Search

Pancreatic cancer is an aggressive disease that typically presents late with poor outcomes, indicating a pronounced need for early detection. In this study, we applied artificial intelligence methods to clinical data from 6 million patients (24,000 pancreatic cancer cases) in Denmark (Danish National Patient Registry (DNPR)) and from 3 million patients (3,900 cases) in the United States (US Veterans Affairs (US-VA)). We trained machine learning models on the sequence of disease codes in clinical histories and tested prediction of cancer occurrence within incremental time windows (CancerRiskNet). For cancer occurrence within 36 months, the performance of the best DNPR model has area under the receiver operating characteristic (AUROC) curve = 0.88 and decreases to AUROC (3m) = 0.83 when disease events within 3 months before cancer diagnosis are excluded from training, with an estimated relative risk of 59 for 1,000 highest-risk patients older than age 50 years. Cross-application of the Danish model to US-VA data had lower performance (AUROC = 0.71), and retraining was needed to improve performance (AUROC = 0.78, AUROC (3m) = 0.76). These results improve the ability to design realistic surveillance programs for patients at elevated risk, potentially benefiting lifespan and quality of life by early detection of this aggressive cancer.

Pancreatic ductal adenocarcinoma (PDAC) is the third most lethal cancer in the United States, with a 5-year life expectancy of 11%. Most symptoms manifest at an advanced stage of the disease when surgery is no longer appropriate. The dire prognosis of PDAC warrants new strategies to improve the outcomes of patients, and early detection has garnered significant attention. However, early detection of PDAC is most often incidental, emphasizing the importance of developing new early detection screening strategies. Due to the low incidence of the disease in the general population, much of the focus for screening has turned to individuals at high risk of PDAC. This enriches the screening population and balances the risks associated with pancreas interventions. The cancers that are found in these high-risk individuals by MRI and/or EUS screening show favorable 73% 5-year overall survival. Even with the emphasis on screening in enriched high-risk populations, only a minority of incident cancers are detected this way. One strategy to improve early detection outcomes is to integrate artificial intelligence (AI) into biomarker discovery and risk models. This expert review summarizes recent publications that have developed AI algorithms for the applications of risk stratification of PDAC using radiomics and electronic health records. Furthermore, this review illustrates the current uses of radiomics and biomarkers in AI for early detection of PDAC. Finally, various challenges and potential solutions are highlighted regarding the use of AI in medicine for early detection purposes.

Immunotherapy with checkpoint inhibitors, such as the programmed death-1 (PD-1) antibodies pembrolizumab and nivolumab, are effective in a variety of tumors, yet not all patients respond. Tumor microsatellite instability-high (MSI-H) has emerged as a biomarker of response to checkpoint blockade, leading to the tissue agnostic approval of pembrolizumab in MSI-H cancers. Here we describe a patient with MSI-H colorectal cancer that was treated with this immune checkpoint inhibitor and exhibited progression of disease. We examined this intrinsic resistance through genomic, transcriptional, and pathologic characterization of the patient's tumor and the associated immune microenvironment. The tumor had typical MSI-H molecular features, including a high neoantigen load. We also identified biallelic loss of the gene for β2-microglobulin (B2M), whose product is critical for antigen presentation. Immune infiltration deconvolution analysis of bulk transcriptome data from this anti-PD-1–resistant tumor and hundreds of other colorectal cancer specimens revealed a high natural killer cell and M2 macrophage infiltration in the patient's cancer. This was confirmed by single-cell transcriptome analysis and multiplex immunofluorescence. Our study provides insight into resistance in MSI-H tumors and suggests immunotherapeutic strategies in additional genomic contexts of colorectal cancer.

Table S1

Supplementary Figures 1 and 2

Table S2

MAF file for copy number variations identified in the cohort

Treatment history for patients enrolled on the PancSeq protocol.

Supplementary Figure S1: Overview of workflow and data generation. Supplementary Figure S2: Analysis of neoplastic cellularity. Supplementary Figure S3: Recurrent copy number alterations. Supplementary Figure S4: Mutational signature analysis. Supplementary Figure S5: Analysis of PDAC gene expression signatures. Supplementary Figure S6: Clinically relevant alterations in the cohort.

Supplementary Experimental Methods

MAF file for mutations identified in the cohort

Objective To assess prevalence of CT imaging-derived sarcopenia, osteoporosis, and visceral obesity in clinically frail and prefrail patients and determine their association with the diagnosis of frailty. Materials and methods This cross-sectional study was constructed using our institution’s pelvic trauma registry and ambulatory database registry. The study included all elderly pelvic trauma patients and ambulatory outpatients between May 2016 and March 2020 who had a comprehensive geriatric assessment and CT abdomen/pelvis within 1 year from the date of the assessment. Patients were dichotomized in prefrail or frail groups. The study excluded patients with history of metastatic disease or malignancy requiring chemotherapy. Results The study cohort consisted of 151 elderly female and 65 male patients. Each gender population was subdivided into frail (114 female [75%], 51 male [78%]) and prefrail (37 female [25%], 14 male [22%]) patients. CT-imaging-derived diagnosis of osteoporosis (odds ratio, 2.5; 95% CI: 1.2–5.5) and sarcopenia (odds ratio, 2.6; 95% CI: 1.2–5.6) were associated with frailty in females, but did not reach statistical significance in males. BMI and subcutaneous adipose tissue at L3 level were statistically lower in the frail male group compared to the prefrail group. BMI showed strong correlation with the subcutaneous area at the L3 level in both genders (Spearman’s coefficient of 0.8, p

