Sequential R-CHOP/(R)-ICE and Dose-Adjusted EPOCH-R Are Both Appropriate Frontline Treatments for Newly Diagnosed Primary Mediastinal B-Cell Lymphoma: Results of a Retrospective Analysis

BACKGROUND Primary mediastinal B-cell lymphoma (PMBL) is a rare subtype of diffuse large B-cell lymphoma associated with high cure rates and historically treated with intensive combined modality approaches. Given the disease's increased frequency in women and median age in the 30s, the optimum regimen would maintain high cure rates and not rely on radiation therapy (RT). Dose-adjusted (DA) EPOCH-R (rituximab, etoposide, prednisone, vincristine, cyclophosphamide, and doxorubicin) has been reported to have excellent outcomes in PMBL without planned RT. At MSKCC, we have investigated sequential R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) induction with (R)-ICE (rituximab, ifosfamide, carboplatin, and etoposide) consolidation without RT in two sequential clinical trials and as a subsequent standard treatment plan. The current retrospective analysis compared the efficacy and long-term outcomes of R-CHOP/(R)-ICE and DA-EPOCH-R for the treatment of newly diagnosed PMBL at MSKCC. PATIENTS AND METHODS We retrospectively identified patients (pts) with newly diagnosed PMBL at MSKCC between January 2002 to January 2022 treated up-front with either R-CHOP/(R)-ICE or DA-EPOCH-R. Pt demographics and clinical data were abstracted from institutional electronic medical records. The analysis included 226 pts: 111 received R-CHOP/(R)-ICE and 115 received DA-EPOCH-R. Progression-free survival (PFS) was calculated from start of treatment until date of progression, death, or censored at date of last follow-up. Overall survival (OS) was calculated from date of diagnosis till death or censored at date of last follow-up. Outcomes were analyzed for the entire cohort as well as in a matched subset: pts who received DA-EPOCH-R were matched to pts who received R-CHOP/(R)-ICE. A 1:1 propensity score matching on sex and R-IPI via the nearest neighbor matching (NNM) approach with a 0.2 caliper cutoff was attempted on 111 pts. Only a subset of 88 pts were matched within the specified caliper distance. Statistical analyses were done using SAS 9.4. RESULTS The 111 R-CHOP/(R)-ICE pts had longer follow-up (median, 7.2 years; 9.5% CI, 6.4 - 8.4), fewer females (48%), and inferior R-IPI prognosis (11% Very Good, 60% Good, 29% Poor) compared to the 115 DA-EPOCH-R pts: median follow-up 4.3 years (9.5% CI, 3.9 - 5.0); 60% female; R-IPI (18% Very Good, 71% Good, 10% Poor). RT was not planned post first-line treatment, with very low utilization: 1% after R-CHOP/(R)-ICE and 3% after DA-EPOCH-R, with no difference between the 2 treatment groups (p = 0.37). For the entire cohort at a median follow-up of 5.5 years (95% CI: 5.3 - 5.8), there was no difference in PFS (HR, 0.79; 95% CI, 0.38 - 1.64; p = 0.53) and OS (HR, 0.76; 95% CI, 0.25 - 2.32; p = 0.63) between the 2 treatment groups. The estimated 5-year PFS and OS for the entire cohort were 86% and 94%, respectively, for R-CHOP/(R)-ICE and 89% and 96%, respectively, for DA-EPOCH-R. The 1:1 propensity score matching was applied to validate this observation. In the propensity score-matched subset (88 pts treated with R-CHOP/(R)-ICE vs 88 pts treated with DA-EPOCH-R), there remained no difference observed between the 2 groups in terms of PFS (HR, 0.57; 95% CI, 0.23 - 1.42; p = 0.23) and OS (HR, 0.53; 95% CI, 0.14 - 2.09; p = 0.37). The estimated 5-year PFS and OS for the matched cohort were 86% and 93%, respectively, for R-CHOP/(R)-ICE and 92% and 98%, respectively, for DA-EPOCH-R. DISCUSSION Both R-CHOP/(R)-ICE and DA-EPOCH-R result in similar highly favorable outcomes in pts with newly diagnosed PMBL, with excellent PFS and OS without the need for RT. In addition, the OS outcomes in our series suggest an encouragingly robust response to subsequent therapy in the relapse setting for PMBL (Vardhana, 2018). Clinical considerations often impact treatment choice. DA-EPOCH-R requires continuous infusion via central access, possible inpatient treatment depending on outpatient treatment constraints, and potential hospitalizations, whereas R-CHOP/(R)-ICE requires fewer treatment days (12 vs 30), peripheral IV access, and only 2-3 inpatient days for 3 of the cycles. Given the benefits of reduced hospitalization and comparable outcomes with R-CHOP/(R)-ICE, our results suggest this regimen may be an appropriate alternative to the widely utilized DA-EPOCH-R in the frontline treatment of PMBL. The optimal regimen will likely depend on local resource utilization.