Busulfan-Dose Escalation in Reduced Intensity Stem Cell Transplantation (RIC-HSCT) Results in Good Outcomes for Patients with MDS and AML without Increased Toxicity

Current evidence indicates that patients receiving myeloablative HSCT have decreased relapse rates but higher transplant-related mortality (TRM) than patients receiving RIC-HSCT. It is well established that there is wide variation in intensity within protocols considered RIC. We have pioneered the use of an alemtuzumab based HSCT with two doses of busulfan (bu) (8mg PO or 6.4mg IV), however it is noted that relapse rates within this context are high (Potter V. et al, BBMT, Jan 2014). Targeted dose escalation of bu in allogeneic transplant conditioning regimens has the potential to reduce relapse rates. We therefore adopted an intensification strategy by increasing the total bu dose (16mg PO or 12.8mg IV) administered during conditioning.

Ninety-five consecutive patients with a median follow-up for survivors of 1083 days (204 - 4627) received FB4C HSCT between 2000 and 2014. The protocol included fludarabine 150mg/m², bu 12.8mg/kg and alemtuzumab (FB4C) as previously described. Patients received pre-emptive DLI for low (<50%) or falling CD3 chimerism. The median age of the cohort was 59 years (19 - 70). Diagnosis included RCMD (n=14), RAEB (n=18), de novo AML (n= 21), AML with multi-lineage dysplasia (AML-MLD) (n=34) and CMML (n= 8). Patients received PBSC (n=89) or BM (n=6) with a respective median cell dose of 6.8 and 2.5 x10⁶CD34/kg. Neutrophil and platelet engraftment occurred at medians of 13 and 16 days respectively. Graft failure occurred in only 6% of patients.

Five year overall survival (OS) for the entire cohort was 56% with a TRM at 100 days of 9.7% and 30% at 5 years. Disease free survival (DFS) at 1 year was 65% and 39% at 5 years. Relapse rates were 24% at 1 and 44% at 5 years. There was no significant difference in outcome for patients receiving sibling vs matched unrelated, 10/10 HLA vs 9/10 HLA matched or BM vs PBSC stem cell source. OS for patients with de novoAML was 62% compared to 43% for MDS-AML, 69% for RCMD, 25% for CMML and 70% for RAEB. These differences were not significant. Twenty patients receiving pDLI did particularly well with a 5 year OS of 79%.

Patients were stratified by age group: <50 years (n= 28), 51-60 years (n= 33) and >60 years (n= 34). These patients were equivalent in terms of disease status at transplantation, IPSS-R, HCT-CI and cytogenetic risk groups. In contrast to our previous results there was no adverse effect of advanced age. Patients <50 years had a 5 year OS of 66% compared to 47% for 50-60 and 51% age >60 (p=NS). TRM in patients <50 years was 25%, 30% for in those aged 50-60 and 36% in those >60 at 5 years. The DFS for those <50 at 5 years was 48%, 35% for 50-60 and 29% for those >60. When compared with patients with a HCT-CI of 0, patients with a HCT-CI score of 1-2 (p=0.04), or ≥3 (0.05), did significantly worse. OS for these groups was 65%, 48% and 50% respectively. The TRM for these groups was not significantly difference suggesting that bu dose intensification can be safely used even in those with additional comorbidities.

Cytogenetics were classified as low (n=5), intermediate (n=61) and poor (n=22) according to NCCN and IPSS-R categories. The small numbers of patients in the very poor and poor IPSS-R groups were grouped together for the final analysis. There was a trend toward significantly worse outcomes for those with poor cytogenetics compared to intermediate (p=0.06) with 5 year OS of 40% vs 62%. Five year DFS was significantly different at 21% vs 45% (p=0.03). OS at 5 years for patients with residual cytogenetic disease pre-HSCT was 41% compared to 63% for those without cytogenetic disease at 63% at the time of transplant (p=0.04). On multivariate analysis this was the only variable to retain significance (HR 1.6, (95% CI 1.02-2.6)). Rates of severe GvHD were low with a cumulative incidence of acute GvHD grade III-IV of 22%. Chronic GvHD rates were low overall at 33% and with NIH severe GvHD occurring in only 13.7%.

These results compare favourably with our previous publication of lower dose bu which demonstrated 5 year OS and DFS of 44% and 33% respectively. Relapse rate was also improved at 44% compared with 51% at five years. The lack of difference in TRM with regard to age and HCT-CI suggests that this protocol is potentially an effective method of optimising outcomes in high risk and older patients with MDS/AML without an increase in toxicity. Although these results are encouraging prospective randomised trials are required to identify the most appropriate setting for bu dose intensification.