Background: Low muscle mass (quantity) is common in patients with advanced cancer, but little is known about muscle radiodensity (quality). We sought to describe the associations of muscle mass and radiodensity with symptom burden, healthcare use, and survival in hospitalized patients with advanced cancer. Methods: We prospectively enrolled hospitalized patients with advanced cancer from September 2014 through May 2016. Upon admission, patients reported their physical (Edmonton Symptom Assessment System [ESAS]) and psychological (Patient Health Questionnaire-4 [PHQ-4]) symptoms. We used CT scans performed per routine care within 45 days before enrollment to evaluate muscle mass and radiodensity. We used regression models to examine associations of muscle mass and radiodensity with patients’ symptom burden, healthcare use (hospital length of stay and readmissions), and survival. Results: Of 1,121 patients enrolled, 677 had evaluable muscle data on CT (mean age, 62.86 ± 12.95 years; 51.1% female). Older age and female sex were associated with lower muscle mass (age: B, –0.16; PPPP=.014), and higher BMI was associated with higher muscle mass (B, 0.58; PPPP=.016), ESAS-Total (B, –0.29; P=.002), PHQ-4-Depression (B, –0.03; P=.006), and PHQ-4-Anxiety (B, –0.03; P=.008) symptoms, as well as decreased hospital length of stay (B, –0.07; P=.005), risk of readmission or death in 90 days (odds ratio, 0.97; PPConclusions: Although muscle mass (quantity) only correlated with survival, we found that muscle radiodensity (quality) was associated with patients’ symptoms, healthcare use, and survival. These findings underscore the added importance of assessing muscle quality when seeking to address adverse muscle changes in oncology.

Background Although CT-based body composition (BC) metrics may inform disease risk and outcomes, obtaining these metrics has been too resource intensive for large-scale use. Thus, population-wide distributions of BC remain uncertain. Purpose To demonstrate the validity of fully automated, deep learning BC analysis from abdominal CT examinations, to define demographically adjusted BC reference curves, and to illustrate the advantage of use of these curves compared with standard methods, along with their biologic significance in predicting survival. Materials and Methods After external validation and equivalency testing with manual segmentation, a fully automated deep learning BC analysis pipeline was applied to a cross-sectional population cohort that included any outpatient without a cardiovascular disease or cancer who underwent abdominal CT examination at one of three hospitals in 2012. Demographically adjusted population reference curves were generated for each BC area. The

Approximately 8-10% of pancreatic cancers do not harbor mutations in KRAS. Understanding the unique molecular and clinical features of this subset of pancreatic cancer (PC) is important to guide patient stratification for clinical trials of molecularly targeted agents. To this end we investigated a cohort of 795 PC patients from Dana-Farber Cancer Institute who had undergone somatic genomic characterization with OncoPanel, a targeted next-generation sequencing panel with coverage of more than 400 cancer-associated genes. A total of 9.2% (73/795) of cases in our cohort were KRAS WT. The KRAS WT cohort was statistically enriched for MSI-H PC and acinar cell carcinomas (p = 0.0035, p < 0.0001 respectively). Actionable alterations in alternative MAPK drivers were identified in 44% (32/73) of KRAS WT cases. BRAF alterations accounted for 56% (18/32) of detected alternative MAPK drivers, the majority of which (72%) were Class II which exhibit dimer-dependent constitutive activity. Receptor Tyrosine Kinase (RTK) fusion events in BRAF, NTRK1, NRG1, NTRK3, ROS1, and FGFR2 accounted for 25% (8/32) of detected alternative MAPK drivers in KRAS WT tumors. BRAF in-frame deletions showed increased sensitivity to dual pan-RAF and MEK inhibition in organoid models and one patient with a ROS1 fusion received prolonged clinical benefit from targeted therapy. In addition to alternative MAPK drivers, mutations in GNAS (p = 0.0014) and ARID2 (p = 0.045) were significantly enriched in KRAS WT PC, whereas TP53 mutations were significantly less frequent in KRAS WT cases. Interestingly, although not statistically significant, rates of mutation in the other canonical tumor suppressor genes (CDKN2A, SMAD4) were also lower in KRAS WT PC. Clinically, KRAS Mutant (MUT) PC were associated with a decreased overall survival (OS) compared to the KRAS WT cohort [median OS 17.5 vs 24.0 months, HR 1.38, p = 0.036], however this relationship was no longer significant after accounting for other clinical factors. For patients with KRAS WT PC, those with SMAD4 alterations had a significantly decreased OS (HR 6.24, p < 0.001), whereas presence of TP53 or CDKN2A mutations had no significant impact. Lastly, we found that KRAS WT PC was associated with a younger age of onset. Interestingly, we noted that KRAS WT PC patients with a younger age of onset had tumors with few oncogenic alterations whereas no such association was seen in KRAS MUT patients. Validation of this finding in a separate dataset is required and is currently ongoing. In summary, our clinical and genomic characterization of KRAS WT PC identifies a high prevalence of alternative MAPK drivers that are amenable to targeted therapies. Our cohort also recapitulates the previously reported clinical characteristics of KRAS WT PC and identifies the presence of SMAD4 alterations as significantly associated with decreased overall survival in KRAS WT PC. Additional analysis from multiple sources will be critical to risk stratify these patients further and to validate age-related findings. Citation Format: Harshabad Singh, Rachel B. Keller, Kevin S. Kapner, Julien Dilly, Srivatsan Raghavan, Chen Yuan, Eizabeth Cohen, Michael Tolstorukov, Emma Hill, Elizabeth Andrews, Lauren K. Brais, Annacarolina Da Silva, Kimberly Perez, Douglas A. Rubinson, Benjamin L. Schlechter, Michael H. Rosenthal, Jason L. Hornick, Valentina Nardi, Yvonne Li, Hersh Gupta, Andrew Cherniack, Mathew L. Meyerson, James M. Cleary, Jonathan A. Nowak, Brian M. Wolpin, Andrew A. Aguirre. Clinical-genomic analysis of KRAS wild-type pancreatic cancer confirms alternative targetable drivers and provides insight for age and risk related clinical stratification [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer; 2022 Sep 13-16; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2022;82(22 Suppl):Abstract nr A001.

