Radiologically occult metastatic pancreatic cancer: how can we avoid unbeneficial resection?

Purpose This study aimed to clarify the key factors for minimizing unsuitable surgical interventions for patients with radiologically occult metastatic pancreatic cancer (ROMPC), defined as a distant metastasis detected during surgery or within 6 months after resection. Methods This study involved 502 patients planned to undergo curative resection for pancreatic cancer between 2008 and 2015. Patients were divided into ROMPC and non-ROMPC groups and evaluated preoperative factors associated with ROMPC. Results Overall survival (OS) was significantly lower in the ROMPC group (n = 145) than the non-ROMPC group (n = 357, median survival time [MST] 10.8 vs. 35.3 months, P < 0.001). In the ROMPC group, OS tended to be worse for patients who had pancreatectomies (n = 84) than those who did not (n = 61, MST 10.1 vs. 13.2 months, P = 0.057), and the next chemotherapy started significantly later in patients who had pancreatectomies (P < 0.001). Moreover, OS was significantly lower for patients with (n = 82) than without (n = 63) liver metastases (MST 9.7 vs. 13.0 months, respectively, P = 0.020). The best indicator for patients at higher risk of ROMPC was a combination of carbohydrate antigen 19-9 concentration >= 300 U/ml and tumor size >= 30 mm. Conclusions In the ROMPC group, patients who underwent pancreatectomy had a poorer prognosis than patients not undergoing pancreatectomy. Given that the liver was the most frequent distant metastatic site for ROMPC and had the poorest prognosis, establishing a strategy featuring new imaging modalities to detect radiologically occult liver metastases is necessary.