Evaluation of preoperative diagnostic methods for resectable pancreatic cancer: a diagnostic capability and impact on the prognosis of endoscopic ultrasound-guided fine needle aspiration

Abstract

Abstract Background A pathological diagnosis of pancreatic cancer should be performed as much as possible to determine the appropriate treatment strategy, but priorities and algorithms for diagnostic methods have not yet been established. In recent years, the endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) has become the primary method of collecting tissues from pancreatic disease, but the effect of EUS-FNA on surgical results and prognosis has not been clarified. Aims To evaluate the diagnostic ability of EUS-FNA and its effect on the preoperative diagnosis, surgical outcome, and prognosis of pancreatic cancer. Methods Between January 2005 and June 2017, 293 patients who underwent surgical resection for pancreatic cancer were retrospectively evaluated. The outcomes of interest were the diagnostic ability of EUS-FNA and its influence on the surgical results and prognosis. Results The diagnostic sensitivity of EUS-FNA was 94.4%, which was significantly higher than that of endoscopic retrograde cholangiopancreatography (ERCP) (45.5%) (p < 0.001). The adverse event rate in ERCP was 10.2%, which was significantly higher than EUS-FNA (1.3%) (p = 0.001). Patients were divided into FNA group (N = 160) and non-FNA group (N = 133) for each preoperative diagnostic method. In the study of surgical curability R0 between the two groups, there was no significant difference in FNA group (65.0% [104/160]) and non-FNA group (64.7% [86/133], p = 1.000). In the prognostic study, 256 patients with curative R0 or R1 had a recurrence rate was 54.3% (70/129) in the FNA group and 57.4% (73/127) in the non-FNA group. Moreover peritoneal dissemination occurred in 34.3% (24/70) in the FNA group and in 21.9% (16/73) in the non-FNA group, neither of which showed a significant difference. The median survival times of the FNA and non-FNA groups were 955 days and 799 days, respectively, and there was no significant difference between the two groups (log-rank p = 0.735). In the Cox proportional hazards model, factors influencing prognosis, staging, curability, and adjuvant chemotherapy were the dominant factors, but the preoperative diagnostic method (EUS-FNA) itself was not. Conclusions EUS-FNA is a safe procedure with a high diagnostic ability for the preoperative examination of pancreatic cancer. It was considered the first choice without the influence of surgical curability, postoperative recurrence, peritoneal dissemination and prognosis.