2021 年 47 巻 6 号 p. 307-315
The criteria for palliative chemotherapy discontinuation have not been adequately systematized. We evaluated the relevance of the neutrophil-lymphocyte ratio, platelet-lymphocyte ratio, prognostic nutritional index, modified Glasgow Prognostic Score (mGPS), and clinicopathological factors as potential factors for chemotherapy discontinuation in patients with recurrent and unresectable pancreatic cancer.
We retrospectively analyzed the data of 91 patients who received palliative chemotherapy for recurrent and unresectable pancreatic cancer at Hiroshima City Hospital between April 2014 and March 2018. Factors significantly related to chemotherapy discontinuation were extracted using Coxʼs proportional-hazard model, and a prognostic model was established by combining these factors.
The median overall survival was 76 days. Multivariate analysis of the factors revealed that the mGPS (0/1-2) (hazard ratio [HR] = 3.053, P = 0.005), the presence of distant metastatic disease (HR = 2.605, P < 0.001), and the status of recurrent or initially unresectable disease (HR = 2.587, P = 0.013) were significantly associated with the discontinuation decision. One point was assigned to each of these three factors to create the prognostic model. A total score index of 0-3 was used to categorize three prognostic risk groups. The high-risk group (3 points) had a significantly lower overall survival than the low- (≤1 point) (P < 0.001) and intermediate-risk (2 points) groups (P < 0.001).
Our study shows that mGPS and this prognostic model can help determine whether chemotherapy should be discontinued in patients with relapsed and unresectable pancreatic cancer.
