Personalized RNA neoantigen vaccines stimulate T cells in pancreatic cancer.

Pancreatic ductal adenocarcinoma (PDAC) is lethal in 88% of patients1, yet harbours mutation-derived T cell neoantigens that are suitable for vaccines 2,3. Here in a phase I trial of adjuvant autogene cevumeran, an individualized neoantigen vaccine based on uridine mRNA-lipoplex nanoparticles, we synthesized mRNA neoantigen vaccines in real time from surgically resected PDAC tumours. After surgery, we sequentially administered atezolizumab (an anti-PD-L1 immunotherapy), autogene cevumeran (a maximum of 20 neoantigens per patient) and a modified version of a four-drug chemotherapy regimen (mFOLFIRINOX, comprising folinic acid, fluorouracil, irinotecan and oxaliplatin). The end points included vaccine-induced neoantigen-specific T cells by high-threshold assays, 18-month recurrence-free survival and oncologic feasibility. We treated 16 patients with atezolizumab and autogene cevumeran, then 15 patients with mFOLFIRINOX. Autogene cevumeran was administered within 3 days of benchmarked times, was tolerable and induced de novo high-magnitude neoantigen-specific T cells in 8 out of 16 patients, with half targeting more than one vaccine neoantigen. Using a new mathematical strategy to track T cell clones (CloneTrack) and functional assays, we found that vaccine-expanded T cells comprised up to 10% of all blood T cells, re-expanded with a vaccine booster and included long-lived polyfunctional neoantigen-specific effector CD8+ T cells. At 18-month median follow-up, patients with vaccine-expanded T cells (responders) had a longer median recurrence-free survival (not reached) compared with patients without vaccine-expanded T cells (non-responders; 13.4 months, P = 0.003). Differences in the immune fitness of the patients did not confound this correlation, as responders and non-responders mounted equivalent immunity to a concurrent unrelated mRNA vaccine against SARS-CoV-2. Thus, adjuvant atezolizumab, autogene cevumeran and mFOLFIRINOX induces substantial T cell activity that may correlate with delayed PDAC recurrence.

Risk Factors for Progression in Patients Undergoing Surveillance for Pancreatic Cysts.

OBJECTIVE: To identify risk factors associated with the progression of pancreatic cysts in patients undergoing surveillance. BACKGROUND: Previous studies of intraductal papillary mucinous neoplasms (IPMNs) rely on surgical series to determine malignancy risk and have inconsistently identified characteristics associated with IPMN progression. METHODS: We conducted a retrospective review of 2197 patients presenting with imaging concerning for IPMN from 2010 to 2019 at a single institution. Cyst progression was defined as resection or pancreatic cancer development. RESULTS: The median follow-up time was 84 months from the presentation. The median age was 66 years, and 62% were female. Ten percent had a first-degree relative with pancreatic cancer, and 3.2% had a germline mutation or genetic syndrome associated with an increased risk of pancreatic ductal adenocarcinoma (PDAC). Cumulative incidence of progression was 17.8% and 20.0% at 12 and 60 months postpresentation, respectively. Surgical pathology for 417 resected cases showed noninvasive IPMN in 39% of cases and PDAC with or without associated IPMN in 20%. Only 18 patients developed PDAC after 6 months of surveillance (0.8%). On multivariable analysis, symptomatic disease [hazard ratio (HR)=1.58; 95% CI: 1.25-2.01], current smoker status (HR=1.58; 95% CI: 1.16-2.15), cyst size (HR=1.26; 95% CI: 1.20-1.33), main duct dilation (HR=3.17; 95% CI: 2.44-4.11), and solid components (HR=1.89; 95% CI: 1.34-2.66) were associated with progression. CONCLUSIONS: Worrisome features on imaging at presentation, current smoker status, and symptomatic presentation are associated with IPMN progression. Most patients progressed within the first year of presentation to Memorial Sloan Kettering Cancer Center (MSKCC). Further investigation is necessary to develop personalized cyst surveillance strategies.

Early-Onset Pancreatic Neuroendocrine Tumors: Clinical Presentation, Pathology Features, and Oncological Outcomes.

