Retrospective analysis of R-CDOP efficacy and safety in frail patients with DLBCL compared to R-CHOP in fit patients

Abstract Introduction: R-CHOP (Rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone) remains a frontline treatment for DLBCL. However, toxicity in older patients remains a challenge. Pegylated liposomal doxorubicin (PLD) compared with doxorubicin has demonstrated a more favorable hematologic toxicity profile in breast cancer treatment. Further data is needed to better establish efficacy and relative safety for PLD substitution in the R-CHOP regimen (R-CDOP). Patients and Methods: A retrospective review was conducted on patients with DLBCL not otherwise specified (NOS) who received treatment with either R-CHOP or R-CDOP at Stony Brook University Hospital between 2014 and 2025. Patients were identified by ICD-10 code C83.3 for DLBCL via TriNetX query. Three independent reviewers manually verified records and excluded cases other than DLBCL NOS such as DLBCL transformed from indolent lymphoma and high-grade B cell lymphoma or other entities. Patient demographics, response rate, overall survival (OS), disease free survival (DFS; events defined as radiographic progression; deaths without confirmed progression were censored), cell counts of baseline, start of cycle, and nadir counts were assessed. Significance testing was performed using the log-rank test for OS and DFS, two tailed t-test for quantitative variables, and Fisher's exact test for categorical variables without adjustment for multiple comparison. Results: Of 450 records screened, 18 patients treated with R-CHOP and 10 with R-CDOP were included for analysis. Patients in the R-CDOP group were older compared to the R-CHOP group (75.4 years vs 57.4 years at diagnosis; p < 0.01), and more likely to have non-GCB subtype (70% versus 44%, p = 0.03). The R-CDOP group had lower baseline ejection fraction (EF) than the R-CHOP group (59% versus 63%, p = 0.04). There was no significant difference in gender, IPI score, LDH, or BMI between the two populations. Each patient in the R-CHOP group completed 6 cycles and 8 in the R-CDOP group completed 6 cycles, one 5 cycles and one 4 cycles. Median OS and DFS were not reached in either group. 5-year DFS was similar (75% for R-CDOP vs 88% for R-CHOP) while OS was significantly lower in the R-CDOP group (5-year OS: 53% vs 94% for R-CHOP; p = 0.01). Rates of any-grade observed adverse events were comparable between R-CDOP and R-CHOP: anemia (100% vs 94%), thrombocytopenia (60% vs 56%), neutropenia (80% vs 72%), and heart failure exacerbation (10% vs 0%). Grade ≥3 toxicities were also similar: anemia (50% vs 22%), thrombocytopenia (40% vs 11%), neutropenia (60% vs 56%), and febrile neutropenia (20% vs 11%). In R-CDOP, neutrophil count declined from 9.67 to 4.14 k/μL by cycle 6 (p = 0.01), with no significant changes in hemoglobin, platelets, or nadir counts between cycles 1 and 6. Conclusions: Patients in the R-CDOP group demonstrated greater age, lower EF, and more aggressive disease subtype than those in the R-CHOP group. There was no evidence of increased hematopoietic or cardiac adverse events in the R-CDOP group compared to the R-CHOP group. DFS was comparable while the OS was lower in the R-CDOP group, suggesting death from non-relapse etiologies.