Neoadjuvant therapy has shown value in various cancer types. The role of neoadjuvant therapy in pancreatic ductal adenocarcinoma (PDAC), however, remains unknown. The aim of the present work is to evaluate the effect of neoadjuvant therapy on the survival of patients with borderline-resectable PDAC. Between 2004 and 2015, 7730 patients with resectable PDAC and 1980 patients with borderline-resectable PDAC were identified from the National Cancer Database (NCDB). Survival was compared between resectable and borderline-resectable patients. Survival and pathologic characteristics were also compared within borderline-resectable patients who received neoadjuvant therapy and those who received adjuvant therapy alone. Kaplan–Meier method and Cox proportional-hazard models were used for analysis. Median overall survival (mOS) of all patients with resectable PDAC was similar to that of patients with borderline-resectable disease treated with neoadjuvant therapy (26.5 versus 25.7 months, p = 0.78). Patients with borderline-resectable disease treated with neoadjuvant therapy had improved mOS compared with borderline-resectable patients treated with adjuvant therapy alone (25.7 versus 19.6 months, p

There has been a rapid increase in the utilization of MRI in rectal cancer staging in the USA essentially replacing endorectal ultrasound and mimicking the trend in Europe seen in the 1990s and 2000s. Accompanying this trend, there is a demand, and recognized need, for greater precision and clarification of confusing, misunderstood and poorly understood concepts, facts, statements and nomenclature regarding rectal cancer and the use of pelvic MRI for diagnosis. As such, this Review, part evidence-based and part expert opinion, will attempt to elucidate and clarify several concepts the authors have encountered in 25 years of imaging rectal cancer, focusing on MRI.

Pancreatic cancer is an aggressive disease that typically presents late with poor patient outcomes. There is a pronounced medical need for early detection of pancreatic cancer, which can be addressed by identifying high-risk populations. Here we apply artificial intelligence (AI) methods to a dataset of 6 million patient records with 24,000 pancreatic cancer cases in the Danish National Patient Registry (DNPR) and, for comparison, a dataset of three million records with 3,900 pancreatic cancer cases in the United States Department of Veterans Affairs (US-VA) healthcare system. In contrast to existing methods that do not use temporal information, we explicitly train machine learning models on the time sequence of diseases in patient clinical histories and test the ability to predict cancer occurrence in time intervals of 3 to 60 months after risk assessment.For cancer occurrence within 36 months, the performance of the best model (AUROC=0.88, DNPR), trained and tested on disease trajectories, exceeds that of a model without longitudinal information (AUROC=0.85, DNPR). Performance decreases when disease events within a 3 month window before cancer diagnosis are excluded from training (AUROC[3m]=0.83). Independent training and testing on the US-VA dataset reaches comparable performance (AUROC=0.78, AUROC[3m]=0.76). These results raise the state-of-the-art level of performance of cancer risk prediction on real-world data sets and provide support for the design of prediction-surveillance programs based on risk assessment in a large population followed by affordable surveillance of a relatively small number of patients at highest risk. Use of AI on real-world clinical records has the potential to shift focus from treatment of late-stage to early-stage cancer, benefiting patients by improving lifespan and quality of life.

