亚洲多种族人群中一线 R-CHOP（类）方案治疗弥漫大 B 细胞淋巴瘤的真实世界结局

Recent breakthrough advances in the treatment of DLBCL, such as the antibody-drug conjugate polatuzumab vedotin, have yielded clinical survival benefit over rituximab, cyclophosphamide, doxorubicin, vincristine and prednisolone (R-CHOP) for the first time in 20 years since the advent of the rituximab era. We thus examine the outcomes of standard immunochemotherapy for DLBCL in our multi-ethnic Asian population, so as to determine the real-world clinical need to adopt new therapeutics in this disease entity.

We conducted a retrospective study involving patients (n = 1071) diagnosed with DLBCL at the National Cancer Centre Singapore from 2010 to 2022, and treated with first-line rituximab-based regimens. The median follow-up duration was 48 months. Survival analyses were performed using the Kaplan-Meier method and multivariate Cox proportional models.

The cohort consisted of 590 male and 481 female patients with a median age of 63.8 years (range, 19.3-93.6). Most were stage III-IV at diagnosis (60.9%) and of non-germinal center B-cell like (non-GCB) subtype by Han's criteria (56.5%). The vast majority received R-CHOP(-like) regimens (n = 997, 93.1%), including rituximab, etoposide, prednisone, vincristine, cyclophosphamide and doxorubicin (EPOCH-R) (n = 95), achieving a 5-year progression-free survival (PFS) and overall survival (OS) of 64.5% and 74.7% respectively. Male sex (p = 0.0294), age > 60 years (p < 0.0001), poor ECOG scores (2-4) (p < 0.0001), advanced stage (III-IV) (p < 0.0001), presence of B-symptoms (p = 0.0305), and raised LDH (p = 0.0161) were independent predictors of OS, 4 of which are risk factors in the International Prognostic Index (IPI). In the intermediate to high-risk subgroup (IPI scores 2-5; n = 752), the 5-year PFS and OS were only 59.0% and 69.8% respectively. EBV status, MYC and/or BCL2/BCL6 rearrangements, were not significantly associated with survival outcomes. EPOCH-R was used more frequently than R-CHOP in patients with MYC rearrangements (n = 82, p < 0.0001), including those with MYC/BCL2 double-hit genetics (n = 31, p < 0.0001). Notably, neither regimen significantly affected survival outcomes, both in MYC-rearranged (PFS: HR 0.60, p = 0.1704; OS: HR 0.49, p = 0.0852), and in MYC/BCL2 double-hit DLBCL (PFS: HR 1.30, p = 0.6433; OS: HR 1.02, p = 0.9803).

Our study demonstrates that our local population has similar clinicopathological and prognostic characteristics of DLBCL as compared to global findings. It also highlights the limitations of R-CHOP(-like) regimens in contemporary DLBCL management and therefore an ongoing need for improved therapeutic strategies.