BACKGROUND: Early-Onset (EO) pancreatic neuroendocrine tumor (PanNET) is a rare disease, but whether it is clinically different from late-onset (LO) PanNET is unknown. Our study aimed to evaluate clinical differences and disease outcomes between EO-PanNET and LO-PanNET and to compare sporadic EO-PanNET with those with a hereditary syndrome. METHODS: Patients with localized PanNET who underwent pancreatectomy at Memorial Sloan Kettering between 2000 and 2017 were identified. Those with metastatic disease and poorly differentiated tumors were excluded. EO-PanNET was defined as <50 and LO-PanNET >50 years of age at the time of diagnosis. Family history and clinical and pathology characteristics were recorded. RESULTS: Overall 383 patients were included, 107 (27.9%) with EO-PanNET. Compared with LO-PanNET, EO-PanNET were more likely to have a hereditary syndrome (2.2% vs. 16%, P <0.001) but had similar pathology features such as tumor grade ( P =0.6), size (2.2 Vs. 2.3 cm, P =0.5) and stageof disease ( P =0.8). Among patients with EO-PanNET, those with hereditary syndrome had more frequently a multifocal disease (65% vs. 3.3%, P <0.001). With a median follow-up of 70 months (range 0-238), the 5-year cumulative incidence of recurrence after curative surgery was 19% (95% CI 12%-28%) and 17% (95% CI 13%-23%), in EO-PanNET and LO-PanNET ( P =0.3). Five-year disease-specific survival was 99% (95% CI 98%-100%) with no difference with respect to PanNET onset time ( P =0.26). CONCLUSIONS: In this surgical cohort, we found that EO-PanNET is associated with hereditary syndromes but has pathologic characteristics and oncological outcomes similar to LO-PanNET. These findings suggest that patients with EO-PanNET can be managed similarly to those with LO-PanNET.

Gemcitabine with Cisplatin Versus Hepatic Arterial Infusion Pump Chemotherapy for Liver-Confined Unresectable Intrahepatic Cholangiocarcinoma.

BACKGROUND: A post-hoc analysis of ABC trials included 34 patients with liver-confined unresectable intrahepatic cholangiocarcinoma (iCCA) who received systemic chemotherapy with gemcitabine and cisplatin (gem-cis). The median overall survival (OS) was 16.7 months and the 3-year OS was 2.8%. The aim of this study was to compare patients treated with systemic gem-cis versus hepatic arterial infusion pump (HAIP) chemotherapy for liver-confined unresectable iCCA. METHODS: We retrospectively collected consecutive patients with liver-confined unresectable iCCA who received gem-cis in two centers in the Netherlands to compare with consecutive patients who received HAIP chemotherapy with or without systemic chemotherapy in Memorial Sloan Kettering Cancer Center. RESULTS: In total, 268 patients with liver-confined unresectable iCCA were included; 76 received gem-cis and 192 received HAIP chemotherapy. In the gem-cis group 42 patients (55.3%) had multifocal disease compared with 141 patients (73.4%) in the HAIP group (p = 0.023). Median OS for gem-cis was 11.8 months versus 27.7 months for HAIP chemotherapy (p < 0.001). OS at 3 years was 3.5% (95% confidence interval [CI] 0.0-13.6%) in the gem-cis group versus 34.3% (95% CI 28.1-41.8%) in the HAIP chemotherapy group. After adjusting for male gender, performance status, baseline hepatobiliary disease, and multifocal disease, the hazard ratio (HR) for HAIP chemotherapy was 0.27 (95% CI 0.19-0.39). CONCLUSIONS: This study confirmed the results from the ABC trials that survival beyond 3 years is rare for patients with liver-confined unresectable iCCA treated with palliative gem-cis alone. With HAIP chemotherapy, one in three patients was alive at 3 years.

Treatment Patterns and Outcomes in Pancreatic Cancer: A Comparative Analysis of Ontario and the USA.

BACKGROUND: Comparative studies evaluating quality of care in different healthcare systems can guide reform initiatives. This study seeks to characterize best practices by comparing utilization and outcomes for patients with pancreatic cancer (PC) in the USA and Ontario, Canada. METHODS: Patients (age ≥ 66 years) with PC were identified from the Ontario Cancer Registry and SEER-Medicare databases from 2006 to 2015. Demographics and treatment (surgery, radiation, chemotherapy, or multimodality (surgery and chemotherapy)) were described. In resected patients, neoadjuvant therapy, readmission, and 30- and 90-day postoperative mortality rates were calculated. Survival was assessed using Kaplan-Meier curves. RESULTS: This study includes 38,858 and 11,512 patients with PC from the USA and Ontario, respectively. More female patients were identified in the USA (54.0%) versus Ontario (46.9%). In the entire cohort, US patients received more radiation in addition to other therapies (18.8% vs. 13.5% Ontario) and chemotherapy alone (34.3% vs. 19.0% Ontario). While rates of resection were similar (13.4% USA vs.12.5% Ontario), multimodality therapy was more common in the UAS (9.0% vs. 6.4%). Among resected patients, neoadjuvant chemotherapy was uncommon in both groups, although more frequent in the USA (12.0% vs. 3.2% Ontario). The 30- and 90-day postoperative mortality rates were lower in Ontario vs. the USA (30-day: 3.26% vs. 4.91%; 90-day: 7.08% vs. 10.96%), however, overall survival was similar between the USA and Ontario. CONCLUSIONS: We observed substantive differences in treatment and outcomes between PC patients in the USA and Ontario, which may reflect known differences in healthcare systems. Close evaluation of healthcare policies can inform initiatives to improve care quality.