T-cell/histiocyte-rich large B-cell lymphoma (TCRLBCL) is an aggressive variant of diffuse large B-cell lymphoma (DLBCL) characterized by rare malignant B cells within a robust but ineffective immune cell infiltrate. The mechanistic basis of immune escape in TCRLBCL is poorly defined and not targeted therapeutically. We performed a genetic and quantitative spatial analysis of the PD-1/PD-L1 pathway in a multi-institutional cohort of TCRLBCLs and found that malignant B cells harbored PD-L1/PD-L2 copy gain or amplification in 64% of cases, which was associated with increased PD-L1 expression (P = .0111). By directed and unsupervised spatial analyses of multiparametric cell phenotypic data within the tumor microenvironment, we found that TCRLBCL is characterized by tumor-immune “neighborhoods” in which malignant B cells are surrounded by exceptionally high numbers of PD-L1–expressing TAMs and PD-1+ T cells. Furthermore, unbiased clustering of spatially resolved immune signatures distinguished TCRLBCL from related subtypes of B-cell lymphoma, including classic Hodgkin lymphoma (cHL) and DLBCL-NOS. Finally, we observed clinical responses to PD-1 blockade in 3 of 5 patients with relapsed/refractory TCRLBCL who were enrolled in clinical trials for refractory hematologic malignancies (NCT03316573; NCT01953692), including 2 complete responses and 1 partial response. Taken together, these data implicate PD-1 signaling as an immune escape pathway in TCRLBCL and also support the potential utility of spatially resolved immune signatures to aid the diagnostic classification and immunotherapeutic prioritization of diverse tumor types. Key Points: • Spatially resolved signatures of PD-1/PD-L1 signaling in the tumor microenvironment define T-cell/histiocyte-rich large B-cell lymphoma. • Three of 5 patients with relapsed/refractory TCRLBCL showed objective clinical responses to single-agent PD-1 blockade (pembrolizumab).

Despite considerable research efforts, pancreatic cancer is associated with a dire prognosis and a 5-year survival rate of only 10%. Early symptoms of the disease are mostly nonspecific. The premise of improved survival through early detection is that more individuals will benefit from potentially curative treatment. Artificial intelligence (AI) methodology has emerged as a successful tool for risk stratification and identification in general health care. In response to the maturity of AI, Kenner Family Research Fund conducted the 2020 AI and Early Detection of Pancreatic Cancer Virtual Summit (www.pdac-virtualsummit.org) in conjunction with the American Pancreatic Association, with a focus on the potential of AI to advance early detection efforts in this disease. This comprehensive presummit article was prepared based on information provided by each of the interdisciplinary participants on one of the 5 following topics: Progress, Problems, and Prospects for Early Detection; AI and Machine Learning; AI and Pancreatic Cancer-Current Efforts; Collaborative Opportunities; and Moving Forward-Reflections from Government, Industry, and Advocacy. The outcome from the robust Summit conversations, to be presented in a future white paper, indicate that significant progress must be the result of strategic collaboration among investigators and institutions from multidisciplinary backgrounds, supported by committed funders.

Background: Weight loss is a common symptom in patients with cancer. However, little is known about how much weight loss and over what time period weight loss precedes cancer diagnosis. Herein, we evaluate the association between weight loss and subsequent 2-year cancer diagnosis using repeatedly measured weight data in two large prospective cohorts. Methods: The analysis included 111,891 US women from the Nurses’ Health Study and 45,498 US men from the Health Professional Follow-up Study, who were enrolled in 1978 and 1988, respectively, and followed through 2012, and who were cancer-free and ≥40 years old at study entry. Current weight and lifestyle information were assessed every 2 years by questionnaire, and weight change in the previous 2 years was calculated by comparing two consecutive questionnaires. Cox proportional hazards regression was used to estimate hazard ratios (HRs) and 95% confidence intervals (95% CIs) for subsequent risk of cancer diagnosis according to weight change percentage. We also examined 20-year trajectories of body mass index (BMI) before cancer diagnosis. Results: During 4.1 million person-years of follow-up, we documented 30,441 incident cancers. Participants with recent weight loss were at increased risk for cancer diagnosis in the subsequent 2 years (Ptrend 10% had a multivariable-adjusted HR for total cancer of 1.09 (95% CI, 1.04-1.13) and 1.17 (95% CI, 1.10-1.24), respectively. For individual cancer types, increased weight loss was associated with a subsequent cancer diagnosis of esophagus, liver, pancreas, leukemia, stomach, myeloma, lung, colorectal, and non-Hodgkin lymphoma (all Ptrend 10% to those without weight loss, multivariable-adjusted HRs were 4.22 (95% CI, 2.50-7.13) for esophageal cancer, 3.30 (95% CI, 1.65-6.61) for liver cancer, 2.46 (95% CI, 1.84-3.30) for pancreatic cancer, 2.40 (95% CI, 1.59-3.61) for leukemia, 2.26 (95% CI, 1.32-3.87) for stomach cancer, 1.80 (95% CI, 1.14-2.83) for myeloma, 1.45 (95% CI, 1.22-1.73) for lung cancer, 1.38 (95% CI, 1.14-1.66) for colorectal cancer, and 1.35 (95% CI, 1.06-1.72) for non-Hodgkin lymphoma. The 20-year trajectory of BMI indicated that most weight loss occurred in the 2 years before cancer diagnosis, with the magnitude of weight loss increasing closer to the time of subsequent cancer diagnosis. Conclusion: Recent weight loss of >10% predicts elevated risk of cancer diagnosis in the forthcoming 2 years, particularly for cancers of the upper gastrointestinal tract, hematological system, colorectum, and lung. Citation Format: Qiao-Li Wang, Ana Babic, Michael H. Rosenthal, Alice A. Lee, Yin Zhang, Xuehong Zhang, Mingyang Song, Leandro F. Rezende, Dong Hoon Lee, Leah Biller, Kimmie Ng, Andrew T. Chan, Meir J. Stampfer, Edward L. Giovannucci, Peter Kraft, Chen Yuan, Brian M. Wolpin. Weight loss and subsequent cancer diagnosis: A prospective cohort study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5934.

