Real World Evidence of Second Line Therapy in Relapsed/Refractory Large B Cell Lymphoma: Spanish Multicenter Geltamo Study

Introduction Salvage treatment for patients (pts) with relapsed/refractory (R/R) large B-cell lymphoma (LBCL) is challenging. High-dose therapy and autologous stem cell transplantation (ASCT) was the choice for eligible pts. For those not candidates there was not a standard second-line (2L) and the available options were not considered curative although recent targeted therapies (TT) have demonstrated durable remissions. CAR-T therapy has been recently approved in 2L for pts with primary refractory (PrR) or early relapse. The aim of this study was to analyze outcomes of different 2L approaches in candidates and not candidates for ASCT. Patients and methods We performed a retrospective multicenter study based on GELTAMO registry. Pts with R/R LBCL who had received at least 1 line of chemoimmunotherapy (CIT) from 2013-2021 were included. Diffuse LBCL NOS, high-grade BCL double/triple hit and NOS, primary mediastinal and transformed follicular lymphoma were included. 2L regimens were grouped into intensive CIT, non-intensive, TT±CIT (including Tafasitamab-Lenalidomide, R-Bendamustine-Polatuzumab and clinical trials), CIT for CNS and palliative treatment (PT). The primary endpoints were relapse rate (RR), progression-free survival (PFS) and overall survival (OS) after 2L in all series and depending on patient eligibility for ASCT. Results Five-hundred and twenty-one pts fulfilled the inclusion criteria. Overall and complete RR was 45% and 33%, respectively. After a median (m) follow-up of 57 months (95%CI 52-62), 32% of the pts were alive and 79% of them free of disease. mPFS and OS were 4.9 (95%CI 3.9-6) and 13.7 (95%CI 10.9-16.4) months, respectively. Two-hundred and eighty-seven (55%) pts were considered candidates for ASCT [mage 56 years (IQR 47-64), ECOG 0-1 84%]; 65% received intensive CIT, 15% TT±CIT and 10% CIT for CNS; 38% of all the candidates finally proceeded to ASCT. mPFS and OS were 5.5 (95%CI 3.9-7.1) and 19.7 (95%CI 14.7-24.8) months, respectively. Besides, 231 (45%) pts were considered not candidates [mage 77 (IQR 72-81), ECOG 0-1 63%]; 40% received non-intensive CIT, 26% TT±CIT, 18% PT. mPFS and OS were 3.7 (95%CI 2.4-5.1) and 9.1 (95%CI 7.1-11) months, respectively. In the ASCT candidate´ subgroup, PFS was influenced by lymphoma subtype, ECOG, LDH, bulky disease (BD), status at 2L and proceeding to ASCT (p<0.01). In the multivariate analysis PrR vs late relapse [HR 1.8 (95%IC 1-3.2), p=0.037] and early vs late relapse [HR 1.8 (95%IC 1-3.4), p=0.047] were the only variables associated with worse PFS and proceeding to ASCT [HR 0.2 (95%IC 0.1-0.3), p<0.001] with higher PFS. OS was influenced by lymphoma subtype, ECOG, LDH, Ann Arbor (AA) stage, BD, extranodal sites, CNS involvement (i), status at 2L, time period, type of 2L and proceeding to ASCT (p<0.01). In the multivariate analysis, ECOG 2-4 [HR 1.8 (95%IC 1.2-2.9), p=0.008], high LDH [HR 1.7 (95%IC 1.1-2.5), p=0.013], III-IV AA stage [HR 1.6 (95%IC 1.1-2.5), p=0.016] and CNSi [HR 2.9 (95%IC 1.6-5), p<0.001] were associated with worse OS and proceeding to ASCT [HR 0.2 (95%IC 0.1-0.3), p<0.001] with better OS. Regarding not candidates, PFS was influenced by lymphoma subtype, ECOG, LDH, BD, status at 2L and type of 2L (p<0.01). In the multivariate analysis PrR disease [HR 1.8 (95%IC 1.2-32.8), p=0.004], high LDH [HR 2.2 (95%IC 1.5-3.3), p<0.001], BD [HR 1.6 (95%IC 1-2.4), p=0.043] and PT [HR 4.3 (95%IC1.7-10.9), p=0.002] were the only independent variables for PFS. OS was influenced by lymphoma subtype, ECOG, AA stage, LDH, BD, extranodal sites, CNSi, status 2L and type of 2L (p<0.01). In the multivariate analysis PrR disease [HR 2 (95%IC 1.3-3), p=0.002], ECOG 2-4 [HR 2 (95%IC 1.4-2.8), p<0.001], high LDH [HR 1.8 (95%IC 1.2-2.6), p=0.004], >1 extranodal sites [HR 1.5 (95%IC 1.1-2.2), p=0.017] and CNSi [HR 2.1 (95%IC 1.2-3.6), p=0.009] were the only independent variables for OS. Conclusions In our series, PrR disease is the variable that was associated with worse PFS for both cohorts. For candidates, response to 2L for proceeding to ASCT was the strongest variable for better PFS and OS. We also confirmed the worst outcome of pts not candidates for ASCT. Currently, with CAR-T therapy approved in 2L for eligible pts, for both candidates and not candidates, these results will probably improve, as well as for ineligible pts with the increasing use of TT. A real-world comparison of this cohort with a series of pts receiving CAR-T in 2L in Spain is planned.