The Influence of Patient Preference on Surgeons' Treatment Recommendations in the Management of Intraductal Papillary Mucinous Neoplasms.

OBJECTIVE: We aimed to determine whether surgeon variation in management of intraductal papillary mucinous neoplasms (IPMN) is driven by differences in risk perception and quantify surgeons' risk threshold for changing their recommendations. BACKGROUND: Surgeons vary widely in management of IPMN. METHODS: We conducted a survey of members of the Americas HepatoPancreatoBiliary Association, presented participants with 2 detailed clinical vignettes and asked them to choose between surgical resection and surveillance. We also asked them to judge the likelihood that the IPMN harbors cancer and that the patient would have a serious complication if surgery was performed. Finally, we asked surgeons to rate the level of cancer risk at which they would change their treatment recommendation. We examined the association between surgeons' treatment recommendations and their risk perception and risk threshold. RESULTS: One hundred and fifty surgeons participated in the study. Surgeons varied in their recommendations for surgery [19% for vignette 1 (V1) and 12% for V2] and in their perception of the cancer risk (interquartile range: 2%-10% for V1 and V2) and risk of surgical complications (V1 interquartile range: 10%-20%, V2 20%-30%). After adjusting for surgeon characteristics, surgeons who were above the median in cancer risk perception were 22 percentage points (27% vs. 5%) more likely to recommend resection than those who were below the median (95% CI: 11.34%; P <0.001). The median risk threshold at which surgeons would change their recommendation was 15% (V1 and V2). Surgeons who recommended surgery had a lower risk threshold for changing their recommendation than those who recommended surveillance (V1: 10.0 vs. 15.0, P =0.06; V2: 7.0 vs. 15.0, P =0.05). CONCLUSIONS: The treatment that patients receive for IPMNs depends greatly on how their surgeons perceive the risk of cancer in the lesion. Efforts to improve cancer risk prediction for IPMNs may lead to decreased variations in care.

Risk Perceptions and Risk Thresholds Among Surgeons in the Management of Intraductal Papillary Mucinous Neoplasms.

OBJECTIVES: We aimed to determine whether surgeon variation in management of intraductal papillary mucinous neoplasm (IPMN) is driven by differences in risk perception and quantify surgeons' risk threshold for changing their recommendations. BACKGROUND: Surgeons vary widely in management of IPMN. METHODS: We conducted a survey of members of the Americas HepatoPancreatoBiliary Association, presented participants with 2 detailed clinical vignettes and asked them to choose between surgical resection and surveillance. We also asked them to judge the likelihood that the IPMN harbors cancer and that the patient would have a serious complication if surgery was performed. Finally, we asked surgeons to rate the level of cancer risk at which they would change their treatment recommendation. We examined the association between surgeons' treatment recommendations and their risk perception and risk threshold. RESULTS: One hundred fifty surgeons participated in the study. Surgeons varied in their recommendations for surgery [19% for vignette 1 (V1) and 12% for V2] and in their perception of the cancer risk (interquartile range: 2%-10% for V1 and V2) and risk of surgical complications (V1 interquartile range: 10%-20%, V2 20-30%). After adjusting for surgeon characteristics, surgeons who were above the median in cancer risk perception were 22 percentage points (27% vs 5%) more likely to recommend resection than those who were below the median (95% CI: 11%-4%; P <0.001). The median risk threshold at which surgeons would change their recommendation was 15% (V1 and V2). Surgeons who recommended surgery had a lower risk threshold for changing their recommendation than those who recommended surveillance (V1: 10.0 vs 15.0, P =0.06; V2: 7.0 vs 15.0, P =0.05). CONCLUSIONS: The treatment that patients receive for IPMNs depends greatly on how their surgeons perceive the risk of cancer in the lesion. Efforts to improve cancer risk prediction for IPMNs may lead to decreased variations in care.

Segmental Versus Diffuse Main Duct Intraductal Papillary Mucinous Neoplasm: Examination of Main Pancreatic Duct Morphology and Implications for Malignancy Risk and Extent of Surgical Resection.