With vast interest in machine learning applications, more investigators are proposing to assemble large datasets for machine learning applications. We aim to delineate multiple possible roadblocks to exam retrieval that may present themselves and lead to significant time delays. This HIPAA-compliant, institutional review board–approved, retrospective clinical study required identification and retrieval of all outpatient and emergency patients undergoing abdominal and pelvic computed tomography (CT) at three affiliated hospitals in the year 2012. If a patient had multiple abdominal CT exams, the first exam was selected for retrieval (n=23,186). Our experience in attempting to retrieve 23,186 abdominal CT exams yielded 22,852 valid CT abdomen/pelvis exams and identified four major categories of challenges when retrieving large datasets: cohort selection and processing, retrieving DICOM exam files from PACS, data storage, and non-recoverable failures. The retrieval took 3 months of project time and at minimum 300 person-hours of time between the primary investigator (a radiologist), a data scientist, and a software engineer. Exam selection and retrieval may take significantly longer than planned. We share our experience so that other investigators can anticipate and plan for these challenges. We also hope to help institutions better understand the demands that may be placed on their infrastructure by large-scale medical imaging machine learning projects.

Skeletal muscle and adipose tissue express the vitamin D receptor and may be a mechanism through which vitamin D supplementation slows cancer progression and reduces cancer death. In this exploratory analysis of a double-blind, multicenter, randomized phase II clinical trial, 105 patients with advanced or metastatic colorectal cancer who were receiving chemotherapy were randomized to either high-dose vitamin D3 (4000 IU) or standard-dose (400 IU) vitamin D3. Body composition was measured with abdominal computed tomography at enrollment (baseline) and after cycle 8 of chemotherapy (16 weeks). As compared with standard-dose vitamin D3, high-dose vitamin D3 did not significantly change body weight [&minus, 0.7 kg, (95% CI: &minus, 3.5, 2.0)], body mass index [&minus, 0.2 kg/m2, 1.2, 0.7)], muscle area [&minus, 1.7 cm2, 9.6, 6.3)], muscle attenuation [&minus, 0.4 HU, 4.2, 3.2)], visceral adipose tissue area [&minus, 7.5 cm2, 24.5, 9.6)], or subcutaneous adipose tissue area [&minus, 8.3 cm2, 35.5, 18.9)] over the first 8 cycles of chemotherapy. Among patients with advanced or metastatic colorectal cancer, the addition of high-dose vitamin D3, vs standard-dose vitamin D3, to standard chemotherapy did not result in any changes in body composition.

Artificial or augmented intelligence, machine learning, and deep learning will be an increasingly important part of clinical practice for the next generation of radiologists. It is therefore critical that radiology residents develop a practical understanding of deep learning in medical imaging. Certain aspects of deep learning are not intuitive and may be better understood through hands-on experience; however, the technical requirements for setting up a programming and computing environment for deep learning can pose a high barrier to entry for individuals with limited experience in computer programming and limited access to GPU-accelerated computing. To address these concerns, we implemented an introductory module for deep learning in medical imaging within a self-contained, web-hosted development environment. Our initial experience established the feasibility of guiding radiology trainees through the module within a 45-min period typical of educational conferences.

BackgroundPreviously, we characterized subtypes of pancreatic ductal adenocarcinoma (PDAC) on computed-tomography (CT) scans, whereby conspicuous (high delta) PDAC tumors are more likely to have aggressive biology and poorer clinical outcomes compared to inconspicuous (low delta) tumors. Here, we hypothesized that these imaging-based subtypes would exhibit different growth-rates and distinctive metabolic effects in the period prior to PDAC diagnosis.Materials and methodsRetrospectively, we evaluated 55 patients who developed PDAC as a second primary cancer and underwent serial pre-diagnostic (T0) and diagnostic (T1) CT-scans. We scored the PDAC tumors into high and low delta on T1 and, serially, obtained the biaxial measurements of the pancreatic lesions (T0-T1). We used the Gompertz-function to model the growth-kinetics and estimate the tumor growth-rate constant (α) which was used for tumor binary classification, followed by cross-validation of the classifier accuracy. We used maximum-likelihood estimation to estimate initiation-time from a single cell (10-6 mm3) to a 10 mm3 tumor mass. Finally, we serially quantified the subcutaneous-abdominal-fat (SAF), visceral-abdominal-fat (VAF), and muscles volumes (cm3) on CT-scans, and recorded the change in blood glucose (BG) levels. T-test, likelihood-ratio, Cox proportional-hazards, and Kaplan-Meier were used for statistical analysis and p-value ResultsCompared to high delta tumors, low delta tumors had significantly slower average growth-rate constants (0.024 month−1 vs. 0.088 month−1, p−1. Leave-one-out-cross-validation showed 80% accuracy in predicting the delta-class (AUC=0.84). High delta tumors exhibited accelerated SAF, VAF, and muscle wasting (p ConclusionImaging-based subtypes of PDAC exhibit distinct growth, metabolic, and clinical profiles during the pre-diagnostic period. Our results suggest that heterogeneous disease biology may be an important consideration in early detection strategies for PDAC.

