Association of overall survival benefit of radiotherapy with progression-free survival after chemotherapy for diffuse large B-cell lymphoma: A systematic review and meta-analysis

Jingnan Wang,Xin Liu,Yunpeng Wu,Qiuzi Zhong,Tao Wu,Yong Yang,Bo Chen,Hao Jing,Yuan Tang,Jing Jin,Yueping Liu,Yongwen Song,Hui Fang,Ningning Lu,Ning Li,Yirui Zhai,Wenwen Zhang,Min Deng,Shulian Wang,Fan Chen,Lin Yin,Chen Hu,Shunan Qi,Yexiong Li
DOI: https://doi.org/10.1016/j.jncc.2024.04.002
2024-04-23
Abstract:Objective: To evaluate whether improved progression-free survival (PFS) from radiotherapy (RT) translates into an overall survival (OS) benefit for diffuse large B-cell lymphoma (DLBCL). Methods: A systematic literature search identified randomized controlled trials (RCTs) and retrospective studies that compared combined-modality therapy (CMT) with chemotherapy (CT) alone. Weighted regression analyses were used to estimate the correlation between OS and PFS benefits. Cohen's kappa statistic assessed the consistency between DLBCL risk-models and PFS patterns. Furthermore, the benefit trend of RT was analyzed by fitting a linear regression model to the pooled hazard ratio (HR) according to the PFS patterns. Results: For both 7 RCTs and 52 retrospective studies, correlations were found between PFS HR (HRPFS) and OS HR (HROS) at trial level (r = 0.639-0.876), and between PFS and OS rates at treatment-arm level, regardless of CT regimens (r = 0.882-0.964). Incorporating RT into CT increased about 18% of PFS, and revealed a different OS benefit profile. Patients were stratified into four CT-generated PFS patterns (>80%, >60-80%, >40-60%, and ≤40%), which was consistent with risk-stratified subgroups (kappa > 0.6). Absolute gain in OS from RT ranged from ≤5% at PFS >80% to about 21% at PFS ≤40%, with pooled HROS from 0.70 (95% CI, 0.51-0.97) to 0.48 (95% CI, 0.36-0.63) after rituximab-based CT. The OS benefit of RT was predominant in intermediate- and high-risk patients with PFS ≤ 80%. Conclusion: We demonstrated a varied OS benefit profile of RT to inform treatment decisions and clinical trial design.
What problem does this paper attempt to address?