OBJECTIVE: To determine whether the morphologic features of the main pancreatic duct (MPD) of main-duct-involved-intraductal papillary mucinous neoplasm (IPMN) (ie, main duct or mixed main duct/side branch) have implications for the risk of malignancy and extent of resection. BACKGROUND: International consensus guidelines acknowledge the presence of various MPD morphologies (ie, diffuse vs segmental main-duct-involved-IPMN) without a precise definition of each entity and with limited data to guide treatment strategy. METHODS: All consecutive main-duct-involved-IPMN patients (2005-2019) with a MPD diameter ≥5 mm by cross-sectional imaging were reviewed from a prospective institutional database. Morphologic features of the MPD were correlated with the identification of high-grade dysplasia or pancreatic ductal adenocarcinoma (HGD/PDAC) by logistic regression modeling. In patients who underwent partial pancreatectomy, preoperative MPD morphologic features were correlated with the future development of HGD/PDAC in the pancreatic remnant by Cox hazards modeling. RESULTS: In a cohort of 214 main-duct-involved-IPMN patients, the overall rate of HGD/PDAC was 54.2%. MPD morphologic characteristics associated with HGD/PDAC included: maximal MPD diameter (5-10 mm: 29.8%; 10-14 mm: 59.0%; 15-19 mm: 78.6%; ≥20 mm: 95.8%; P <0.001), segmental extent of maximal dilation (<25%: 28.2%; 25%-49%: 54.9%; 50%-74%: 63.1%; ≥75%: 67.9%; P =0.002), and nonsegmental MPD diameter (<5 mm: 21.5% vs ≥5 mm: 78.5%, P <0.001). Diffuse MPD dilation involving ≥90% extent was rare (5.6%). After a median follow-up of 50 months, 7 (7.2%) patients who underwent partial pancreatectomy for IPMN without associated PDAC developed HGD/PDAC in the pancreatic remnant. Maximal MPD diameter, segmental extent of maximal dilation, or nonsegmental MPD diameter were not associated with the development of HGD/PDAC in the pancreatic remnant. However, a mural nodule on preoperative imaging was associated with the development of HGD/PDAC in the pancreatic remnant. CONCLUSIONS: "Diffuse" involvement with homogenous dilation of the MPD was rare. For the majority of patients with segmental main-duct-involved-IPMN, the MPD morphology conferred malignancy risk. Duct morphology was not predictive for the development of HGD or invasive disease in the pancreatic remnant, implying the safety of limited pancreatic resection for initial surgical management.

Surgical Margin of Resected Colorectal Liver Metastases:: How Accurate is Surgeon Prediction?

OBJECTIVE: The aim of this study was to describe the surgeon's ability to accurately predict the margin following resection of colorectal liver metastases (CRLM). SUMMARY BACKGROUND DATA: The decision to resect CRLM is based on the surgeon's ability to predict tumor free resection margins. However, to date, no study has evaluated the accuracy of surgeon margin prediction. METHODS: In this single-institution prospective study, the operating attending and fellow independently completed a preoperative and postoperative questionnaire describing their expected resection margin in 100 consecutive cases (200 assessments) of colorectal liver metastasis resections. In cases with multiple metastases, the closest margin was assessed as the margin of interest for the primary outcome. Surgeon assessments were compared to the gold-standard histopathologic assessment. RESULTS: After excluding aborted cases, 190 preoperative and 190 postoperative assessments from 95 cases were included in the analysis. The pathologic margin was noted to be wide (≥1 cm), 1 mm to 1 cm, narrow (<1 mm), and positive in 28 (29.5%), 55 (57.9%), 5 (5.3%), and 7 (7.4%) cases, respectively. The 88 cases with negative margins were all predicted to be negative. None of the cases with positive margins were predicted to be positive. Ninety-one (48%) preoperative and 104 (55%) postoperative predictions were accurate. The sensitivity of predicting a margin <1 mm was 8.3% preoperatively and 16.7% postoperatively. The positive predictive value for preoperative and postoperative predictions of margin <1 mm was 18.2% and 26.7%, respectively. CONCLUSIONS: Surgeons are inaccurate at predicting positive and close surgical margins following resection of CRLM. A predicted close margin should not necessarily preclude resection.

A Model to Predict Treatment Failure in Patients Undergoing Upfront Surgery for Resectable Colorectal Liver Metastases.

INTRODUCTION: Patients who recur in the first year after resection of colorectal liver metastases (CRLM) do poorly. The aim of our study was to predict treatment failure in patients undergoing upfront resection with a nomogram. METHODS: Data from patients resected between 1991 and 2019 were randomly split (70:30) into two cohorts. Treatment failure was defined as any recurrence or death within 12 months. A nomogram was constructed using multivariable logistic regression on the training cohort and validated using the testing cohort. RESULTS: Overall, 783 patients were included. Primary tumor characteristics included 50% left-sided: 75.2% T3/4 and 56.5% node-positive. The median disease-free interval was 10 months, median number of metastases was 1 (1-50), and with a median size of 3.6 cm (0.2-22); 222 (28.3%) patients recurred within 1 year. Recurrence was mostly extrahepatic with or without liver involvement (150/222, 67.6%). Curative-intent treatment was possible in 37.8% of these patients. Primary location, T-stage and node status, disease-free interval, and number and size of metastases were associated with treatment failure. The area under the curve from the validation of the model was 0.6 (95% confidence interval 0.52-0.68). Patients with a high-risk of treatment failure (≥40%) had a worse survival from the landmark time of 12 months from surgery compared with those with low-risk (2-years: 82% vs. 70%; p = 0.0002). CONCLUSIONS: Primary location, T stage, node status, disease-free interval, and number and size of metastases are associated with treatment failure. The survival of patients with a probability of treatment failure ≥40% is unfavorable. Future trials investigating the role of neoadjuvant therapy in these high-risk patients are warranted.

