Split dose R-CHOP mitigates toxicity while preserving dose intensity and response in frail elderly patients with diffuse large B cell lymphoma (DLBCL) - a Wisconsin oncology network (WON) study

Abstract Introduction Management of DLBCL in older adults (OA) is challenging, requiring a personalized approach that accounts for age-related comorbidities and frailty. Many OA receive non-anthracycline based treatment (tx), or attenuated regimens (i.e. R-miniCHOP) to avoid toxicity. However, DLBCL remains the leading cause of death for OA treated with these regimens. Torka et al (ASH 2024) defined a population with DLBCL at increased risk of developing severe toxicities (STox) from tx. Predictors of STox included age >75, ADL and IADL dependence, poor nutritional status, patient rated Karnofsky performance status (prKPS) ≤80 and timed up and go (TUG) time of ≥12 secs. SD R-CHOP is an alternative approach where pts receive CHOP at 50% dose reduction on days 1 and 15 of each 28-day cycle (C) with R on Day 1 for 6 cycles. The cumulative dose of one cycle of SD R-CHOP is equal to one cycle of R-CHOP-21. Interim data was previously presented at ASCO 2023. We present updated results including data from planned geriatric assessments (GA). Methods A Phase II, prospective, open label trial of SD-R-CHOP was conducted in frail OA through the WON. Eligible pts had untreated de-novo or transformed (t) DLBCL or high-grade B cell lymphoma NOS stage IIX-IV with KPS ≥ 50 and were either age ≥ 75 years old or 70-74 years with a Cumulative Illness Rating scale for Geriatrics (CIRS-G) of at least 10 or ≥1 comorbidity scored 3-4. Pts with CNS involvement were excluded. Impaired organ function was permitted as long as R-CHOP agents were not contraindicated. Pre-phase steroids and intrathecal CNS prophylaxis were allowed but not required. G-CSF support was mandatory. Tx consisted of two cycles of SD R-CHOP followed by interim response assessment with PET/CT and peripheral blood clonoSEQ MRD testing. Pts with interim PET/CT score of Deauville 1-3 and undetectable (u) MRD were offered abbreviated therapy with two additional cycles of SD-R-CHOP. All others completed 6 cycles of tx. Dose reductions for cytopenias were not permitted. Subsequent dose was delayed until tx parameters were met, up to 28 days. Pts were evaluated prospectively using the Cancer and Aging Research Group (CARG) GA at baseline, interim disease assessment, and end of tx (EOT). The primary endpoint was EOT complete response (CR) rate. A Simon 2-stage design was used with a target of 16 of 26 pts (>60%) achieving an EOT CR to consider this regimen successful. Results 26 pts received at least one dose of SD-R-CHOP. The median age was 80 years (71-88) and most had stage III/IV disease (85%). IPI was ≥ 3 for all, and 6 pts (23%) had tDLBCL. 20 pts (77%) had KPS ≤80, and 14 (56%) had prKPS ≤80. At baseline 12 of 26 (48%) had impaired IADLs, 5/25 (20%) had impaired ADL and TUG was ≥12 secs for 20 (77%) pts, with a median TUG of 20 secs (IQR 12-75). 7 pts were unable to perform baseline TUG. The EOT CR rate for all pts was 65% (n=17) exceeding target threshold. Nine pts (35%) were both uMRD/PET- at interim assessment after C2; of these, 8 completed abbreviated tx, while one opted for a full 6 cycles. Eight of these 9 pts (89%) achieved an EOT CR. 17 pts were assigned to the full tx arm (6 cycles) due to interim MRD+ or PET+ and in this group, the EOT CR rate was 53%. At 18-months, overall survival (OS) was 67% and progression free survival (PFS) was 58%. Neither median OS nor PFS were reached to date. Only one pt with interim uMRD/PET- has relapsed to date despite 8 of 9 receiving abbreviated therapy. Grade 4+ toxicity occurred in 13 pts (50%) and SAE in 9 (35%). 15 pts (58%) had a dose modification/delay. 3 pts (12%) had >1 dose delay, 4 (15%) had >1 cycle at reduced dosing. Two pts (7.6%) experienced non-relapse mortality: 1 infection, 1 cardiac event possibly related to tx. No component of the CARG GA predicted STox in this cohort. Conclusions SD-R-CHOP offers frail OA with DLBCL equivalent dose intensity as R-CHOP-21 through a fractionated dosing schedule. The EOT CR rate for all pts was 65% exceeding the Phase II threshold set for this trial, demonstrating the value of maintaining dose intensity for older pts. Despite significant comorbidity and a high prevalence of factors associated with STox, events were less common than expected in this group. De-escalation of therapy appears feasible for pts with interim uMRD/PET- response, limiting toxicity without impacting long-term outcomes. These findings suggest that SD R-CHOP warrants further study in frail OA with DLBCL at high risk for STox.