Co-Morbidity and Disease Status at the Time of Transplant Predict Outcome Following Allogeneic Haematopoietic Stem Cell Transplantation (HSCT) for Poor Risk Myelodysplastic Syndrome (MDS) and Acute Myeloid Leukaemia (AML).

Allogeneic haematopoietic stem cell transplantation (allo-HSCT) remains the standard of care in poor risk myeloid malignancies. Identifying factors which predict both non-relapse mortality and relapse risk are imperative to patient selection. We report the outcome of 198 cases of poor risk MDS (n=70) and AML (n=128) who underwent allo-HSCT with either myeloablative (MA) (n=67) or reduced intensity conditioning (RIC) (n=131). A T-cell depleted regimen using Alemtuzumab was administered to all patients receiving RIC and for volunteer unrelated donor (VUD) MA transplants. 13% had a diagnosis of refractory anaemia excess blasts (RAEB I), 14% RAEB II, 15% AML with multi-lineage dysplasia, 49% de-novo AML and 9% chronic myelomonocytic leukaemia. The median age of the RIC cohort was 53years (21–72 years) compared with 31years (17–59 years) in the MA group. On comparison of RIC versus MA cohorts, 31% v 45% had a co-morbidity score of 0, 17% v 25% score 1, 17% v 16% score 2 and 35% v 14% score ≥3 (p=0.025). The Charlson co-morbidity index was used to define the pre-transplant score. No difference in disease status at time of transplant was observed between the RIC and MA groups (p=0.399) with 80% v 71% in first complete remission (CR1), 4% v 6% in second CR (CR2), 2% v 6% partial remission (PR), 9% v 9% progressive disease (PrD) and 5% v 8% dysplastic remission (DysCR). A higher number of VUD transplants 61% versus 45% (p=0.042) were observed in the RIC cohort, whereas a higher proportion of cases with poor risk cytogenetics (42% v 26%) (p=0.01) were observed in the MA group. No difference in FAB classification, prior therapy or HLA disparity was observed between the RIC and MA groups. The median follow-up for the MA and RIC groups was 1782days (50–4300days) and 1014days (168–2211days) respectively. The transplant related mortality (TRM) for the MA and RIC cohorts at day+100 was 31% v 10% and at 3years 47% v 30% respectively. The 3year overall survival (OS) was 41% v 47% (p=0.062), 3year disease free survival (DFS) 36% v 43% (p=0.156) and 3year relapse rate 34% v 40% (p=0.28) for the MA and RIC groups respectively. Multiple cox regression analysis identified co-morbidity risk score (p=0.003) disease status at time of transplant (p=0.012) and cytogenetic risk group (p=0.041) as the most significant factors affecting OS for all allogeneic HSCT. Analysis of factors affecting DFS further identified co-morbidity risk group (p=0.007) disease status at time of transplant (p=0.015) and cytogenetic risk group (p=0.022) as retaining significance. On analysis of RIC and MA cohorts separately co-morbidity and disease status remain the most significant variables affecting OS and DFS. With respect to disease status, PR and PrD consistently demonstrate significantly worse outcome than CR1, CR2, or DysCR subgroups. A co-morbidity index of ≥2 displayed inferior outcome to a score of 0 or 1. Age, sex, HLA disparity or sibling v VUD showed no effect on OS or DFS. In summary, the results demonstrate a trend towards improved OS following RIC transplantation, with no significant difference in DFS. Patient selection for allo-HSCT should not be based on age alone, whereas response to chemotherapy and the use of co-morbidity scores more accurately predict outcome following both MA and RIC regimens.