Rates and Patterns of Recurrence After Microwave Ablation of Colorectal Liver Metastases: A Per Lesion Analysis of 416 Tumors in the Era of 2.45 GHz Generators.

BACKGROUND: For some patients with colorectal liver metastases (CRLMs), surgical resection of all visible disease can lead to long-term survival and even cure. When complete resection is not feasible, microwave ablation (MWA) can help achieve hepatic disease control. As modern 2.45-GHz MWA generators gain popularity, the characteristics of tumors most likely to benefit from this method remain unclear. This study aimed to evaluate local recurrence (LR) rates, patterns of recurrence, and factors contributing to treatment failure after 2.45-GHz MWA of CRLM. METHODS: Patients with CRLM who underwent operative 2.45-GHz MWA between 2011 and 2019 were identified in a prospectively maintained single-institution database. Recurrence outcomes were ascertained for each lesion by imaging review. Factors associated with LR were analyzed. RESULTS: The study enrolled 184 patients bearing 416 ablated tumors. Most of the patients (65.8%) had high clinical risk scores (3-5), and 165 (90%) underwent concurrent liver resection. The median tumor size was 10 mm. After a median follow-up period of 28.8 months, LR was observed in 45 tumors, and the cumulative incidence of LR at 24 months was 10.9% (95% confidence interval [CI], 8.0-14.3%]. In 7%, LR was the first recurrence site, often combined with recurrence elsewhere. The cumulative incidence of LR at 24 months was 6.8% (95% CI 3.8-11.0%) for tumors 10 mm in size or smaller, 12.4% (95% CI 7.8-18.1%) for tumors 11 to 20 mm in size, and 30.2% (95% CI 14.2-48.0%) for tumors larger than 20 mm. In the multivariable analysis, tumors larger than 20 mm with a subcapsular location were significantly associated with increased LR risk. CONCLUSIONS: Treatment of CRLM with 2.45-GHz MWA offers excellent local control at 2 years and is most successful for small tumors deep within the parenchyma.

Combined Hepatic Arterial Infusion Pump and Systemic Chemotherapy in the Modern Era for Chemotherapy-Naive Patients with Unresectable Colorectal Liver Metastases.

PURPOSE: Chemotherapy-naive patients with unresectable colorectal liver metastases (CRLM) have been the best responders to hepatic arterial infusion (HAI) therapy. The current treatment paradigm has drifted away from HAI in the first-line setting. We aimed to analyze outcomes of combined first-line systemic therapy with HAI therapy (HAI+SYS) in the modern era. METHODS: We conducted a retrospective study of consecutive chemotherapy-naive patients with unresectable CRLM who received HAI+SYS between 2003 and 2019. Patients were selected from a prospectively maintained database. Outcomes included radiological response rate, conversion to resection (CTR) rate, and overall survival (OS). RESULTS: Fifty-eight chemotherapy-naive patients were identified out of 546 patients with unresectable CRLM managed with HAI. After induction treatment, 4 patients (7%) had a complete radiological response, including two durable responses. In total, 32 patients (55%) underwent CTR. CTR or complete response without resection was achieved after seven cycles of systemic therapy and four cycles of HAI therapy. Median OS for the whole cohort was 53.0 months (95% confidence interval 23.0-82.9). Three- and 5-year OS in patients who achieved CTR or complete response versus patients who did not was 88% and 72% versus 27% and 0% respectively. Of patients who underwent CTR, complete and major pathological response (no and <10% viable tumor cells, respectively) was observed in 7 (22%) and 12 patients (38%). CONCLUSIONS: Combined HAI+SYS in chemotherapy-naive patients resulted in durable and substantial response in a large proportion of patients. Nearly two-thirds of patients achieved a complete response or proceeded to conversion surgery, which was associated with prolonged survival.

Recurrence-free survival versus overall survival as a primary endpoint for studies of resected colorectal liver metastasis: a retrospective study and meta-analysis.