Artificial intelligence and machine learning (AI-ML) have taken center stage in medical imaging. To develop as leaders in AI-ML, radiology residents may seek a formative data science experience. The authors piloted an elective Data Science Pathway (DSP) for 4th-year residents at the authors' institution in collaboration with the MGH & BWH Center for Clinical Data Science (CCDS). The goal of the DSP was to provide an introduction to AI-ML through a flexible schedule of educational, experiential, and research activities. The study describes the initial experience with the DSP tailored to the AI-ML interests of three senior radiology residents. The authors also discuss logistics and curricular design with common core elements and shared mentorship. Residents were provided dedicated, full-time immersion into the CCDS work environment. In the initial DSP pilot, residents were successfully integrated into AI-ML projects at CCDS. Residents were exposed to all aspects of AI-ML application development, including data curation, model design, quality control, and clinical testing. Core concepts in AI-ML were taught through didactic sessions and daily collaboration with data scientists and other staff. Work during the pilot period led to 12 accepted abstracts for presentation at national meetings. The DSP is a feasible, well-rounded introductory experience in AI-ML for senior radiology residents. Residents contributed to model and tool development at multiple stages and were academically productive. Feedback from the pilot resulted in establishment of a formal AI-ML curriculum for future residents. The described logistical, planning, and curricular considerations provide a framework for DSP implementation at other institutions. Supplemental material is available for this article. © RSNA, 2020.

Purpose The anatomic variability of the hepatic artery has important implications in the treatment of periampullary cancers. The purpose is of this report is to describe the importance of the replaced right hepatic artery originating from the gastroduodenal artery, which is typically removed during a pancreatic head resection, and to demonstrate how this variant can markedly alter treatment planning and intraoperative decision making. Materials and methods In this report, two cases of a replaced right hepatic artery originating from the gastroduodenal artery are described in the setting of a borderline-resectable pancreatic adenocarcinoma identified on preoperative CT imaging. Results The importance of identifying this variant preoperatively is highlighted. In addition, the implications involved in preserving the replaced right hepatic artery are emphasized, requiring preservation of a portion of the gastroduodenal artery within the pancreatic parenchyma during pancreaticoduodenectomy. Conclusion The location of this variant in the setting of a pancreatic head malignancy mandates its attention by a multidisciplinary team of radiologists, radiation oncologists, medical oncologists, and surgeons, prior to pancreaticoduodenectomy.

INTRODUCTION Upgrading computerized tomography (CT) scanners to iterative reconstruction techniques (IRT) decreases radiation dose. This reduction, combined with changes in surveillance protocols in clinical stage I testicular cancer (CS1TC) measurably decrease the lifetime attributable risk (LAR) of dying of radiation-associated cancer. MATERIALS AND METHODS This IRB-approved study enrolled 24 CS1TC patients who had CT scans on the same Toshiba Aquilion 64 CT before and after IRT software installation. Dose-length product and CT dose index volume were recorded. A physicist calculated effective doses. Radiation doses were compared using the Wilcoxon signed rank test. Median effective dose per scan was multiplied by scan number based on 16 versus 7 scans in 5-year AS protocols to calculate estimated cumulative dose (ECD). LAR of dying of radiation-associated solid tumor was estimated using ECD for a single exposure at age 35 with the excess absolute risk transport model from the BEIR VII analysis of long-term atomic bomb survivors. RESULTS Median preupgrade and postupgrade effective doses were 12.5 and 7.7 mSv, respectively (P

Clinically relevant subtypes exist for pancreatic ductal adenocarcinoma (PDAC), but molecular characterization is not yet standard in clinical care. We implemented a biopsy protocol to perform time-sensitive whole-exome sequencing and RNA sequencing for patients with advanced PDAC. Therapeutically relevant genomic alterations were identified in 48% (34/71) and pathogenic/likely pathogenic germline alterations in 18% (13/71) of patients. Overall, 30% (21/71) of enrolled patients experienced a change in clinical management as a result of genomic data. Twenty-six patients had germline and/or somatic alterations in DNA-damage repair genes, and 5 additional patients had mutational signatures of homologous recombination deficiency but no identified causal genomic alteration. Two patients had oncogenic in-frame BRAF deletions, and we report the first clinical evidence that this alteration confers sensitivity to MAPK pathway inhibition. Moreover, we identified tumor/stroma gene expression signatures with clinical relevance. Collectively, these data demonstrate the feasibility and value of real-time genomic characterization of advanced PDAC. Significance: Molecular analyses of metastatic PDAC tumors are challenging due to the heterogeneous cellular composition of biopsy specimens and rapid progression of the disease. Using an integrated multidisciplinary biopsy program, we demonstrate that real-time genomic characterization of advanced PDAC can identify clinically relevant alterations that inform management of this difficult disease. Cancer Discov; 8(9); 1096–111. ©2018 AACR. See related commentary by Collisson, p. 1062. This article is highlighted in the In This Issue feature, p. 1047

Malignancy is accompanied by changes in the metabolism of both cells and the organism1,2. Pancreatic ductal adenocarcinoma (PDAC) is associated with wasting of peripheral tissues, a metabolic syndrome that lowers quality of life and has been proposed to decrease survival of patients with cancer3,4. Tissue wasting is a multifactorial disease and targeting specific circulating factors to reverse this syndrome has been mostly ineffective in the clinic5,6. Here we show that loss of both adipose and muscle tissue occurs early in the development of pancreatic cancer. Using mouse models of PDAC, we show that tumour growth in the pancreas but not in other sites leads to adipose tissue wasting, suggesting that tumour growth within the pancreatic environment contributes to this wasting phenotype. We find that decreased exocrine pancreatic function is a driver of adipose tissue loss and that replacement of pancreatic enzymes attenuates PDAC-associated wasting of peripheral tissues. Paradoxically, reversal of adipose tissue loss impairs survival in mice with PDAC. When analysing patients with PDAC, we find that depletion of adipose and skeletal muscle tissues at the time of diagnosis is common, but is not associated with worse survival. Taken together, these results provide an explanation for wasting of adipose tissue in early PDAC and suggest that early loss of peripheral tissue associated with pancreatic cancer may not impair survival.

