Alkylator-Based Conditioning for Allogeneic Hematopoietic Stem Cell Transplantation (alloHSCT) in Lymphoma: A Registry Study Comparing Thiotepa-, Busulfan-, Melphalan- and Treosulfan-Containing Regimens On Behalf of the EBMT Lymphoma Working Party

Thiotepa (TT) is an alkylating agent approved for conditioning for alloHSCT. TT-based alloHSCT has been pioneered in a variety of lymphoma subtypes with promising results, but the available information about the value of thiotepa in this indication compared to other alkylator-based regimens is still limited.

Primary objective was to compare the outcome of TT-based alloHSCT with that of alloHSCT conditioned with other alkylator regimens (non-TT) separately for diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), and peripheral T Cell lymphoma (PTCL). Primary endpoint was event-free survival (EFS); secondary endpoints were overall survival, non-relapse mortality (NRM), and relapse incidence. Eligible were patients >18 years who had received TT-, busulfan (BU)-, melphalan- (MEL), or treosulfan- (TREO) based conditioning for T-replete alloHSCT between 2003–2010 for FL, DLBCL, or PTCL. Statistical analysis was based on multivariable comparisons using stratified Cox and Fine & Gray regression models.

201 patients with TT fulfilled the inclusion criteria and were compared with 578 non-TT patients (BU 55%, MEL 35%, TREO 10%). The most frequently used specific regimens were TT-cyclophosphamide combinations (75%) in the TT group and BU-fludarabine combinations (52%) in the non-TT group. Of the total 779 patients, 43%% had FL, 39% DLBCL, and 18% PTCL. TT and non-TT patients were comparable for age, sex, time from diagnosis, remission status at HSCT, and proportion of unrelated donor transplants. However, the TT group contained significantly more patients with PTCL (24% vs 15%), with poor performance status (PS; 12.5% vs 3%), and with BM as HSCT source (14% vs 9%). By multivariate comparisons considering conditioning, age, sex, remission status, PS, and time from diagnosis, EFS was significantly affected by active disease at HSCT and poor PS in all three lymphoma subtypes. In contrast, conditioning with TT had no significant impact (Hazard ratio (HR) 1.06 (0.67–1.67) for FL; 1.01 (0.67–1.51) for DLBCL; 1.33 (0.75–2.36 for PTCL) on EFS or any other endpoint. Similar results were seen when the analysis was broken down to specific conditioning regimens (TT-CY vs BU-based). MEL- and TREO-based regimens did not show a significant difference compared to BU for any endpoint.

This study failed to identify significant outcome differences between the four conditioning regimen types tested. However, the limitations inherent to registry analyses have to be considered. In particular, conclusions on differential regimen toxicity apart from NRM will require additional, ideally prospective studies.

Dreger:Riemser G, Greifswald, Germany: Consultancy, Honoraria, Research Funding. Off Label Use: Treosulfan for conditioning for allogeneic HSCT. Schmitz:Riemser AG, Greifswald, Germany: Honoraria.