Allogeneic Stem Cell Transplantation As Salvage Therapy for First Relapse after Autografting in Multiple Myeloma Patients

Despite the significant improvement in outcomes for myeloma patients (pts) in the more recent years (yrs), the disease remains incurable. Because of high morbidity and mortality the role of allo-SCT in treatment of myeloma remains controversial. Registry data from the European Society of Blood and Marrow Transplantation (EBMT) suggest an increasing use of allo-SCT as salvage therapy after failure to auto-SCT. In present single center study we provide data on the use of allo-SCT focusing on pts with first relapse/progressive disease after autografting.

A total of 89 pts (males, 62%) with median age of 53y (24-73) relapsed after an autograft (single, 64%; tandem, 36%) and received an allo-SCT during 2000 - 2015 yrs at University of Hamburg were included. The majority of the pts experienced advanced disease (stage III, 80%) and short remission duration after auto-SCT (≤24 mo, 81%) with median of 15 mo (2-91). A total of 76% of pts received a reinduction therapy prior to allo-SCT, whereas 24% did not. Twelve pts (14%) had high risk cytogenetics (defined as detection of del17p and/or t(4;14) by FISH). Twelve pts experienced extramedullary relapses. At time of transplantation pts were in CR (10%), vgPR (22%), PR (27%), SD (16%) or PD (25%). Allo-SCTs were performed mostly with peripheral stem cell (92%) and from unrelated (matched, 40%; mismatched, 32%) donors. Myeloablative conditioning was used in 70% of pts (busulfan-based, 61%; treosulfan-based, 9%). A total of 28 pts (31%) received a maintenance therapy (lenalidomide, 79%) starting at median 6 mo (4-8) post-transplant. Furthermore, 19 pts (24%) received DLIs (positive immunofixation, n=15; prophylaxis, n=4) within a median of 10 mo (4-27).

The cumulative incidence (CI) of acute GvHD (grade II-IV) at d100 was 43% (32-55%). The CI of chronic GvHD at 5y post-transplant was 51% (39-63%; mild, n=19, moderate, n=12, severe, n=5). Of the 19 pts receiving DLIs, 5 (26%) developed GvHD (grade II-IV). Six (40%) of 15 pts responded (CR).

The CI of TRM at d100 and at 5y were 8% (4-16%) and 19% (12-28%), respectively. We observed improved TRM in pts who didn't receive radiation before allo-SCT (p=0,03). The CI of relapses at 5y was 61% (49-72%). We observed a trend to lower relapse rate of 29% in pts with longer remission duration (>24 mo, p=0,07).

The 5y probability of EFS was 28% (19-39%). The median EFS was 29 mo (22-36). In univariate analysis, we observed decreased survival probability for pts with high risk cytogenetics (p=0,026), remission duration ≤24 mo (p=0,008) and non-responders to salvage therapy (p=0,037). In multivariate analysis the negative impact of remission duration ≤24 mo (HR 6,9; 1,8-25,6; p=0,004) and failure to salvage therapy (HR 3,3; 1,2-9,0; p=0,021) were confirmed. In landmark analysis at d100 we observed improved survival for responding pts (p<0,001), pts in CR (p=0,028), those in "stringent" CR compared to CR (p=0,014) and those with negative MRD (minimal residual disease, FACS, p=0,037). Intriguingly, the median EFS for pts in "stringent" CR was not reached. In landmark analysis at 6 mo we observed improved survival for pts receiving preemptive/prophylactic DLIs (p=0,045).

After a median follow up of 48 mo (6-170), the 5y probability of OS was 57% (46-78%). The median OS was 76 mo (42-110). In univariate analysis, we observed decreased survival probability for pts with extramedullary relapse (p=0,039), high risk cytogenetics (p=0,001), remission duration ≤6 mo (p=0,044) and those who received reduced intensity conditioning (p=0,022). In multivariate analysis the negative impacts of extramedullary relapses (HR 7,3; 2,7-20; p<0,001), reduced intensity conditioning (HR 3,5; 1,4-8,5; p=0,006) and short remission duration (HR 2,6; 1,1 - 6,2; p=0,038) were confirmed. In landmark analysis at d100 we observed improved survival for responding pts (p<0,001), pts in CR (p=0,028) and those in "stringent" CR compared to CR (p=0,014). In landmark analysis at 6 mo we observed improved survival for pts receiving DLIs (p<0,001) and maintenance therapy (p=0,002).

Allo-SCT being performed as salvage therapy after autograft failure resulted in around 30% 5y EFS. The response on d100 post-transplant, augmented by the MRD detection, seems to be prognostic important and might be used for planning further therapeutic strategies post-transplant. The use of DLIs and maintenance therapy post-transplant represent a possible approach to further improve survival.