Faculty Bibliography

Pancreatic cancer is a lethal disease in a large part due to the systemic nature at the time of diagnosis. In those patients who undergo a potentially curative resection of pancreatic cancer, the overwhelming majority will have systemic relapse. Circulating tumor cells are an important mediator of the development of metastases. Circulating tumor cells have been identified in patients with clinically localized resectable pancreatic cancer and exist as several phenotypes. Mesenchymal and stem cell-like phenotypes of circulating tumor cells predict early recurrence and worse survival. This review focuses on the current understanding of circulating tumor cells in pancreatic cancer and how this information can be used in developing more effective therapy in the future.

INTRODUCTION:Approximately, 20% of patients with pancreatic ductal adenocarcinoma have resectable disease at diagnosis. Given improvements in locoregional and systemic therapies, some patients with borderline resectable pancreatic cancer (BRPC) can now undergo successful resection. The outcomes of patients with BRPC after neoadjuvant therapy remain unclear. METHODS:A prospectively maintained single-institution database was utilized to identify patients with BRPC who were managed at the Johns Hopkins Pancreas Multidisciplinary Clinic (PMDC) between 2013 and 2016. BRPC was defined as any tumor that presented with radiographic evidence of the involvement of the portal vein (PV) or superior mesenteric vein (SMV) that was deemed to be technically resectable (with or without the need for reconstruction), or the abutment (< 180° involvement) of the common hepatic artery (CHA) or superior mesenteric artery (SMA), in the absence of involvement of the celiac axis (CA). We collected data on treatment, the course of the disease, resection rate, and survival. RESULTS:Of the 866 patients evaluated at the PMDC during the study period, 151 (17.5%) were staged as BRPC. Ninety-six patients (63.6%) underwent resection. Neoadjuvant chemotherapy was administered to 142 patients (94.0%), while 78 patients (51.7%) received radiation therapy in the neoadjuvant setting. The median overall survival from the date of diagnosis, of resected BRPC patients, was 28.8 months compared to 14.5 months in those who did not (p < 0.001). Factors associated with increased chance of surgical resection included lower ECOG performance status (p = 0.011) and neck location of the tumor (p = 0.001). Forty-seven patients with BRPC (31.1%) demonstrated progression of disease; surgical resection was attempted and aborted in 12 patients (7.9%). Eight patients (5.3%) were unable to tolerate chemotherapy; six had disease progression and two did not want to pursue surgery. Lastly, four patients (3.3%) were conditionally unresectable due to medical comorbidities at the time of diagnosis due to comorbidities and failed to improve their status and subsequently had progression of the disease. CONCLUSION:After initial management, 31.1% of patients with BRPC have progression of disease, while 63.6% of all patients successfully undergo resection, which was associated with improved survival. Factors associated with increased likelihood of surgical resection include lower ECOG performance status and tumor location in the neck.

Liquid biopsy (LB) is a non-invasive approach representing a promising tool for new precision medicine strategies for cancer treatment. However, a comprehensive analysis of its reliability for pancreatic cancer (PC) is lacking. To this aim, we performed the first meta-analysis on this topic. We calculated the pooled sensitivity, specificity, positive (LR+) and negative (LR-) likelihood ratio, and diagnostic odds ratio (DOR). A summary receiver operating characteristic curve (SROC) and area under curve (AUC) were used to evaluate the overall accuracy. We finally assessed the concordance rate of all mutations detected by multi-genes panels. Fourteen eligible studies involving 369 patients were included. The overall pooled sensitivity and specificity were 0.70 and 0.86, respectively. The LR+ was 3.85, the LR- was 0.34 and DOR was 15.84. The SROC curve with an AUC of 0.88 indicated a relatively high accuracy of LB for molecular characterization of PC. The concordance rate of all mutations detected by multi-genes panels was 31.9%. LB can serve as surrogate for tissue in the molecular profiling of PC, because of its relatively high sensitivity, specificity and accuracy. It represents a unique opportunity to be further explored towards its introduction in clinical practice and for developing new precision medicine approaches against PC.

BACKGROUND:Intraductal papillary mucinous neoplasms (IPMNs) are precursor lesions to the development of pancreatic adenocarcinoma. We determined if non-O blood groups are more common in patients with IPMN and if blood group is a risk factor for progression to invasive pancreatic cancer among patients with IPMN. METHODS:The medical records were reviewed of all patients undergoing resection of an IPMN at Johns Hopkins Hospital from June 1997 to August 2016. Potential risk factors of high-grade dysplasia and associated adenocarcinoma were identified through a multivariate logistic regression model. RESULTS:Seven hundred and seventy-seven patients underwent surgical resection of an IPMN in which preoperative blood type was known. Sixty-two percent of IPMN patients had non-O blood groups (vs. 57% in two large US reference cohorts, P = 0.002). The association between non-O blood group was significant for patients with IPMN with low- or intermediate-grade dysplasia (P < 0.001), not for those with high-grade dysplasia (P = 0.68). Low- and intermediate-grade IPMNs were more likely to have non-type O blood compared to those with high-grade IPMN and/or associated invasive adenocarcinoma (P = 0.045). Blood type O was an independent predictor of having high-grade dysplasia without associated adenocarcinoma (P = 0.02), but not having associated invasive cancer (P = 0.72). The main risk factor for progression to invasive cancer after surgical resection was IPMN with high-grade dysplasia (P = 0.002). CONCLUSION:IPMN patients are more likely to have non-O blood groups than controls, but type O blood group carriers had higher odds of having high-grade dysplasia in their IPMN. These results indicate blood group status may have different effects on the risk and progression of IPMNs.

