AUTHORS
Elizabeth Brem, MD, Hongli Li, MS2, Michael LeBlanc, PhD3, Rebecca L Olin, MD4, Norah Lynn Henry, MD5, Ajay Major, MD, MBA6, Jennifer Effie Amengual, MD7, Jonathan W. Friedberg, MD8, Sonali Smith, MD, FASCO9
AFFILIATIONS
1 University of California, Irvine, Orange, California, United States
2 SWOG Statistics and Data Management Center, Seattle, Washington, United States
3 SWOG Statistics and Data Management Center, Seattle, Washington, United States
4 University of San Francisco, San Francisco, California, United States
5 University of Michigan, Ann Arbor, Michigan, United States
6 University of Colorado, Denver, Colorado, United States
7 Columbia University, New York, New York, United States
8 University of Rochester, Rochester, New York, United States
9 University of Chicago, Chicago, Illinois, United States
ABSTRACT
We are in an era of unprecedented gains in the treatment of aggressive lymphomas, yet outcomes in older adults have lagged with lower cure rates and increased morbidity/mortality. One-third of DLBCL in the United States occurs in patients (pts) > 75y, but older adults are significantly underrepresented in prospective trials. Geriatric assessments are not widely used and have not been systematically implemented in clinical trial settings. This leads to many older adults with DLBCL either being undertreated, not treated, or overtreated with dire consequences. With this in mind, S1918 included prospective frailty assessment as an integrated biomarker.
S1918 (NCT04799275) was a randomized, Phase II/III study of pts > 75 years with newly diagnosed DLBCL comparing oral azacitidine (CC-486) added to R-miniCHOP (Arm 1) versus the control of R-miniCHOP (Arm 2). Overall, 136 pts (134 eligible) were accrued to the Phase II portion, 67 in each arm. Study accrual closed April 01, 2025. The median age of patients enrolled was 81.8y (Arm 1) and 81.6y (Arm 2); 69% on Arm 1 and 67% on Arm 2 were age ≥80 years. The Phase II component did not meet the pre-specified criterion for continuation to Phase III; details are reported in a separate ASH 2026 abstract.
Baseline frailty assessment and serial functional assessment were integrated biomarkers. Frailty was objectively assessed at study entry using the simplified Geriatric Assessment (sGA; Merli et al 2021). Pt functioning was assessed longitudinally using a Comprehensive Geriatric Assessment (CGA; Hurria A et al 2011) administered prior to treatment initiation, and 6, 12, and 24 months following registration. This CGA has both a patient and provider component and assesses multiple dimensions of functioning.
Frailty assessment pre-registration was performed in 97.8% of enrolled pts. Among 131 of 134 eligible pts assessed for sGA (97.8%), 19 pts (14.5%) were fit, 47 (35.9%), were unfit, 27 (20.6%) were frail, and 38 (29%) were superfrail. By study arm, 52.2% (Arm 1) and 46.9% (Arm 2) scored as frail or superfrail; frailty was not a stratification factor. The CGA assessment was readily completed prior to registration (99.2% of eligible and evaluable patients for patient and 95.5% for provider portion respectively). Pt and provider portions were readily completed at 6m (79.3% pt, 77.5% provider) and 12m (78.8%, 74.8%). The CGA was less likely to be completed at the 24-month time point (43.0%, 40.9%).
Superfrail patients had a significantly reduced PFS compared to fit, unfit, and frail patients, independent of treatment arm. 1-year PFS was 72% (95% CI 46-87%) for fit patients, 71% (95% CI 55-82%) for unfit patients, 71% (95% CI 49-85%) for frail patients, and 47% (95% CI 31-62%) for superfrail patients.
More pts had grade 4-5 toxicities on Arm 1 regardless of age. For those age 75-79 who had baseline frailty scored, eight (40%) experienced grade 4-5 events on Arm 1 vs two (11%) on Arm 2. For those 80+, 12 (28%) experienced events on Arm 1 vs nine (20%) on Arm 2. When evaluated by frailty status, toxicities were more balanced between arms. In pts scoring frail/superfrail on sGA, 10 pts (29%) experienced grade 4-5 events on Arm 1 vs 8 pts (27%) on Arm 2.
Superfrail pts disproportionately accounted for the deaths on study. On Arm 1, five superfrail pts died compared to two unfit and one frail. On Arm 2, four superfrail patients had grade 5 events (none in the other categories). Overall, non-relapse mortality (NRM) was similar between frail/superfrail and fit/unfit patients (52% [n=14] vs 54% [n=7]).
In conclusion, S1918 showed feasibility of prospective frailty assessment and serial CGA collection as part of a prospective randomized trial in an intergroup setting. High grade AEs and deaths were more likely to occur in frail/superfrail pts, independent of assigned treatment arm. One-year PFS was significantly reduced for superfrail pts compared to all other sGA categories. Increased toxicity, especially in superfrail patients, appears to drive inferior outcomes and future trials should focus on reducing toxicity. Our experience strongly supports frailty as a powerful integral biomarker and stratification factor for clinical trials in older adults, especially in this new era of targeted agents that hold great promise if safely applied.
Funding: This research was funded by NIH/NCI grants U10CA180888, U10CA180819, UG1CA189974, and in part by Celgene Corporation (now part of Bristol-Myers Squibb).
DISCLOSURES
Elizabeth Brem, MD: Abbvie/Genmab - Honoraria; AstraZeneca - Honoraria; BMS - Honoraria; Genentech - Consultancy (Includes expert testimony); Caribou Biosciences - Consultancy (Includes expert testimony); ADC Therapeutics - Consultancy (Includes expert testimony); BeOne - Consultancy (Includes expert testimony); Incyte - Consultancy (Includes expert testimony); BMS - Consultancy (Includes expert testimony); Regeneron - Consultancy (Includes expert testimony); AstraZeneca - Consultancy (Includes expert testimony); Poseida Therapeutics - Consultancy (Includes expert testimony)