Impact of Preoperative Weight Loss on Prognosis in Patients with Pancreatic Cancer

<b>Background/Objectives</b>: This study aimed to investigate the impact of preoperative weight loss on long-term postoperative survival and the significance of preoperative nutrition on perioperative weight change in patients with pancreatic cancer. <b>Methods</b>: Overall, 125 patients who underwent radical resection for invasive pancreatic ductal carcinoma were retrospectively analyzed. The preoperative weight loss rate (pre-%WL) from the initial visit to immediately before surgery was calculated. Patients were classified by pre-%WL into the weight-loss (≥6% loss) and weight-maintenance (<6% loss) groups. The association of pre-%WL with postoperative outcomes and long-term survival was assessed. We evaluated preoperative nutrition's effect on perioperative weight change. <b>Results</b>: The study cohort included 91 (73%) and 34 (27%) patients with weight maintenance and weight loss, respectively. Specifically, the weight-loss group had a longer operative time (<i>p</i> = 0.025) and greater blood loss (<i>p</i> = 0.018) than the weight-maintenance group. Patients with weight loss had significantly poorer recurrence-free survival (RFS; 8.7 vs. 17.8 months, <i>p</i> = 0.004) and overall survival (OS; 18.1 vs. 45.2 months, <i>p</i> = 0.002) than those with weight maintenance. Multivariate analysis revealed weight loss as an independent prognostic indicator of poor RFS (hazard ratio = 2.07; <i>p</i> = 0.003) and OS (hazard ratio = 2.55; <i>p</i> = 0.0008). The presence or absence of preoperative nutritional therapy was not correlated with the pre-%WL but was associated with postoperative (by the time of discharge) weight loss rate (median weight change rate: -2.9% vs. -5.6%, <i>p</i> = 0.001). <b>Conclusions</b>: Preoperative weight loss ≥ 6% was associated with poor RFS and OS in patients with pancreatic cancer. Although preoperative nutritional therapy did not suppress preoperative weight loss, it suppressed postoperative weight loss.