981-1997) to determine if the year which patients received HSCT impacted survival compared to post-2009 data. RESULTS: Median OS and five-year survival probability were 11.4 months and 24.4% (95% CI, 20.5-28.5%) in the global analysis, and 12.0 months and 28.4% (95% CI, 22.1-34.9%) in the subgroup analysis. The generalised gamma and Gompertz models fit longer-term extrapolation criteria. The generalised gamma model predicted survival at 10.4% vs. 14.8% (15 years), 8.3% vs. 12.8% (20 years), and 6.9% vs. 11.4% (25 years) for the global and subgroup analysis, respectively. The Gompertz model predicted survival to plateau at 23% vs. 25.6% just before 10 years. The sensitivity analysis excluding older data found median survival increased two-fold (25.3 vs. 12 months). CONCLUSIONS: Results synthesize long-term evidence of outcomes for HSCT-receiving patients, providing a basis for treatment comparison. Risk of death is low beyond four years and newer data appears correlated with improved outcomes.
