Background: Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most common malignant tumors both in Japan and Mongolia. Surgical resection is considered to be the potentially curative treatment for PDAC without distant metastases. Nevertheless, the 5-year overall survival rate after pancreatic resection alone has been reported only 8 to 10 %. The most significant factor affecting survival is the high rate of postoperative recurrence. The introduction of chemotherapy in adjuvant settings has been proved beneficial and prolonged overall survival rate. However, some patients are failed to complete adjuvant chemotherapy and obtain significantly less survival benefit. In this study, we retrospectively reviewed the impact of completion of adjuvant chemotherapy on survival and revealed independent risk factors of incompletion of adjuvant chemotherapy. Methods: From January 2000 and September 2012, a total of 236 consecutive patients at Kobe University Hospital who were planned pancreatic resection for PDAC. Forty-five patients who had metastatic disease (liver, peritoneum and/or distant lymph nodes) at laparotomy, 13 patients with R2 resection, and 46 patients who did not receive adjuvant chemotherapy were excluded, and the remaining 132 patients received adjuvant chemotherapy (gemcitabine alone, S1 alone, or gemcitabine plus S1 regimen) for 6 months. Excluding 27 patients discontinued therapy due to recurrent disease during adjuvant chemotherapy, 105 patients (55 men and 50 women) were enrolled in this analysis. Preoperative factors (sex, age, body mass index [BMI], comorbidity, carcinoembrionic antigen [CEA] level, carbohydrate antigen 19-9 [CA19-9] level, white blood cell count [WBC], lymphocyte count, C-reactive protein [CRP] level, total protein level, and serum albumin level), intraoperative factors (type of operative procedure, duration of operation, and intraoperative blood transfusion), postoperative factors (tumor size, stages, histological type, lymph node metastasis, postoperative complications, body weight loss, CEA level, CA19-9 level, WBC, lymphocyte count, CRP level, total protein level, serum albumin level, and serum cholinesterase level) were evaluated. Results: Seventy-five (71 %) patients completed adjuvant chemotherapy (complete group), and 30 (29 %) patients discontinued therapy due to adverse events (incomplete group). Patient survival was significantly better in complete group than in incomplete group (median survival time; 48.9 months vs 17.9 months, 5-year survival rate; 42.7 % vs 17.1 %). The statistically significant factors in incomplete adjuvant chemotherapy by univariate analysis were high preoperative WBC, type of pancreatectomy (pancreatoduodenectomy or total pancreatectomy vs distal pancreatectomy), postoperative body weight loss ≥ 7.5 %, and low postoperative serum albumin level). Among these factors, multivariate analyses identified that low postoperative serum albumin level was identified as significant predictors for failure to complete adjuvant chemotherapy (odds ratio 0.02, 95 % confidence interval 0.002 to 0.14; P < 0.001). The cutoff value of the postoperative serum albumin level was determined using receiver operating characteristic curve, the area under the curve was 0.79, and the cutoff value was 3.1 mg/dl. The adjuvant chemotherapy was completed in 85.5 % and 44.4 % (p < 0.001) in lower (< 3.1 mg/dl) and high (≥ 3.1 mg/dl) postoperative serum albumin level groups, respectively. The overall 1-, 3-, 5-year survival rates were 73.9 %, 22.4 %, 18.0 % in a lower group, and 96.8 %, 64.6 %, and 41.0 % in a higher group respectively (p < 0.001). Conclusions: Patients who completed adjuvant chemotherapy had significantly better survival rate than those who failed to complete adjuvant chemotherapy. Postoperative serum albumin level was identified as an independent risk factor of incomplete adjuvant chemotherapy. Postoperative serum albumin level ≥ 3.1 mg/dl is a good marker of predicting completion of adjuvant chemotherapy and better survival. Maintenance of postoperative nutritional condition may be one of the important factors to achieve best prognosis in the treatment of PDAC.

