双原发癌对非小细胞肺癌患者生存的影响：大规模队列研究

Background: Although survival outcomes for non-small cell lung cancer (NSCLC) have improved with advances in treatment, the increasing number of long-term survivors has brought greater attention to double primary cancers (DPCs). This study aimed to evaluate the frequency and distribution of double primary cancers (DPC) in patients with NSCLC and to assess their impact on survival outcomes. Methods: This study evaluated clinical variables of patients diagnosed with NSCLC between 2008 and 2024. The types of DPCs were identified, and the impact of DPC occurrence and timing on the survival outcomes of NSCLC was analyzed. To minimize the influence of lead-time bias, this study applied Cox regression models incorporating time-varying covariates. Multivariable analyses were then performed to evaluate the prognostic impact of DPC timing on survival, adjusting for age, smoking status, stage, histology, DPC duration, multiple primary cancer status, and the presence of EGFR mutations or ALK rearrangements. Results: Among a total of 34,593 patients diagnosed with NSCLC, 4,339 (12.5%) were identified as having DPC, and 455 (1.3%) had multiple primary cancers (MPCs) involving three or more malignancies. Among the total 34,593 patients with NSCLC, DPC was identified before NSCLC in 2,345 patients (6.8%), concurrently in 1,039 (3.0%), and after NSCLC diagnosis in 1,155 (3.3%). The incidence of DPC was slightly higher in males (12.9%) than in females (12.0%) and showed an inverse association with NSCLC stage-occurring in 17.8%, 13.0%, 9.4%, and 6.4% of patients with stage I, II, III, and IV disease, respectively. In males, the most frequent DPCs were gastric cancer (2.7%), colorectal cancer (2.1%), prostate cancer (2.1%), and liver cancer (1.0%). Among 32,170 patients after excluding those with post-NSCLC DPC, pre-existing DPC was associated with worse survival compared with no DPC (adjusted HR = 1.146, P = 0.0018). However, DPC diagnosed more than 5 years before NSCLC showed no significant effect on survival (adjusted HR = 1.094, P = 0.1965), whereas DPC within 5 years conferred poorer outcomes (adjusted HR = 1.180, P = 0.0018). In a time-varying Cox model including 31,827 patients with concurrent or post-NSCLC DPC, the development of DPC after NSCLC remained independently associated with worse survival (adjusted HR = 1.377, P < 0.0001). Conclusions: DPC diagnosed more than 5 years before NSCLC had no impact on survival, whereas DPC occurring within 5 years before or after NSCLC was independently associated with worse outcomes. These findings underscore the importance of increased vigilance and individualized management strategies for NSCLC patients with recent or subsequent DPCs, particularly in an era of improving long-term survival.