Management and Outcome of Patients with Diffuse Large B-Cell Lymphoma Relapsed after Autologous Hemopoietic Stem Cell Transplantation (auto-HSCT): A Retrospective Analysis of the Lymphoma Working Party (LWP) of the European Society for Blood and Marrow Transplantation (EBMT)

Introduction: The standard therapy for relapsed/refractory diffuse large B-cell lymphoma (DLBCL) patients is salvage immuno-chemotherapy followed by high-dose therapy and auto-HSCT, but 50%-70% of them will have recurrence of the disease. There is no consensus on further therapy at this point. The present study is intended to analyze management and outcome of patients with DLBCL relapsing after auto-HSCT and reported to the EBMT database.

Patients and Methods: Adult patients (age ≥ 18 years) diagnosed with DLBCL who underwent a first auto-HSCT between Jan/2003 and Dec/2013 and who relapsed after the auto-HSCT, were included in the analysis. Five hundred and forty-three patients (pts) were identified, but therapy after auto-HSCT relapse was not reported in 164 pts. This group of pts was more heavily pretreated before auto-HSCT [3 or more lines in 65 (40%); p<0.0001], and had a shorter time from auto-HSCT to relapse [median: 3.5 months (mo); p=0.002] than the 379 pts with known therapy. This study focuses in the group of 379 pts with known therapy after first auto-HSCT relapse. Overall survival (OS) was calculated from the initiation of therapy after auto-HSCT relapse.

Results: 229 (60%) pts were male, with a median age at auto-HSCT of 54 years (range: 19-75); at relapse: 34% had an R-IPI 3-5, 28% had B symptoms, 55% raised LDH, 33% bulky disease ≥5 cm and 76% Karnosfsky score above 80. Auto-HSCT was performed after one line of chemotherapy in 91 pts (24%), after two lines in 203 pts (54%), and after three or more lines in 85 pts (22%). Median time from first auto-HSCT to relapse was 5.7 mo (range: 0.4-75), 113 pts (30%) relapsed more than 1 year after auto-HSCT. The most frequent therapy combinations at relapse were: platinum containing regimens in 134 pts (35%), other active combinations (doxorubicin, cytosine arabinoside, ifosfamide, gemcitabine) in 97 pts (26%), lenalidomide / bendamustine containing regimens in 26 pts (7%), palliative (steroids, single agent, radiotherapy alone) in 77 pts (20%), other treatments in 45 pts (12%). Response after therapy was complete remission (CR) in 86 pts (23%), partial remission (PR) in 42 pts (11%), stable/progressive disease (SD/PD) in 171 pts (45%), unknown in 73 pts (19%) and death without response assessment in 7 pts (2%). Ninety-six (25%) pts underwent a second HSCT [allogeneic HSCT (allo-HSCT) in 82 patients]; 37 of them were in CR, 16 in PR, 39 with active disease, and in 4 status was unknown. Median follow-up of alive pts was of 42 mo (range: 26-65), 288 pts (76%) died: 239 (63%) of disease progression, 32 (8.4%) due to transplant related mortality, 5 (1.3%) of secondary malignancies, and 12 from other causes. OS at 36 mos for the whole cohort was 26% (95% CI 21-31). Results were significantly more favorable in pts who relapsed more than 1 year after auto-HSCT, with an OS at 36 mo of 43% (95% CI 34-53) compared with those who relapsed less than 1 year after auto-HSCT, with an OS at 36 mo of 18% (95% CI 14-24) (p<0.001). OS after a second allo-HSCT at 36 mos was 37% (95% CI 27-51).

Conclusions: In our series, 76% of pts who received therapy for relapse after auto-HSCT maintained a good performance status, in spite of the fact that relapse after auto-HSCT was an early event (median time from auto-HSCT to relapse, less than six months). Around 35% of the pts who received therapy responded, and 25% underwent a second HSCT, mainly allo-HSCT. The outcome of pts who relapsed longer than 1 year after auto-HSCT was relatively good, with a 36-mo OS of 43%, supporting consideration of active, curative approaches for these pts at relapse. In contrast, pts relapsing less than 1 year after auto-HSCT have significantly worse outcomes. These patients remain an unmet clinical need, and novel therapies and strategies are required.