Aim: We studied body composition throughout the first year after diagnosis and its association with overall survival (OS) among esophageal cancer patients. Methods: We utilized clinical abdominal computerized tomography (CT) scans from esophageal adenocarcinoma (EA) and squamous cell carcinoma (SCC) patients, who were diagnosed between 1999 and 2018 and enrolled in an ongoing study at Massachusetts General Hospital. A deep learning based automated segmentation of each CT at the L3 slice to obtain muscle (MA), subcutaneous adiposity tissue (SAT), and visceral adiposity tissue (VAT). These measures were normalized by age-, sex-, and race using a published cohort of non-cancer patient CT body composition database. We defined baseline sarcopenia using the Martin et. al. 2013 cut points. Among available scans, we obtained HRs for body composition metrics throughout the first year after diagnosis (3-6 months, 6-9 months, 9-12 months) using conditional follow-up and time-varying metrics. We obtained OS hazard ratios (HR) for pre-treatment baseline metrics using an extended cox proportional hazard model. Multivariable models included smoking status, histology, chemotherapy, radiation, and time-dependent surgery as covariates, and baseline hazards were stratified by lymph node status at diagnosis (lymph node negative, lymph node positive, and metastatic). We performed additional sensitivity baseline analyses restricting to EA patients, patients who received trimodality therapy, and stratified by lymph node status. Results: 251 patients (80.9% EA) had pre-treatment CT scans, and 438 patients (85.4% EA) had at least one CT scan throughout the first year after diagnosis. In the complete cohort, normalized baseline body composition measures and sarcopenia were not associated with OS in the univariable or multivariable analyses. Among EA only, baseline MA was associated with longer survival (HR: 0.85 per standard deviation increase; 95% CI 0.73-0.99; p=0.04). 3-6 months after diagnosis (N=206 patients), higher normalized MA, SAT, and VAT were all significantly associated (p6 months after diagnosis. The strongest association was 9-12 months after diagnosis (N=131), where a standard deviation increase in MA was associated with 0.56 hazard of death (95%CI 0.42-0.75; p Conclusions: Muscle area in the first year after diagnosis is strongly associated with survival in esophageal cancer. Since associations were strongest during and after treatment, directionality of this association is uncertain, and we will investigate additional factors to determine the clinical utility of these associations. Citation Format: Elizabeth Loehrer, Jakob Weiss, Raymond H. Mak, Junwei Lu, Andrea Shafer, Michael H. Rosenthal, Hugo J. Aerts, David C. Christiani. Body composition and overall survival in esophageal cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 761.

Background Diagnostic imaging plays a critical role in the initial diagnosis and therapeutic monitoring of pancreatic adenocarcinoma. Over the past decade, the concept of ‘borderline resectable’ pancreatic cancer has emerged to describe a distinct subset of patients existing along the spectrum from resectable to locally advanced disease for whom a microscopically margin-positive (R1) resection is considered relatively more likely, primarily due to the relationship of the primary tumor with surrounding vasculature. Materials and methods This review traces the conceptual evolution of borderline resectability from a radiological perspective, including the debates over the key imaging criteria that define the thresholds between resectable, borderline resectable, and locally advanced or metastatic disease. This review also addresses the data supporting neoadjuvant therapy in this population and discusses current imaging practices before and during treatment. Results A growing body of evidence suggests that the borderline resectable group of patients may particularly benefit from neoadjuvant therapy to increase the likelihood of an ultimately margin-negative (R0) resection. Unfortunately, anatomic and imaging criteria to define borderline resectability are not yet universally agreed upon, with several classification systems proposed in the literature and considerable variance in institution-by-institution practice. As a result of this lack of consensus, as well as overall small patient numbers and lack of established clinical trials dedicated to borderline resectable patients, accurate evidence-based diagnostic categorization and treatment selection for this subset of patients remains a significant challenge. Conclusions Clinicians and radiologists alike should be cognizant of evolving imaging criteria for borderline resectability given their profound implications for treatment strategy, follow-up recommendations, and prognosis.

