Prognostic accuracy of the international classification for grading the resectability of pancreatic cancer: is it time for a fourth dimension?

Background The 2017 international consensus criteria expanded the definition of anatomically borderline resectable pancreatic ductal adenocarcinoma to include both a biological and a conditional dimension. An inflammatory variable was mentioned but not included.

Aim The main aim of this study was determing for the first time the prognostic accuracy of all 12 possible subgroups regarding oncological outcomes and longterm survival. Secondly, we include and evaluate modified Glasgow prognostic score (mGPS) and Neutrophil-to-lymphocite ratio (NLR) to define a borderline resectable inflammatory variable (BR-I) as a new potential variable of the ICC.

Methods In the period 2010-2022, a total of 419 patients who underwent pancreatic resection due to adenocarcinoma were included retrospectively in the study. They were divided into 12 groups according to the international consensus classification with all possible combinations of the different variables. We also considered inflammatory factors and created a novel variable.

Results The cohort included 222 male and 197 female patients with a median age of 69 years and median BMI of 25 kg/m². The median overall survival (OS) was 24 months, with a median disease-free survival (DFS) of 9 months for the entire cohort. A significant difference in median OS between patients with PR-B and BR-B – 29 months, compared to 19 months. The subgroup analysis and the results showed that groups consisting of patients with CA 19-9 value below the cut-off value of 500 U/ml had favorable outcomes compared to groups with BR-B status. We analyzed the mGPS and NLR values of the patients, with complete data available for 137 patients. 74 had a NLR value less than 3. Patients with NLR<3 had a significant difference in median OS of 29 months compared to those with NLR≥3 (23 months). For mGPS, patients with a score of 0 had a median OS of 29 months, those with a score of 1 had a median OS of 23 months, and those with a score of 2 had a median OS of 16 months. The median DFS did not differ significantly between groups for NLR or mGPS scores. 26.5% of patients met the criteria for BR-I, and the median OS and DFS were 23 and 11 months, respectively.

Conclusion The preoperative classification into anatomical as well as biological and conditional resectability criteria has a high prognostic accuracy. The inflammatory variable has been demonstrated to potentially have a significant predictive role in determining the outcome.

 

MissingLinkText

Publication History

Article published online:
28 August 2023