Background: Borderline resectable pancreatic adenocarcinoma (BR-PAC) is defined as locally advanced tumor of the pancreas without metastasis that is, although potentially resectable (R), at high risk for positive resection margin following surgery. The therapeutic strategy has remained unestablished because BR-PAC is biologically a heterogeneous subset in which the preoperative prognostic factors are undetermined. Recently, several prognostic factors related to systemic inflammation have been explored in various kinds of cancers: the combination of serum C-reactive protein (CRP) and albumin as the modified Glasgow prognostic factor; a combination of CRP and white blood cell count in the prognostic index; a combination of albumin and lymphocyte counts in Onodera’s prognostic nutritional index; the neutrophil-lymphocyte ratio (NLR); and the platelet-lymphocyte ratio (PLR). Although these prognostic factors have been explored in some small cohort studies of PAC patients, the results still remain controversial especially because PAC patients with diverse clinical stages were included in the cohorts. It has never been reported whether or not the systemic inflammatory response is validated as a predictive risk factor in cohorts of only advanced BR-PAC patients. Method: Between January 2003 and June 2012 at Kobe University Hospital, 136 consecutive pancreatic adenocarcinoma (PAC) patients who underwent surgical curative resection were retrospectively studied. Prior to surgery, the PAC patients were stratified into R- and BR-PAC patients according to the National Comprehensive Cancer Network guidelines. To evaluate the independent prognostic significance of NLR and PLR, univariate and multivariate Cox proportional-hazard models were applied. Results: The median survival in PAC patients with preoperative NLR > 3 (n=45) and NLR < 3 (n=91) was 17.5 months and 31.1 months, respectively (P=0.0037). However, the median survival in PAC patients with PLR > 225 (n=32) and PLR < 225 (n=104) was 21.8 months and 26.2 months, showing no significant difference in overall survival between the two groups (P=0.2526). The median survival in the R-PAC patients with NLR > 3 (n=38) and NLR < 3 (n=70) was 18.1 months and 33.1 months, respectively (P=0.0138). However, the median survival in the R-PAC patients with PLR > 225 (n=27) and PLR < 225 (n=81) was 24.1 months and 25.8 months, showing no significant difference in overall survival between the two groups (P=0.6533). The median survival in BR-PAC patients with preoperative NLR > 3 (n=7) and NLR < 3 (n=21) was 14.8 months and 27.2 months, respectively (P=0.0068). In addition, median survival in BR-PAC patients with preoperative PLR > 225 (n=5) and PLR < 225 (n=23) was 14.8 months and 26.2 months, respectively (P=0.0050). Preoperative NLR > 3 (HR=21.437, 95% CI=4.119-142.980; P=0.0002) and PLR > 225 (HR=30.993, 95% CI=3.844- 384.831; P=0.0009) were the only independent prognostic factors in BR-PAC patients. Conclusion: Preoperative NLR and PLR offer independent prognostic information regarding overall survival in BR-PAC patients following curative resection. The workup is only to obtain a blood sample of 3 mL from PAC patients immediately before treatment. In the near future, these factors associated with the systemic inflammatory response may have the potential to become criteria for BRPAC candidates to undergo neoadjuvant chemotherapy and/or neoadjuvant chemoradiation followed by surgical resection

Background: Borderline resectable pancreatic adenocarcinoma (BR-PAC) isdefined as locally advanced tumor of the pancreas without metastasis that is,although potentially resectable (R), at high risk for positive resection marginfollowing surgery. The therapeutic strategy has remained unestablished becauseBR-PAC is biologically a heterogeneous subset in which the preoperativeprognostic factors are undetermined. Recently, several prognostic factors relatedto systemic inflammation have been explored in various kinds of cancers: thecombination of serum C-reactive protein (CRP) and albumin as the modifiedGlasgow prognostic factor; a combination of CRP and white blood cell countin the prognostic index; a combination of albumin and lymphocyte counts inOnodera’s prognostic nutritional index; the neutrophil-lymphocyte ratio (NLR);and the platelet-lymphocyte ratio (PLR). Although these prognostic factors havebeen explored in some small cohort studies of PAC patients, the results still remaincontroversial especially because PAC patients with diverse clinical stages wereincluded in the cohorts. It has never been reported whether or not the systemicinflammatory response is validated as a predictive risk factor in cohorts of onlyadvanced BR-PAC patients.Method: Between January 2003 and June 2012 at Kobe University Hospital,136 consecutive pancreatic adenocarcinoma (PAC) patients who underwentsurgical curative resection were retrospectively studied. Prior to surgery, the PACpatients were stratified into R- and BR-PAC patients according to the NationalComprehensive Cancer Network guidelines. To evaluate the independentprognostic significance of NLR and PLR, univariate and multivariate Coxproportional-hazard models were applied.Results: The median survival in PAC patients with preoperative NLR > 3 (n=45)and NLR < 3 (n=91) was 17.5 months and 31.1 months, respectively (P=0.0037).However, the median survival in PAC patients with PLR > 225 (n=32) and PLR< 225 (n=104) was 21.8 months and 26.2 months, showing no significantdifference in overall survival between the two groups (P=0.2526). The mediansurvival in the R-PAC patients with NLR > 3 (n=38) and NLR < 3 (n=70) was 18.1months and 33.1 months, respectively (P=0.0138). However, the median survivalin the R-PAC patients with PLR > 225 (n=27) and PLR < 225 (n=81) was 24.1months and 25.8 months, showing no significant difference in overall survivalbetween the two groups (P=0.6533). The median survival in BR-PAC patientswith preoperative NLR > 3 (n=7) and NLR < 3 (n=21) was 14.8 months and 27.2months, respectively (P=0.0068). In addition, median survival in BR-PAC patientswith preoperative PLR > 225 (n=5) and PLR < 225 (n=23) was 14.8 months and26.2 months, respectively (P=0.0050). Preoperative NLR > 3 (HR=21.437, 95%CI=4.119-142.980; P=0.0002) and PLR > 225 (HR=30.993, 95% CI=3.844-384.831; P=0.0009) were the only independent prognostic factors in BR-PACpatients.Conclusion: Preoperative NLR and PLR offer independent prognostic informationregarding overall survival in BR-PAC patients following curative resection. Theworkup is only to obtain a blood sample of 3 mL from PAC patients immediatelybefore treatment. In the near future, these factors associated with the systemicinflammatory response may have the potential to become criteria for BRPACcandidates to undergo neoadjuvant chemotherapy and/or neoadjuvantchemoradiation followed by surgical resection