BACKGROUND: Recurrence-free survival has been used as a surrogate endpoint for overall survival in trials involving patients with resected colorectal liver metastases. We aimed to assess the correlation between recurrence-free survival and overall survival after resection of colorectal liver metastases to determine the adequacy of this surrogate endpoint. METHODS: In this retrospective study and meta-analysis, we compiled an institutional cohort of consecutive patients who had complete resection of colorectal liver metastases from the Memorial Sloan Kettering Cancer Center (New York, NY, USA) prospective database. Patients were eligible for inclusion if they were aged 18 years or older, and underwent hepatectomy, with or without operative ablation, between Jan 1, 1991, and April 30, 2019. We estimated overall survival and recurrence-free survival probabilities at various timepoints using the Kaplan-Meier method, and we assessed pairwise associations between these endpoints using Spearman's rank correlation. We also did a meta-analysis of adjuvant phase 3 clinical trials for colorectal liver metastases to assess the correlation between hazard ratios (HRs) for recurrence-free survival and overall survival. We searched MEDLINE for articles of phase 3 randomised controlled trials analysing adjuvant treatment strategies for resected colorectal metastases from database inception to Jan 1, 2022. The titles and abstracts of identified studies were screened before full-text screening and summary data were either recalculated or extracted manually from the published Kaplan-Meier curves (depending on data availability). FINDINGS: Data were available for 3299 patients in the institutional database, of whom 2983 were eligible for inclusion in our cohort. Median follow-up was 8·4 years (95% CI 7·9-9·1) , during which time there were 1995 (67%) disease recurrences and 1684 (56%) deaths. Median recurrence-free survival was 1·3 years (95% CI 1·3-1·4) and median overall survival was 5·2 years (95% CI 5·0-5·5). 1428 (85%) of 1684 deaths were preceded by recurrence, and median time from recurrence to death was 2·0 years (IQR 1·0-3·4). Pairwise correlations between recurrence-free survival and overall survival were low to moderate, with a correlation estimate ranging from 0·30 (SD 0·17) to 0·56 (0·13). In the meta-analysis of adjuvant clinical trials, the Spearman's correlation coefficient between recurrence-free survival HR and overall survival HR was r=0·20 (p=0·71). INTERPRETATION: We found a minimal correlation between recurrence-free survival and overall survival after resection of colorectal liver metastases. Recurrence-free survival is an inadequate surrogate endpoint for overall survival in this disease setting. FUNDING: US National Cancer Institute.

Genomic Stratification of Resectable Colorectal Liver Metastasis Patients and Implications for Adjuvant Therapy and Survival.

OBJECTIVE: To determine whether genomic risk groups identified by somatic mutation testing of colorectal liver metastasis (CRLM) can be used for "molecularly-guided" selection for adjuvant systemic chemotherapy and hepatic artery infusion of FUDR (SYS+HAI-FUDR). BACKGROUND: Several genomic biomarkers have been associated with clinical phenotype and survival for patients with resectable CRLM. It is unknown whether prognostication afforded by genomic stratification translates into enhanced patient selection for adjuvant hepatic artery infusion therapy. METHODS: Consecutive patients with resected CRLM and available mutational characterization via Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets were reviewed from a prospective institutional database. Patients were stratified into three genomic risk groups based on previously defined alterations in SMAD4, EGFR and the RAS/RAF pathway. The association between SYS+HAI-FUDR and overall survival, relative to adjuvant chemotherapy alone (SYS), was evaluated in each genomic risk group by Cox proportional hazard regression and propensity score matched analyses. RESULTS: A total of 334 patients (SYS+HAI-FUDR 204; SYS 130) were identified; the rates of RAS/RAF alterations and SMAD4 inactivation were 47.4% and 11.7%, respectively. After a median follow-up of 58 months, adjuvant SYS+HAI-FUDR was independently associated with a reduced risk of death (HR 0.50, 95%CI 0.26-0.98, P = 0.045) in the low-risk genomic group, but not in the moderate-risk (HR 1.07, 95%CI 0.5-2.07, P = 0.749) or high-risk (HR 1.62, 95%CI 0.29-9.12, P = 0.537) cohorts. Following propensity score matching, adjuvant SYS+HAI-FUDR remained associated with significant improvements in long-term survival selectively in the low-risk genomic cohort (5-year actuarial survival: 89% vs. 68%, P = 0.019). CONCLUSIONS: Genomic alterations in RAS/RAF, SMAD4, and EGFR may be useful to guide treatment selection in resectable CRLM patients and warrant external validation and integration in future clinical trial design.

Timing of Primary Tumor Resection in Synchronous Metastatic Colon Cancer Patients Undergoing Hepatic Arterial Infusion Pump Placement.

