Alemtuzumab-Based RIC HSCT with Pre-Emptive DLI as An Effective Strategy to Achieve Long-Term Disease Remission In Patients with High Risk Myelodysplastic Syndrome (MDS) and Acute Myeloid Leukaemia (AML).

Abstract 1304

Disease relapse following allogeneic haematopoietic stem cell transplant (HSCT) for MDS/AML remains one of the primary causes of treatment failure, particularly in the setting of T-cell depleted reduced-intensity conditioning (RIC). Pre-emptive immunotherapy such as the use of donor lymphocyte infusions (DLI) is one strategy utilized to prevent relapse, although this approach may be associated with an increased risk of GvHD. We present our experience, evaluating toxicity and efficacy of pre-emptive DLI for the treatment of mixed donor chimerism (MDC) in a cohort of high risk MDS/AML patients who received a uniform alemtuzumab-based RIC HSCT protocol.

37 patients received a RIC protocol using fludarabine, busulphan and alemtuzumab. The stem cell source was BM (n=8) and PBSC (n=29) with a median cell dose of 2.40 and 5.34×106̂ CD34/kg respectively. The donor source was from HLA-matched sibling (n=20), HLA-matched unrelated (n=14) with 3 1-allele mismatched unrelated donors. Peripheral blood CD3 chimerism was monitored at day 30, 60, 100, 180 and 1 year post-HSCT. Patients with MDC (defined as CD3 <50%, or fall of CD3 of greater than 20% in a month) received pre-emptive DLI at a starting dose of 5×105̂CD3/kg or 1×106̂CD3/kg. In order to receive pre-emptive DLI, patients had to be in cytogenetic and morphological remission, to be off immunosuppression and have no evidence of active GvHD. Escalating doses of DLI were administered at 6–8 week intervals in the absence of improvement in donor chimerism (increase of <20% donor CD3).

The median age of the cohort at transplant was 59 yrs (25-71). At the time of initial diagnosis 70% of the patients had an advanced stage disease (RAEBI/II (24%) or AML(46%)). Mean donor CD3 chimerism at the time of DLI was 44% (range: 0–76%) and median time to DLI was 195 days (range: 22 to 732 days). The median number of DLI doses was 3 (range:1-5) with a median total CD3 dose of 4.9×106̂ CD3/kg (range: 5×105̂ to 4.15×108̂). The majority of patients (83%) achieved full donor chimerism (FDC, >95%) at a median time of 193 days following the start of DLI (range: 47–732 days).

The incidence of GVHD was 54%, however the incidence of severe (acute grade III,IV or chronic extensive) GvHD was limited at 21%. Skin was the commonest site of involvement with GvHD. The number of DLI doses, the total DLI dose infused, recipient age, donor source and disease stage all had no association with the development of GvHD or attainment of FDC post DLI. At last follow-up, 10 patients had died due to disease relapse(4), PTLD(1), GvHD/CMV(1), multi-organ failure(1), cerebral haemorrhage(1), infection(1) and encephalopathy(1). Evidence of relapsed disease was seen in 7 patients, and 3 patients proceeded to a second RIC HSCT.

The 2-yr actuarial OS, TRM and DFS was 81%+/−6%, 14%+/−6%, and 68%+/−8% respectively. Encouragingly, the median OS for survivors post-DLI was 2537 days (range: 805–3287 days). Neither donor source, disease stage nor the presence of GvHD post-DLI had any impact on TRM, relapse or OS. Patients who achieved FDC post DLI had a 2-yr OS of 84%+/−7% when compared with those who failed to achieve FDC: 60%+/−21%, p=0.23. In conclusion, the use of pre-emptive DLI HSCT is associated with a low incidence of severe GvHD and a low TRM and relapse. In combination with an alemtuzumab-based RIC HSCT, this approach facilitates the attainment of durable long-term remission in patients whilst reducing the procedure related toxicity.