This study was approved by the institutional review board of Kagawa University. A total of 142 consecutive patients undergoing pancreatectomy for PDAC between January 2000 and May 2016 were retrospectively examined. Informed consent was obtained from all patients according to the institutional protocol of our hospital. All 142 patients had a PDAC that was histologically examined by at least two pathologists. Of the 142 patients, 27 patients were excluded. Ten patients were censored within 6 months, 10 were classified as unresectable category based on the 2016 NCCN guideline [11], 5 had unclear recurrence timing, 1 underwent R2 resection, and 1 had perioperative mortality. The data from the remaining 115 patients were retrospectively analyzed.
The patients were diagnosed with R (n=86) or borderline resectable ([BR], n=29) PDAC according to the 2016 NCCN guidelines [11]. All surgical procedures were divided into the following three types: classic, pylorus-preserving, or subtotal stomach-preserving pancreaticoduodenectomy (PD) in 75 patients (65%); distal pancreatectomy in 36 (31%); and total pancreatectomy in 4 (3%). Systematic lymph node dissection was performed in all operations. R0 resection was achieved in 104 patients (90%) and R1 was achieved in 11 (10%). R0 resection was defined as negative margin based on Union for International Cancer Control (UICC) definition. Preoperative CRT and postoperative ACT were given to 51 (44%) and 85 (74%) patients, respectively. Among the 115 patients, 34 (30%) experienced early recurrence within 6 months postoperatively (group E) and 81 (70%) did not (group NE).
Outcome measures
The variables included age; sex; body mass index (BMI); tumor location; resectability; serum C-reactive protein, serum albumin, hemoglobin, and serum CA19-9 levels; neutrophil/lymphocyte ratio (NLR); lymphocyte count; modified Glasgow Prognostic Score [10]; the standardized uptake value (SUV) seen on 18F-fluorodeoxy glucose positron emission tomography (FDG-PET); and presence or absence of preoperative CRT. The intraoperative data, including the operation time, estimated blood loss, blood transfusion, and portal vein resection; and postoperative data on morbidity according to the Clavien-Dindo classification [12], CA19-9 normalization status, and postoperative ACT induction, were reviewed and included. A receiver operating characteristic (ROC) curve was constructed to estimate the optimal cutoff value for the recurrence of PDAC within 6 months postoperatively, which was determined as the point closest to the upper left-hand corner of the graph. The ROC curves demonstrated that the cutoff points of the preoperative serum CA19-9 level, NLR, lymphocyte count, SUV on FDG-PET, and tumor size were 173, 4.65, 1648, 4.73, and 3.0, respectively; and the areas under the curve (AUC) were 0.637, 0.466, 0.556, 0.593, and 0.639, respectively.
Preoperative chemoradiotherapy
We introduced short-term neoadjuvant hypofractionated chemoradiotherapy with S1 in patients with R and BR PDAC between January 2009 and May 2016, and already reported the efficacy and safety [13]. Hypofractionated, external-beam radiotherapy (30 Gy in 10 fractions) with concurrent S1 (60 mg/m2) was delivered 5 days per week for 2 weeks prior to pancreatectomy. All of the patients with preoperative CRT (n=51) in this study received radiotherapy in a similar way. Meanwhile, all patients before December 2008 underwent upfront surgery.
Adjuvant therapy and follow-up
ACT was applied postoperatively unless contraindicated by the patients’ conditions. The patients received gemcitabine, referring to the results of the CONKO-001 trial [14] between 2006 and 2012; or S-1, referring to the results of the JASPAC01 trial since 2013 [15], according to the recommended protocols. S-1 is an oral fluoropyrimidine consisting of tegafur, a prodrug of fluorouracil, and two biochemical modulators. It is characterized by the inhibition of dihydropyrimidine dehydrogenase activity by gimeracil, the maintenance of a high concentration of fluorouracil, and by the suppression of fluorouracil’s phosphorylation in the gastrointestinal tract by oteracil potassium, thereby reducing gastrointestinal toxicity. Gemcitabine at a dose of 1,000 mg/m2 was administered weekly for 3 weeks, followed by 1 week of rest; oral S-1 (80 mg/m2/day) was administered from days 1 to 28, followed by a 2-week rest period or from days 1 to 14, followed by a 1-week rest period. Chemotherapy was initiated within 2 months postoperatively in all patients who were considered eligible for the treatment. The follow-up examinations were performed every 2–3 months for 1 year and every 6 months thereafter, until the disease progressed. Enhanced computed tomography was performed every 6 months. We moved the examination date forward or added magnetic resonance imaging or FDG-PET, if necessary.
Statistical analysis
The clinicopathological features of patients in the groups E and NE were compared. The categorical variables were compared between the groups using the chi-square test and Fisher’s exact test. Survival was calculated using the Kaplan–Meier method and was compared between the groups using the log-rank test. A multivariate analysis using the backward elimination method of the Cox proportional hazards model was undertaken, using variables from the univariate analysis that were considered to potentially affect survival (P<0.10). P-values <0.05 were considered statistically significant. All statistical analyses were performed using the SPSS Statistics 25.0 for Windows software program (SPSS, Inc., Chicago, IL, USA).