INTRODUCTION: Hepatic arterial infusion (HAI) chemotherapy is associated with improved survival in stage IV colon cancer with liver metastases. Whether simultaneous colon resection and HAI pump (HAIP) placement is associated with increased morbidity has not been specifically studied. The purpose of this study was to compare perioperative outcomes of simultaneous colon resection and HAIP placement versus HAIP placement alone. METHODS: This was a retrospective study of consecutive patients with colon cancer and synchronous liver metastases who underwent HAIP placement between 2007 and 2018. Clinicopathologic characteristics, operative data, complications, and time to first cycle of HAIP chemotherapy were compared between patients who underwent colon resection simultaneously with HAIP placement and those who underwent HAIP placement alone. RESULTS: A total of 258 patients underwent simultaneous colectomy and HAIP placement, and 116 patients underwent HAIP placement alone. Grade 1-2 complications were more common in patients who underwent simultaneous colectomy and HAIP placement (36.8% vs. 19.0%, P < 0.001), but grade 3-4 complications were not observed more frequently (14.3% vs. 12.9%, P = 0.872). The median interval between HAIP placement and start of HAIP chemotherapy did not differ between groups (simultaneous colectomy, 27 days [interquartile range (IQR) 17-34]; HAIP placement alone, 30 days [IQR 21-34]; P = 0.924). Infection of the pump causing either delay of initiation of chemotherapy or explantation of the pump occurred in five patients with simultaneous colectomy and in one patient with HAIP placement alone (P = 0.671). CONCLUSIONS: Simultaneous HAIP implantation and colectomy is safe in patients with liver metastases of colon carcinoma.

Induction FOLFIRINOX for patients with locally unresectable pancreatic ductal adenocarcinoma.

OBJECTIVES: Patients with locally advanced pancreatic adenocarcinoma (PDAC) receive induction chemotherapy with or without radiation, with the goal of R0 resection and improving survival. Herein, we evaluate the outcomes of patients who presented with Stage III PDAC and received induction FOLFIRINOX. METHODS: An institutional database was queried for consecutive patients who received induction FOLFIRINOX for locally unresectable PDAC between 2010 and 2016. Clinical and radiographic parameters were assessed pre- and posttreatment, and clinical outcomes were evaluated. RESULTS: There were 200 patients who met the inclusion criteria. The median number of cycles of FOLFIRINOX was 8, 70% (n = 140) received radiation, and 18% (n = 36) underwent resection. Median overall survival (OS) in resected patients was 36 months (95% confidence interval [CI]: 24-56), and this group had improved OS compared to patients that did not undergo resection (hazard ratio (95% CI): 0.41 (0.26-0.64), p < 0.001). Patients (n = 112) who did not progress on induction therapy but remained unresectable had a median OS of 23.9 months (95% CI: 21.1-25.4). CONCLUSION: Nearly 20% of patients with locally advanced PDAC responded sufficiently to induction FOLFIRINOX to undergo resection, which was associated with improved OS compared to patients that did not undergo resection. Patients with stable disease who remain unresectable represent a group of patients with locally advanced PDAC who may benefit from optimization of additional nonoperative treatment.

Machine learning radiomics can predict early liver recurrence after resection of intrahepatic cholangiocarcinoma.

BACKGROUND: Most patients recur after resection of intrahepatic cholangiocarcinoma (IHC). We studied whether machine-learning incorporating radiomics and tumor size could predict intrahepatic recurrence within 1-year. METHODS: This was a retrospective analysis of patients with IHC resected between 2000 and 2017 who had evaluable computed tomography imaging. Texture features (TFs) were extracted from the liver, tumor, and future liver remnant (FLR). Random forest classification using training (70.3%) and validation cohorts (29.7%) was used to design a predictive model. RESULTS: 138 patients were included for analysis. Patients with early recurrence had a larger tumor size (7.25 cm [IQR 5.2-8.9] vs. 5.3 cm [IQR 4.0-7.2], P = 0.011) and a higher rate of lymph node metastasis (28.6% vs. 11.6%, P = 0.041), but were not more likely to have multifocal disease (21.4% vs. 17.4%, P = 0.643). Three TFs from the tumor, FD1, FD30, and IH4 and one from the FLR, ACM15, were identified by feature selection. Incorporation of TFs and tumor size achieved the highest AUC of 0.84 (95% CI 0.73-0.95) in predicting recurrence in the validation cohort. CONCLUSION: This study demonstrates that radiomics and machine-learning can reliably predict patients at risk for early intrahepatic recurrence with good discrimination accuracy.

Treatment patterns and survival in patients with early-onset pancreatic cancer.

