Background Diffuse large B-cell lymphoma (DLBCL) is the most prevalent aggressive subtype of non-Hodgkin lymphoma, characterized by substantial clinical and molecular heterogeneity. Approximately one-third of DLBCL cases originate from extranodal sites, including the gastrointestinal tract, nasal cavity, and breast. Although R-CHOP remains the standard first-line therapy, extranodal DLBCL patients generally exhibit inferior outcomes compared to nodal DLBCL, underscoring the need for novel therapeutic strategies. Emerging evidence suggests that DLBCL with multiple extranodal sites harbors dysregulated tumor microenvironments, showing significant downregulation of T-cell differentiation-related genes and cytokine signaling pathways, indicative of potential immune escape. Additionally, the activated B-cell-like subtype may exhibit programmed death ligand 1 (PD-L1) overexpression through genetic alterations, further facilitating immune evasion. These findings rationalize investigating programmed death 1 (PD-1) blockade combination therapy in the frontline setting to enhance therapeutic efficacy in extranodal DLBCL.

Methods This multicenter, prospective, single-arm trial enrolled treatment-naïve patients aged 18-65 years with histologically confirmed primary extranodal DLBCL from three centers (NCT05093140). Patients received two 21-day cycles of immune priming with camrelizumab (200 mg) and rituximab (375 mg/m²on Day 1), followed by six 21-day cycles of standard R-CHOP, and subsequent camrelizumab maintenance monthly for 6 cycles in those who achieved complete response (CR). The primary endpoint was CR rate after camrelizumab-R-CHOP therapy per 2014 Lugano criteria. Secondary endpoints included objective response rate (ORR) post-induction and consolidation, duration of response, progression-free survival (PFS), overall survival (OS), and safety assessed via NCI CTCAE v5.0.

Results Between June 2022 and April 2025, 30 patients were enrolled (median age 51.5 years, range 26–58). Advanced-stage disease), and 26.7% (8/30) had an age-adjusted International Prognostic Index (aa-IPI) >1. Primary extranodal sites included stomach (40.0%, 12/30), nasal cavity (26.7%, 8/30), intestine (16.7%, 5/30), and breast (6.7%, 2/30); 10.0% (3/30) had >1 extranodal site. Pathology subtypes comprised germinal center B-cell-like (GCB, 26.7%, 8/30), non-GCB (43.3%, 13/30), double-expressor (20.0%, 6/30), and double-hit (10.0%, 3/30) lymphoma. Following induction with camrelizumab-rituximab, ORR was 63.3% (19/30), with CR in 26.7% (8/30). Site-specific ORRs were: gastric 75.0% (9/12; CR 41.7%), intestinal 60.0% (3/5; CR 0%), breast 100% (2/2; CR 50.0%), nasal 50.0% (4/8; CR 37.5%), and multisite 33.3% (1/3; CR 0%). Among the 28 patients evaluable for consolidation (2 patients still on treatment), camrelizumab-R-CHOP resulted in an ORR of 96.4% (27/28) and a CR rate of 89.3% (25/28). At a median follow-up of 23.9 month, the estimated 2-year PFS and OS were 86.8% (95% CI: 60.5-91.7) and 90.3% (95% CI: 69.6-96.4), respectively. The most common treatment-emergent adverse events were leukopenia (56.6%), neutropenia (53.3%), and lymphopenia (46.7%). Grade ≥3 adverse events (AEs) occurred in 63% (19/30) of patients, with neutropenia (43.3%) and leukopenia (33.3%) being the most frequent. Reactive capillary endothelial proliferation, a known camrelizumab-related AE, occurred in 11 patients, all of which were grade 1.

ConclusionCamrelizumab combined with rituximab demonstrated promising efficacy in primary extranodal DLBCL. Sequential R-CHOP therapy further improved the CR rate, with a favorable safety profile. These results support further validation of PD-1-based strategy in frontline management of extranodal DLBCL in prospective randomized studies.

This content is only available as a PDF.
2025
Sign in via your Institution