Determinants of eligibility for second-line chemotherapy following gemcitabine plus nab-paclitaxel therapy in patients with unresectable pancreatic cancer: a retrospective study.

内容摘要

Second-line chemotherapy (2L) is recommended after gemcitabine plus nab-paclitaxel (GnP) first-line chemotherapy (1L) for unresectable pancreatic cancer (UR-PC). However, in routine clinical practice, not all patients proceed to 2L. This retrospective study aimed to identify baseline clinical and biological factors associated with 2L eligibility. We retrospectively reviewed the data of 124 consecutive patients with UR-PC who received 1L GnP between 2016 and 2024 at a single center, excluding those with postoperative recurrence. Patients were grouped by 2L receipt [2L(&#x2009;+), n&#x2009;=&#x2009;63] or non-receipt [2L(&#x2009;-), n&#x2009;=&#x2009;61]. Ascites was assessed based on baseline imaging findings and additionally classified as none, mild, or moderate-to-severe. Overall survival (OS) was assessed from completion or discontinuation of 1L GnP, and progression-free survival (PFS) was assessed from 1L GnP initiation, using Kaplan-Meier and Cox regression analyses. Cox regression analysis for OS included only baseline variables. The 2L(&#x2009;+) group less frequently had baseline ascites and better Eastern Cooperative Oncology Group performance status. The distribution of ascites severity also differed significantly between the groups. Median OS was 7.1 vs. 1.9&#xa0;months (p&#x2009;<&#x2009;0.001) and median PFS was 5.9 vs. 3.0&#xa0;months (p&#x2009;=&#x2009;0.004) for 2L(&#x2009;+) vs. 2L(&#x2009;-). Multivariate Cox analysis identified the absence of ascites as an independent factor associated with longer OS (hazard ratio [HR] 0.595, 95% confidence interval 0.367-0.963; p&#x2009;=&#x2009;0.035). Multivariate logistic regression revealed that the absence of ascites independently predicted 2L eligibility (adjusted odds ratio 4.435, 95% CI 1.583&#x2009;-&#x2009;12.423, p&#x2009;=&#x2009;0.005). Baseline ascites was independently associated with reduced eligibility for 2L after 1L GnP in patients with UR-PC. Nevertheless, patients who received 2L had longer survival than those who did not, including among those with baseline ascites. However, this association should be interpreted cautiously because of the retrospective design and potential treatment-selection bias. These findings may support individualized reassessment and proactive supportive care to maximize opportunities for sequential chemotherapy.