Nationwide data on the effect of primary treatment on median overall survival (mOS) in an entire unselected population of patients (pts) with pancreatic cancer (PC) have not been reported before. The aim of the present study was to investigate the effect of initial treatment on mOS in all incident pts with PC in Denmark in a recent five-year period (2011 to 2016).
From 1 May 2011 to 30 April 2016, 4260 pts diagnosed with PC were identified in the national Danish Pancreatic Cancer Database. Last follow up was 10 September 2017. Excluded were 99 pts (2%), 56 due to combined preoperative chemotherapy followed by resection, 22 due to other malignancies, 13 due to incorrect registration of treatment, 6 pts treated with unknown chemotherapy and 2 were lost to follow up, leaving 4161 pts included. The mOS was analysed from the date of the initial treatment, either resection or chemotherapy or from the date of diagnosis in case of best supportive care (BSC).
Initial treatment and mOS for all 4161 pts.Table: 743P
|Treatment||Pts, No.||mOS, months||95% CI|
Abbreviations: CI: confidence interval, N: lymph node status, without – or with +, Gem: gemcitabine, FOLFIRINOX: 5-flourouracil, leukovorin, irinotecan and oxaliplatin, Cap: capecitabine, S1: tegafur/gimeracil/oteracil, Pac: nab-paclitaxel, others: other regimens. *There were 38 pts without histopathological reports on lymph node status.
The initial resected lymph node negative pts had the longest survival; double that of lymph node positive pts. Pts initially treated with chemotherapy had slightly shorter mOS than found in randomized trials, reflecting patient characteristics in an unselected population. The outcome of gemcitabine monotherapy was poor, possibly reflecting less treatment effect and selection of less fit pts. The BSC group was larger than expected and further investigations, particularly in early diagnosis of PC are of utmost importance.
Clinical trial identification
Legal entity responsible for the study
The Axel Muusfeldt foundation.
All authors have declared no conflicts of interest.