The Oncology Grand Rounds series is designed to place original reports published in the Journal into clinical context. A case presentation is followed by a description of diagnostic and management challenges, a review of the relevant literature, and a summary of the authors’ suggested management approaches. The goal of this series is to help readers better understand how to apply the results of key studies, including those published in Journal of Clinical Oncology, to patients seen in their own clinical practice. A 64-year-old woman with a history of hypertension and type 2 diabetes had been in her usual state of health until she developed symptoms of diarrhea, abdominal bloating, and discomfort in the midepigastrium. Evaluation with a contrast-enhanced abdominopelvic computed tomography (CT) scan demonstrated a mass in the pancreatic body that was approximately 3.1 cm × 2 cm × 2.1 cm in size with abutment of the portal vein–superior mesenteric vein confluence for less than 180°. The confluence was narrowed but without thrombosis. No tumor–vessel interface was noted at the superior mesenteric artery, celiac artery, or common hepatic artery. Several peripancreatic lymph nodes were observed that measured up to 11 mm × 5 mm. No evidence for distant spread of disease was identified. An upper endoscopy with endoscopic ultrasound was performed and fine-needle aspirates of the pancreas mass were positive for malignant cells that were consistent with adenocarcinoma. Chest CT scan without intravenous contrast demonstrated no evidence of metastatic disease. The patient came to the clinic to discuss management of her newly diagnosed malignancy.

Standardized terminology is critical to providing consistent reports to referring clinicians. This lexicon aims to provide a reference for terminology frequently used in rectal cancer and reflects the consensus of the Society of Abdominal Radiology Disease Focused Panel in Rectal cancer. This lexicon divided the terms into the following categories: primary tumor staging, nodal staging, treatment response, anal canal anatomy, general anatomy, and treatments.

Immunotherapy with checkpoint inhibitors, such as the programmed death-1 (PD-1) antibodies pembrolizumab and nivolumab, are effective in a variety of tumors, yet not all patients respond. Tumor microsatellite instability-high (MSI-H) has emerged as a biomarker of response to checkpoint blockade, leading to the tissue agnostic approval of pembrolizumab in MSI-H cancers. Here we describe a patient with MSI-H colorectal cancer that was treated with this immune checkpoint inhibitor and exhibited progression of disease. We examined this intrinsic resistance through genomic, transcriptional, and pathologic characterization of the patient's tumor and the associated immune microenvironment. The tumor had typical MSI-H molecular features, including a high neoantigen load. We also identified biallelic loss of the gene for b2-microglobulin (B2M), whose product is critical for antigen presentation. Immune infiltration deconvolution analysis of bulk transcriptome data from this anti-PD-1–resistant tumor and hundreds of other colorectal cancer specimens revealed a high natural killer cell and M2 macrophage infiltration in the patient's cancer. This was confirmed by single-cell transcriptome analysis and multiplex immuno-fluorescence. Our study provides insight into resistance in MSI-H tumors and suggests immunotherapeutic strategies in additional genomic contexts of colorectal cancer., Cancer Research UK (C10674/A27140 Grand Challenge Award), Entertainment Industry Foundation. Stand Up to Cancer Colorectal Cancer Dream Team (Grant Number: SU2C-AACR-DT22-17)

Background: Pancreatic cancer is associated with development of cachexia, a wasting syndrome thought to limit survival. Few studies have longitudinally quantified peripheral tissues or identified biomarkers predictive of future tissue wasting. Methods: Adipose and muscle tissue were measured by computed tomography (CT) at diagnosis and 50 to 120 days later in 164 patients with advanced pancreatic cancer. Tissue changes and survival were evaluated by Cox proportional hazards regression. Baseline levels of circulating markers were examined in relation to future tissue wasting. Results: Compared with patients in the bottom quartile of muscle change per 30 days (average gain of 0.8 ± 2.0 cm2), those in the top quartile (average loss of 12.9 ± 4.9 cm2) had a hazard ratio (HR) for death of 2.01 [95% confidence interval (CI), 1.12–3.62]. Patients in the top quartile of muscle attenuation change (average decrease of 4.9 ± 2.4 Hounsfield units) had an HR of 2.19 (95% CI, 1.18–4.04) compared with those in the bottom quartile (average increase of 2.4 ± 1.6 Hounsfield units). Changes in adipose tissue were not associated with survival. Higher plasma branched chain amino acids (BCAA; P = 0.004) and lower monocyte chemoattractant protein-1 (MCP-1; P = 0.005) at diagnosis were associated with greater future muscle loss. Conclusions: In patients with advanced pancreatic cancer, muscle loss and decrease in muscle density in 2 to 4 months after diagnosis were associated with reduced survival. BCAAs and MCP-1 levels at diagnosis were associated with subsequent muscle loss. Impact: BCAAs and MCP-1 levels at diagnosis could identify a high-risk group for future tissue wasting.

Despite advances in molecular biology, imaging, and treatment, gastric neoplasms remain a significant cause of morbidity and mortality; gastric adenocarcinoma is the fifth most common malignancy and third most common cause of death worldwide (Brenner et al., Methods Mol Biol 472:467-477, 2009; Howson et al. Epidemiol Rev 8:1-27, 1986; Roder, Gastric Cancer 5(Suppl 1):5-11, 2002; Ferlay et al., GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. International Agency for Research on Cancer, 2013). Because of both the frequency at which malignant gastric tumors occur as well as the worldwide impact, gastric neoplasms remain important lesions to identify and characterize on all imaging modalities. Despite the varied histologies and behaviors of these neoplasms, many have similar imaging features. Nonetheless, the treatment, management, and prognosis of gastric neoplasms vary by pathology, so it is essential for the radiologist to make every effort to differentiate between these lesions and raise the less common entities as differential diagnostic considerations when appropriate.