Locally advanced pancreatic cancer (LAPC) has several definitions but essentially is a nonmetastasized pancreatic cancer, in which upfront resection is considered not beneficial due to extensive vascular involvement and consequent high chance of a nonradical resection. The introduction of FOLFIRINOX chemotherapy and gemcitabine-nab-paclitaxel (gem-nab) has had major implications for the management and outcome of patients with LAPC. After 4-6 months induction chemotherapy, the majority of patients have stable disease or even tumor-regression. Of these, 12 to 35% are successfully downstaged to resectable disease. Several studies have reported a 30-35 months overall survival after resection; although it currently remains unclear if this is a result of the resection or the good response to chemotherapy. Following chemotherapy, selection of patients for resection is difficult, as contrast-enhanced computed-tomography (CT) scan is unreliable in differentiating between viable tumor and fibrosis. In case a resection is not considered possible but stable disease is observed, local ablative techniques are being studied, such as irreversible electroporation, radiofrequency ablation, and stereotactic body radiation therapy. Pragmatic, multicenter, randomized studies will ultimately have to confirm the exact role of both surgical exploration and ablation in these patients. Since evidence-based guidelines for the management of LAPC are lacking, this review proposes a standardized approach for the treatment of LAPC based on the best available evidence.

Background: Intraductal papillary mucinous neoplasm (IPMN) is premalig-nant pancreatic lesion. International guidelines suggest several treatments and malignancy predictors but offer limited predictors of individual risk. A nomo-gram to predict individual IPMN malignancy risk was released with good diagnostic performance, basedona cohort of2258 Korean or Japanese patients with IPMN. This study validated a nomogram to predict malignancy risk and inva-siveness of IPMN, using Eastern and Western cohorts.
Method(s): We collected clinicopathological and radiological data of patients who underwent pancreas resection for IPMN at 4 centers each in Eastern and Western countries. After excluding patients with >=1 missing malignancy predictor in the nomogram (main pancreatic duct diameter, cyst size, presence of mural nodule, serum CEA and CA 19-9 levels, and age), we analyzed data of the remaining 393 patients (Eastern: n = 265; Western: n = 128).
Result(s): Although mean age, sex, log value of serum CA 19-9, tumor location, main duct diameter, cyst size and presence of mural nodule differed between the Korea/Japan, Eastern and Western cohorts, rates of malignancy and invasive cancer did not significantly differ. Areas under the receiver operating characteristics curve (AUC) values using the nomogram to predict malignancy were Eastern: 0.745, Western: 0.856, and combined cohorts: 0.776; and to predict in-vasiveness were Eastern: 0.736, Western: 0.891, and combined cohorts: 0.788.
Conclusion(s): External validation of the nomogram showed good performance in predicting malignancy and invasive cancer in both Eastern and Western IPMN patients. The nomogram could be globally applicable to decide customized treatment options for patients with IPMN

Background and Objectives: We previously proposed a nomogram predicting individual risks of malignancy and invasiveness of intraductal papillary mucinous neoplasms and validated it in an external cohort. However, it is difficult to apply if data on tumor marker are lacking. The aim of the current study was to develop a new nomogram based on radiologic findings using previous nomogram development and an external validation cohort.
Material(s) and Method(s): A total of 3049 patients who underwent surgical resection at 30 tertiary institutes in 7 countries were enrolled and clinicopathologic data were retrospectively analyzed. Based on fitted model, area under the receiver operating characteristics curve (AUC) was calculated using 10-fold cross validation by exhaustive search.
Result(s): The study consisted of 1914 (62.8%) patients for previous nomogram development and 1135 patients (37.2%) in the external validation cohort. Among patients, 1898 (62.3%) had low, 577 (18.9%) had high grade dysplasia, and 574 (18.8%) had invasive carcinoma. Patients were allocated randomly into model development and test sets to construct the nomogram, with fixed ratios according to malignancy and invasiveness. Exhaustive search resulted in three variables (cyst size, duct dilatation, and mural nodule) for malignancy and four variables (cyst size, duct dilatation, mural nodule, and location) for invasiveness being selected to construct the nomogram, and AUC was 0.742 and 0.741, respectively. AUC for test set was 0.727 and 0.704, respectively, and Hosmer-Lemeshow goodness of fit test showed good discrimination power (p = 0.066 and 0.067, respectively).
Conclusion(s): The new nomogram based on radiologic findings is accurate and helpful in identifying patients at risk of malignancy and invasiveness and selecting treatment options in clinical settings.
Copyright