Pembrolizumab With R-CHOP in Previously Untreated DLBCL: Sustained, High Efficacy, and Safety With Long-Term Follow-Up.

BACKGROUND: While generally ineffective in relapsed diffuse large B cell lymphoma (DLBCL), immune checkpoint inhibitors (ICIs) may hold greater promise in untreated, immunocompetent patients. We previously reported safety and early efficacy of pembrolizumab plus rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (PR-CHOP) in a phase I trial of untreated DLBCL, noting responses in 90% of patients (complete response 77%) and a 2-year progression-free survival (PFS) of 83%. We herein report long-term safety and efficacy at 5-year follow up. PATIENTS AND METHODS: Adult patients with untreated DLBCL or grade 3b follicular lymphoma, intended to receive 6 cycles of R-CHOP were eligible. Patients (N = 30) were treated with pembrolizumab 200 mg IV and R-CHOP in 21-day cycles for 6 cycles. RESULTS: At median follow up of 4.8 years, 5-year PFS was 71% (CI, 54%-94%) and 5-year overall survival was 83% (CI, 71%-98%). Immune-related adverse events (IRAEs) occurred in 7 (23%) patients (10% grade 3/4). Three IRAEs (rash, thyroiditis, rheumatoid arthritis) occurred beyond 3 months of treatment completion. PD-L1 tumor expression was documented in 19 of 23 (83%) tested patients. None of the 19 patients who had any PD-L1 expression have relapsed, whereas 2 out of the 4 patients with no PD-L1 expression have relapsed. CONCLUSION: PR-CHOP has led to durable responses in most patients, with the best outcomes in PD-L1-expressing disease. Furthermore, the safety profile was manageable, with no consistent pattern of late events. These data support ongoing strategies incorporating ICIs in frontline DLBCL therapy and confirmation of predictive biomarkers including tumor PD-L1 expression.

Pembrolizumab with R-CHOP in previously untreated diffuse large B-cell lymphoma: potential for biomarker driven therapy.

Tumor programmed death-ligand 1 (PD-L1) expression in diffuse large B-cell lymphoma (DLBCL) is associated with inferior outcomes. The first-line immunologically-replete setting may be an opportune time for PD-1 inhibition. We evaluated pembrolizumab in combination with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) in untreated patients with DLBCL. Eligible patients were age 18 or older, had adequate organ function, and had DLBCL requiring full-course therapy. Patients received pembrolizumab 200 mg/cycle with R-CHOP, primarily to assess toxicity. Response assessment utilized standard criteria, and PD-L1 staining was performed at a validated central laboratory. Among 30 patients, toxicity was comparable to standard R-CHOP but with two grade ≥3 immune related adverse events (rash, pneumonitis). The overall and complete response rate was 90% and 77%. With 25·5 months of median follow-up, 2-year progression-free survival (PFS) is 83%. PD-L1 expression was associated with non-GCB subtype, and improved PFS and survival. Pembrolizumab can safely be added to R-CHOP, and is associated with a high CR rate and 2-year PFS. Improved PFS with PR-CHOP in PD-L1 expressing tumors contradicts historical data in R-CHOP treated patients, supporting evaluation of PD-L1 as a biomarker to identify DLBCL patients who may benefit from this first-line strategy.