The current role of blood-based biomarkers in surgical decision-making in patients with localised pancreatic cancer: A systematic review

The role of blood-based biomarkers in surgical decision-making in patients with localised pancreatic cancer remains unclear. This review aimed to report the utility of blood-based biomarkers focusing on prediction of response to neoadjuvant therapy, prediction of surgical resectability and early relapse after surgery. MEDLINE/PubMed, Embase and Web of Science were searched till October 2019. Studies published between January 2000 and September 2019 with a minimum of 20 patients with pancreatic adenocarcinoma, reporting the utility of at least one blood-based biomarker in predicting response to neoadjuvant therapy and predicting surgical resectability or early relapse after surgery were included. A total of 2604 studies were identified, of which 24 comprising of 3367 patients and 12 blood-based biomarkers were included. All included studies were observational. Levels of carbohydrate antigen (CA)19-9 were reported in the majority of the studies. Levels of CA19-9 predicted the response to neoadjuvant therapy and early relapse in 10 studies. CA125 levels above 35 U/ml were predictive of surgical irresectability in two studies. However, marked variation in both timing of sampling and cut-off values was noted between studies. Despite some evidence of potential benefit, the utility of currently available blood-based biomarkers in aiding surgical decision-making in patients undergoing potentially curative treatment for pancreatic cancer is limited by methodological heterogeneity. Standardisation of future studies may allow a more comprehensive analysis of the biomarkers described in this review. •Biomarkers may aid in surgical decision-making in patients with pancreatic cancer.•Low CA19-9 and CA125 levels were predictive of surgical resectability.•Dynamic change in CA19-9 levels was predictive of response to neoadjuvant therapy.•Elevated pre- and postoperative levels of CA19-9 were prognostic of early relapse.•A heterogeneity in cut-off levels and time of sampling for biomarkers was noted.