BACKGROUND: Pancreatic cancer is uncommon in patients younger than 50 years, although its incidence is increasing. This study characterizes treatment utilization for early-onset pancreatic cancer (EOPC) versus average-age-onset pancreatic cancer (AOPC) and identifies factors associated with failure to receive treatment. METHODS: The National Cancer Data Base (NCDB) was queried for patients with EOPC (age < 50 years) or AOPC (age ≥ 50 years) from 2004 to 2016. Multinomial regression was used to compare utilization (single modality vs multimodal treatment with or without surgery vs no treatment) between EOPC and AOPC. Kaplan-Meier methods were used to estimate overall survival (OS). RESULTS: Of 248,634 patients, 15,710 (6.3%) had EOPC. There were more male patients (56% vs 50%), non-White patients, and privately insured patients (61% vs 30%) with EOPC versus AOPC, without notable differences in clinical stage distribution. Patients with EOPC received more chemotherapy (38% vs 29%), surgery (9% vs 6.9%), chemoradiation (12% vs 9.2%), and multimodal treatment (21% vs 15%). The odds of receiving multimodal curative therapy were significantly higher for patients with EOPC versus patients with AOPC after adjustments for confounders (odds ratio, 3.89; 95% confidence interval [CI], 3.66-4.15; P < .001). Nineteen percent of patients with EOPC, in contrast to 39% of patients with AOPC, received no treatment. Patients with AOPC more frequently declined chemotherapy (15% vs 9.5%). One-year OS was higher for EOPC versus AOPC across each stage (0/I/II, 72% [95% CI, 71%-74%] vs 53% [95% CI, 53%-54%]; III, 48% [95% CI, 45%-50%] vs 38% [95% CI, 37%-38%]; IV, 25% [95% CI, 24%-26%] vs 15% [95% CI, 15%-15%]) and treated patients (0/I/II, 75% [95% CI, 74%-77%] vs 64% [95% CI, 63%-64%]; III, 51% [95% CI, 49%-54%] vs 47% [95% CI, 47%-48%]; IV, 29% [95% CI, 28%-31%] vs 23% [95% CI, 23%-24%]). CONCLUSIONS: Patients with EOPC receive more oncologic therapy than patients with AOPC, although the intensity, type, and duration of chemotherapy are not available in the NCDB; however, 19% and 39%, respectively, receive no therapy. Underutilization may explain suboptimal oncologic outcomes. Efforts to improve access and treatment utilization in all age groups are warranted.

Local Control and Survival After Induction Chemotherapy and Ablative Radiation Versus Resection for Pancreatic Ductal Adenocarcinoma With Vascular Involvement.

OBJECTIVE: We sought to compare overall survival (OS) and disease control for patients with localized pancreatic ductal adenocarcinoma (PDAC) treated with ablative dose radiotherapy (A-RT) versus resection. SUMMARY BACKGROUND DATA: Locoregional treatment for PDAC includes resection when possible or palliative RT. A-RT may offer durable tumor control and encouraging survival. METHODS: This was a single-institution retrospective analysis of patients with PDAC treated with induction chemotherapy followed by A-RT [≥98 Gy biologically effective dose (BED) using 15-25 fractions in 3-4.5 Gy/fraction] or pancreatectomy. RESULTS: One hundred and four patients received A-RT (49.8%) and 105 (50.2%) underwent resection. Patients receiving A-RT had larger median tumor size after induction chemotherapy [3.2 cm (undetectable-10.9) vs 2.6 cm (undetectable-10.7), P < 0.001], and were more likely to have celiac or hepatic artery encasement (48.1% vs 11.4%, P <0.001), or superior mesenteric artery encasement (43.3% vs 9.5%, P < 0.001); however, there was no difference in the degree of SMV/PV involvement (P = 0.123). There was no difference in locoregional recurrence/progression at 18-months between A-RT and resection; cumulative incidence was 16% [(95% confidence interval (CI) 10%-24%] versus 21% (95% CI 14%-30%), respectively (P= 0.252). However, patients receiving A-RT had a 19% higher 18-month cumulative incidence of distant recurrence/progression [58% (95% CI 48%-67%) vs 30% (95% CI 30%-49%), P= 0.004]. Median OS from completion of chemotherapy was 20.1 months for A-RT patients (95% CI 16.4-23.1 months) versus 32.9 months (95% CI 29.7-42.3 months) for resected patients (P < 0.001). CONCLUSION: Ablative radiation is a promising new treatment option for PDAC, offering locoregional disease control similar to that associated with resection and encouraging survival.

The Landmark Series: Hilar Cholangiocarcinoma.

Hilar cholangiocarcinoma (HC) is a rare and highly aggressive biliary tract neoplasm. As such, the data driving the management of this disease generally are not based on prospective clinical trial data but rather consist of retrospective experiences and limited level 1 data. Surgical resection offers the best chance of a long-term survival, but local and distant recurrences are common. This report presents landmark articles that form the basis of preoperative, operative, and adjuvant strategies for HC.