Real-world evidence of improved survival with pola-based regimen in newly diagnosed DLBCL patients with extranodal involvement

Abstract

Abstract Background: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease, with extranodal (EN) involvement in nearly 50% of newly diagnosed cases, often linked to higher tumor burden and poorer outcomes. The phase III POLARIX trial established Pola-R-CHP as a new first-line standard, improving progression-free survival (PFS) over R-CHOP. However, subgroup analysis showed limited benefit in EN patients. Real-world evidence is needed to further evaluate Pola-containing regimens in this high-risk group. We conducted a study to assess its efficacy and safety in newly diagnosed DLBCL patients with EN involvement in routine clinical practice. Methods: In this study, eligible patients were newly diagnosed with DLBCL, had at least one site of EN involvement confirmed by PET/CT, and received at least one cycle of first-line Pola-containing regimens therapy from April 2023 and May 2025. The main clinical characteristics, short-term efficacy, survival status, and adverse reactions in this population are analyzed. Results: A total of 126 treatment-naïve DLBCL patients receiving Pola-containing regimens were included. Among them, 96 (76.2%) had at least one EN site involved. The median age was 60.5 years (range, 14–88); 73 (57.9%) were <65 years, 53 (42.1%) were ≥65 years, and 13 (10.3%) were ≥80 years. The cohort comprised 53 males and 73 females. Most patients (n=97, 77.0%) presented with Ann Arbor stage III–IV disease. IPI scores were 0–1 in 17 (13.5%), 2–3 in 73 (57.9%), and 4–5 in 34 (27.0%) patients.The most frequently involved EN sites included the gastrointestinal tract (n=27, 28.1%), bone (n=27, 28.1%), liver (n=18, 18.8%), and lung (n=14, 14.6%). Among patients with EN disease, 42 (43.8%) had 1 site involved, 31 (32.3%) had 2 sites, and 23 (24.0%) had ≥3 sites.Regarding treatment, 95 patients received Pola-R-CHP, 25 received Pola-R-miniCHP, and 6 received other regimens (Pola-R plus BTKi, or Pola-R-EPCH). Among the 81 response-evaluable patients, 13 without EN disease achieved a CR in 11 (84.6%) and a PR in 2 (15.4%). Among the 68 response-evaluable EN patients, CR and PR were observed in 49 (72.1%) and 11 (16.2%) patients, respectively, with 8 (11.8%) having stable or progressive disease.After excluding patients who had not completed treatment, 93 patients were included in the survival analysis. At a median follow-up of 12.9 months, median PFS and OS were not reached. The estimated 12-month PFS and OS rates in the overall cohort were 71.1% (±5.0%) and 86.2% (±3.9%), respectively. Among patients with EN involvement, 12-month PFS and OS rates were 67.0% (±5.7%) and 85.0% (±4.4%), respectively. When stratified by number of EN involvement sites, the 12-month PFS rates were 71.1% (±8.2%) for patients with 1 site, 73.5% (±9.4%) for 2 sites, and 62.5% (±12.1%) for ≥3 sites. Corresponding 12-month OS rates were 85.7% (±6.7%), 87.0% (±7.0%), and 80.8% (±10.0%), respectively. Pola-based regimens were generally well tolerated. The most common grade 3–4 AEs were hematologic toxicities, including neutropenia (31.0%), febrile neutropenia (8.7%), anemia (14.3%), and thrombocytopenia (9.5%). The infection rate was 9.5%. Conclusions: In this real-world cohort of newly diagnosed DLBCL patients, Pola-containing regimens demonstrated promising efficacy and manageable safety among those with extranodal involvement. A high complete response rate was observed. Although the 12-month PFS was modest, it translated into an OS benefit, even in patients with advanced-stage or multi-site EN disease. The trend toward decreasing survival with increasing number of EN sites highlights the need for early identification and tailored management in this high-risk subgroup. Pola-based regimens may offer a viable frontline treatment option for these patients in routine clinical practice.