Postoperative outcomes in elderly patients undergoing pancreatic resection for pancreatic adenocarcinoma: A systematic review and meta-analysis

Pancreatic cancer is a disease of the elderly. Surgical resection is usually offered to patients in early stage disease; however, pancreatic resection in the elderly is controversial.MEDLINE, EMBASE and Cochrane Library, were searched for studies comparing short- and long-term outcomes of elderly (above the age of 70) with non-elderly patients (below the age of 70) following pancreatic resection for pancreatic adenocarcinoma over the period from the inception of electronic database to 2017. Twelve articles documenting 4860 patients were included. A meta-analysis of data on patient characteristics, operative techniques, and perioperative outcomes were analysed. Our primary endpoint was postoperative mortality, defined as 30-day mortality or in-hospitalisation mortality.There were 919 patients in the elderly group and 3941 patients in the non-elderly group. Elderly patients had worse ASA scores (p 0.001) and more cardiovascular comorbidities (p = 0.002). Tumour size, T-stage, N-stage and tumour grade were similar between the elderly and non-elderly group (p 0.05). Fewer elderly patients received a concomitant venous resection with their pancreatectomy (RR0.80, p = 0.003, I2 = 0%), achieved a negative margin status (RR0.76, p = 0.02, I2 = 28%) and underwent adjuvant chemotherapy treatment (RR0.69, p 0.001, I2 = 42%). Overall complication (RR1.15, p 0.001, I2 = 47%), in particular, respiratory complications (RR2.33, p = 0.004, I2 = 39%), was higher in the elderly group. There was no difference in postoperative pancreatic fistula formation, postoperative haemorrhage, intraabdominal abscess and length of hospital stay between both groups (p 0.05). Postoperative mortality was similar between both groups (p = 0.17). Subgroup analysis according to the time of enrolment (2000, ≥2000) showed a significant subgroup effect (Chi2 = 3.44, p = 0.06, I2 = 70.9%) and revealed that postoperative mortality in the elderly group improved over time (Before 2000: n = 1654, subtotal RR2.27, p = 0.02, I2 = 0%; From 2000 onwards: n = 3206, subtotal RR1.00, p = 0.99, I2 = 0%).Fewer elderly patients received chemotherapy and portal vein resection to achieve a clear margin. Pancreatic resection of pancreatic adenocarcinoma can be performed safely on elderly patients with acceptable risks in experienced centres by specialist hepatobiliary surgeons. Age alone should not be the only determinant for the selection of patients for surgical treatment of pancreatic adenocarcinoma.