Positive Intraoperative Peritoneal Lavage Cytology is a Negative Prognostic Factor in Pancreatic Ductal Adenocarcinoma: A Retrospective Single-Center Study

Objective The aim of this study is to evaluate the prognostic significance of intraoperative peritoneal lavage cytology (PLC) in pancreatic invasive ductal adenocarcinoma. Methods Intraoperative PLC was evaluated in 162 patients with resectable pancreatic invasive ductal adenocarcinoma. The results were analyzed for correlations with clinicopathological parameters and/or prognoses. Results In the 162 cases of resectable pancreatic ductal adenocarcinoma, 18 (11%), 141 (87%), and 3 (2%) were positive, negative, and equivocal for intraoperative PLC, respectively. Intraoperative PLC positivity was associated with older patients (over 65 years), large tumor size (over 35 mm), tumor location in the body/tail of the pancreas, and distant metastasis. Univariate analysis showed that larger tumor sizes (≥35 mm, P = 0.001), lymph node metastases (P = 0.005), distant metastasis (P = 0.004), advanced stage (stage IIB or III, P = 0.006), advanced tumor histological grade (G3, P < 0.001), or positive intraoperative PLC (P = 0.002) are associated with a shorter survival. Multivariate analysis revealed that larger tumor sizes (≥35 mm, P = 0.026), lymph node metastasis (P = 0.021), advanced tumor histological grade (G3, P < 0.001), and positive intraoperative PLC (P = 0.002) were independent prognostic factors. Conclusion Intraoperative PLC is an independent prognostic factor for resectable pancreatic invasive ductal adenocarcinoma.


Introduction
Pancreatic ductal adenocarcinoma is the most common pancreatic neoplasm, and is highly aggressive and malignant with a very low-survival rate. Several factors, such as large tumor size, retroperitoneal invasion, location of the tumor in the body or tail of the pancreas, and lymph node metastasis, are correlated with poor prognosis (1). Ascites or peritoneal lavage fluids are occasionally analyzed in patients of several cancer types to evaluate the peritoneal dissemination of tumor cells. In ovarian and fallopian tube cancers, the detection of tumor cells in ascites or peritoneal lavages is considered a negative prognostic factor, and peritoneal cytology was adopted as one of the factors determining the stage of the tumor by the International Federation of Gynecology and Obstetrics as well as the Union for International Cancer Control/American Joint Committee on Cancer (UICC/AJCC) classification system (2)(3)(4). In pancreatic ductal adenocarcinoma, the incidence of positive peritoneal cytology was 5-30% in resectable cases (5)(6)(7)(8)(9)(10)(11)(12)(13)(14) and 20-57% in unresectable cases (6,8,9,11,15). However, the prognostic significance of positive peritoneal cytology has been controversial for pancreatic ductal adenocarcinoma. Some studies reported that positive peritoneal cytology is associated with poor prognosis (5,6,(8)(9)(10)(13)(14)(15)(16)(17). In contrast, other studies reported that peritoneal cytology findings do not influence the patients' prognosis (6-8, 11, 12, 18).
In this study, we retrospectively analyzed the relationship between intraoperative peritoneal lavage cytology (PLC) outcomes, clinicopathological parameters, and patient survival in resectable pancreatic ductal adenocarcinoma cases treated at Tokai University Hospital.

Cases
We retrospectively analyzed 162 consecutive cases of surgically resected, conventional pancreatic ductal adenocarcinoma with intraoperative PLC examination treated between 2006 and 2013 at Tokai University Hospital. All cases were diagnosed based on routine histological examination. We excluded cases of special histological types, such as adenosquamous carcinoma, mucinous carcinoma, undifferentiated carcinoma, and intraductal papillary mucinous neoplasms with invasion, as well as cases that had undergone preoperative chemotherapy. Tumor resectability was evaluated with computed tomography and at laparotomy. Lesions eligible for resection included those without distant metastases and those with or without involvement of the portal or superior mesenteric vein as long as resection and reconstruction margins were deemed safe. Furthermore, tumors with gastroduodenal artery encasement up to the hepatic artery, with either short segment encasement or direct abutment to the hepatic artery, were deemed resectable, as were lesions with no extensions to the celiac axis and tumors encasing the superior mesenteric artery in a manner not exceeding 180°of the circumference of the vessel wall (19). pT-stage, pN-stage, and tumor grade (G) were categorized according to the UICC/AJCC classification (3,4). After laparotomy, the Douglas fossa was washed with 200-300 mL of saline, and 10 mL of peritoneal ascites was aspirated using a syringe. The collected peritoneal lavage fluid was centrifuged for 10 min at 600 × g. After discarding the supernatant, precipitated cellular components were smeared onto glass slides, fixed in 95% ethanol, and subjected to Papanicolaou staining and/or were dried and subjected to Giemsa staining. Cytological results were classified as negative, positive, or equivocal. The clinical data were collected from individual patient records; patients' clinicopathological features are summarized in Table 1. This study was approved by

Correlation Between Clinicopathological Parameters and PLC
The correlation between PLC results and clinicopathological parameters was analyzed ( Table 2). Of 159 cases examined after excluding 3 patients with equivocal cytology results, positive PLC was associated with tumor sizes 35 mm or larger (P = 0.01), ages 65 years and older (P = 0.01), tumor location in the body/tail of the pancreas (P = 0.047), and the occurrence of distant metastasis (P = 0.002).

Discussion
Some studies reported that positive ascitic cytology or PLC are positively correlated with lymph node metastasis, tumor location in the body/tail of the pancreas, larger tumor size, vascular invasion, serosal invasion, positive surgical margins, pT-factor, and advanced stage (5,7,(9)(10)(11)(12)(13)(14)17). Our study revealed that positive intraoperative PLC in resectable pancreatic ductal adenocarcinoma was associated with larger tumor size (≥35 mm), older age (≥65 years), tumor location in the body/tail of the pancreas, and distant metastasis. Hence, our results are similar to those previously reported in the literature. As in previous studies, our data indicated that PLC positivity occurred in resectable pancreatic ductal adenocarcinomas in advanced stages. The relationship between positive peritoneal cytology results and survival of resectable pancreatic ductal adenocarcinoma patients has been controversial. While several studies concluded that positive peritoneal cytology is associated with poor  survival (5,6,9,10,13,14), other studies reported no significant difference in prognosis between positive and negative peritoneal cytology (7,8,11,12). Even with multivariate analysis, two conflicting results have been reported, one concluding that peritoneal cytology is an independent prognostic factor (14) and another stating that it is not (13).
The results of intraoperative PLC may be valuable for determining the subsequent treatment plan. Surgery for pancreatic cancer is highly invasive, and postoperative complications sometimes occur as well. Therefore, positive intraoperative PLC results may lead surgeons to decide to abandon the invasive resection procedure, particularly in patients who are older or in bad condition. Conversely, positive results in patients who are otherwise in good condition may prompt the pursuit of more aggressive postoperative therapy options.
In conclusion, this study revealed that positive intraoperative PLC in resectable pancreatic ductal adenocarcinoma was associated with older age, larger tumor size, tumor location in the body/tail of the pancreas, and distant metastasis. Patients with positive intraoperative PLC showed shorter survival times than those with negative PLC. Our data demonstrate that intraoperative positive PLC in resectable pancreatic ductal adenocarcinoma is a negative predictor of patient prognosis.