Deep learning can predict microsatellite instability directly from histology in gastrointestinal cancerMicrosatellite instability determines whether patients with gastrointestinal cancer respond exceptionally well to immunotherapy. However, in clinical practice, not every patient is tested for MSI, because this requires additional genetic or immunohistochemical tests. Here we show that deep residual learning can predict MSI directly from H&E histology, which is ubiquitously available. This approach has the potential to provide immunotherapy to a much broader subset of patients with gastrointestinal cancer. A deep residual learning framework identifies microsatellite instability in histology slides from patients with cancer and can be used to guide immunotherapy.
Pancreatogastrostomy Versus Pancreatojejunostomy for RECOnstruction After PANCreatoduodenectomy (RECOPANC, DRKS 00000767)OBJECTIVES: To assess pancreatic fistula rate and secondary endpoints after pancreatogastrostomy (PG) versus pancreatojejunostomy (PJ) for reconstruction in pancreatoduodenectomy in the setting of a multicenter randomized controlled trial. BACKGROUND: PJ and PG are established methods for reconstruction in pancreatoduodenectomy. Recent prospective trials suggest superiority of the PG regarding perioperative complications. METHODS: A multicenter prospective randomized controlled trial comparing PG with PJ was conducted involving 14 German high-volume academic centers for pancreatic surgery. The primary endpoint was clinically relevant postoperative pancreatic fistula. Secondary endpoints comprised perioperative outcome and pancreatic function and quality of life measured at 6 and 12 months of follow-up. RESULTS: From May 2011 to December 2012, 440 patients were randomized, and 320 were included in the intention-to-treat analysis. There was no significant difference in the rate of grade B/C fistula after PG versus PJ (20% vs 22%, P = 0.617). The overall incidence of grade B/C fistula was 21%, and the in-hospital mortality was 6%. Multivariate analysis of the primary endpoint disclosed soft pancreatic texture (odds ratio: 2.1, P = 0.016) as the only independent risk factor. Compared with PJ, PG was associated with an increased rate of grade A/B bleeding events, perioperative stroke, less enzyme supplementation at 6 months, and improved results in some quality of life parameters. CONCLUSIONS: The rate of grade B/C fistula after PG versus PJ was not different. There were more postoperative bleeding events with PG. Perioperative morbidity and mortality of pancreatoduodenectomy seem to be underestimated, even in the high-volume center setting.