Reporting of older subgroups in registration trials of FDA and EMA approvals between 2017 and 2023 for diffuse large B-cell lymphoma: A systematic review

Abstract Introduction: Diffuse large B-cell lymphoma (DLBCL) is primarily a disease of older adults (OA). The median age at diagnosis is in the mid-sixties, and 55% of patients (pts) are >=65 years (yrs) at time of diagnosis. The United States Food and Drug Administration (FDA) has recommended in its guideline document for adequate representation of OA in clinical trials to maintain the generalizability of results. In addition to increasing participation, detailed subgroup reporting of the OA cohort enrolled to registrational trials is critical to contextualise efficacy and toxicities in this subgroup. In this systematic review, we analyse subgroup reporting of baseline pt characteristics, efficacy, toxicities, and quality of life (QoL) of OA in contemporary FDA and European Medicines Agency (EMA) registration trials between 2014 to 2023, leading to approval between 2017 and 2023. Methods: FDA and EMA approvals for DLBCL between 2017 and 2023 were obtained from publicly available FDA and EMA resources. Literature searches, according to trial name, National Clinical Trial registration number, and drug name, were conducted in Medline ALL (Ovid), Embase, Scopus, Web of Science Core Collection, Cochrane CENTRAL, and CINAHL, for all publications relating to included trials. Study efficacy endpoints were extracted from ClinicalTrials.gov registry data, published endpoints, and the most recent study protocol/statistical data analysis plan. Primary and secondary publications reporting on included studies were identified through a two-step screening process within Covidence. The availability and completeness of older adult subgroup data relating to protocol-defined efficacy endpoints, as well as data relating to baseline pt characteristics, health-related quality of life outcomes, and toxicity outcomes, were assessed according to predefined criteria. Results: Between 2017 and 2023, the FDA and EMA approved 10 therapies with 12 indications for DLBCL, based on 12 registration trials (2017-ZUMA-1; 2018-JULIET; 2019-GO29365; 2020-L-MIND, SADAL; 2021- TRANSCEND, LOTIS-2; 2022-ZUMA-7, TRANSFORM; 2023-POLARIX, EPCORE NHL-1, NP30179). 11 trials were in the setting of relapsed/refractory disease. 1 trial, POLARIX, was in the first-line setting. 5 trials assessed CAR-T cell therapies, 3 antibody-drug-conjugates, 2 bispecific antibodies, 1 monoclonal antibody, and 1 a small molecule. A total of 7369 results were retrieved through a literature search of registrational trials, and 92 were included for data extraction. 12 primary endpoints (ORR, CRR, EFS, PFS) and 66 secondary endpoints (OS, PFS, RFS, ORR, CRR, PRR, DOR, BOR, DCR, 2-year EFS, 2-year PFS, 2-year DFS) were assessed. Of the 12 primary endpoints, 10 (83.4%) were reported completely, 1 (8.3%) partially, and 1 (8.3%) was not reported among older subgroups. Of the 66 secondary endpoints, 25/66 (37.9%) were completely reported, 6/66 (9.1%) were partially reported, and 35/66 (53.0%) were unreported. A dedicated secondary report on OA was available for five trials (ZUMA-1, SADAL, LOTIS-2, ZUMA-7, and POLARIX). The existence of dedicated reports correlated significantly with the completeness of reporting (for primary and secondary endpoints) (p=0.03). Baseline characteristics were completely reported in 4 (ZUMA-1, ZUMA-7, LOTIS-2, POLARIX) (33.3%) and partially reported in 1 (SADAL) (8.4%) trial. 7 of 12 (58.3%) trials did not report basic baseline characteristics in older subgroups. QoL using a validated instrument was assessed in 8/12 (66.7%) trials. QoL was completely reported in 1/8 (ZUMA-7) (12.6%) trial, partially in 2/8 (LOTIS-2, POLARIX) (25.0%) trials, and was unreported in 5/8 (TRANSCEND, JULIET, TRANSFORM, SADAL, EPCORE NHL-1) (62.5%) trials. None of the trials completely reported toxicities in older OA, but 6 (ZUMA-1, ZUMA-7, TRANSCEND, SADAL, LOTIS-2, POLARIX) (50%) reported partially, whereas the remaining 6 (50%) trials left toxicities unreported in OA. Conclusion: Our analysis highlights a need for improvement in OA subgroup reporting in DLBCL trials, particularly among secondary endpoints. Standardization of subpopulation analysis in OA and dedicated trials and abstracts reporting the data will better enable a benefit-risk assessment of new agents for OA with DLBCL.