Automated pancreatic segmentation and fat fraction evaluation based on a self-supervised transfer learning network

Accurate segmentation of the pancreas from abdominal computed tomography (CT) images is challenging but essential for the diagnosis and treatment of pancreatic disorders such as tumours and diabetes. In this study, a dataset with 229 sets of high-resolution CT images was generated and annotated. We proposed a novel 3D segmentation model named nnTransfer (nonisomorphic transfer learning) net, which employs generative model structure for self-supervision to facilitate the network's learning of image attributes from unlabelled data. The effectiveness for pancreas segmentation of nnTransfer was assessed using the Hausdorff distance (HD) and Dice similarity coefficient (DSC) on the dataset. Additionally, a histogram analysis with local thresholding was used to achieve automated whole-volume measurement of pancreatic fat (fat volume fraction, FVF). The proposed technique performed admirably on the dataset, with DSC: 0.937 ± 0.019 and HD: 2.655 ± 1.479. The mean pancreas volume and FVF of the pancreas were 91.95 ± 23.90 cm3 and 12.67 % ± 9.84 %, respectively. The nnTransfer functioned flawlessly and autonomously, facilitating the use of the FVF to evaluate pancreatic disease, particularly in patients with diabetes.

Volumetric Pancreas Segmentation on Computed Tomography: Accuracy and Efficiency of a Convolutional Neural Network Versus Manual Segmentation in 3D Slicer in the Context of Interreader Variability of Expert Radiologists

PURPOSE

This study aimed to compare accuracy and efficiency of a convolutional neural network (CNN)-enhanced workflow for pancreas segmentation versus radiologists in the context of interreader reliability.

METHODS

Volumetric pancreas segmentations on a data set of 294 portal venous computed tomographies were performed by 3 radiologists (R1, R2, and R3) and by a CNN. Convolutional neural network segmentations were reviewed and, if needed, corrected ("corrected CNN [c-CNN]" segmentations) by radiologists. Ground truth was obtained from radiologists' manual segmentations using simultaneous truth and performance level estimation algorithm. Interreader reliability and model's accuracy were evaluated with Dice-Sorenson coefficient (DSC) and Jaccard coefficient (JC). Equivalence was determined using a two 1-sided test. Convolutional neural network segmentations below the 25th percentile DSC were reviewed to evaluate segmentation errors. Time for manual segmentation and c-CNN was compared.

RESULTS

Pancreas volumes from 3 sets of segmentations (manual, CNN, and c-CNN) were noninferior to simultaneous truth and performance level estimation-derived volumes [76.6 cm 3 (20.2 cm 3 ), P < 0.05]. Interreader reliability was high (mean [SD] DSC between R2-R1, 0.87 [0.04]; R3-R1, 0.90 [0.05]; R2-R3, 0.87 [0.04]). Convolutional neural network segmentations were highly accurate (DSC, 0.88 [0.05]; JC, 0.79 [0.07]) and required minimal-to-no corrections (c-CNN: DSC, 0.89 [0.04]; JC, 0.81 [0.06]; equivalence, P < 0.05). Undersegmentation (n = 47 [64%]) was common in the 73 CNN segmentations below 25th percentile DSC, but there were no major errors. Total inference time (minutes) for CNN was 1.2 (0.3). Average time (minutes) taken by radiologists for c-CNN (0.6 [0.97]) was substantially lower compared with manual segmentation (3.37 [1.47]; savings of 77.9%-87% [ P < 0.0001]).

CONCLUSIONS

Convolutional neural network-enhanced workflow provides high accuracy and efficiency for volumetric pancreas segmentation on computed tomography.

Pancreatic atrophy and recovery after allogeneic hematopoietic cell transplantation

BACKGROUND

Pancreatic atrophy after allogeneic hematopoietic cell transplantation (HCT) is one of the symptoms associated with chronic graft-versus-host disease (GVHD). Although pancreatic atrophy has been considered to cause exocrine insufficiency and weight loss, it is not yet clear what kinds of recipients can be expected to recover their body weight (BW) or pancreatic thickness. In addition, the effect of pancreatic atrophy on the prognosis has not been clarified.

METHODS

We retrospectively analyzed 170 recipients who received allogeneic bone marrow transplantation or peripheral blood stem cell transplantation, and evaluated them using the CT scan images obtained closest to 1, 2, 3, and 4 years after HCT.

RESULTS

Fifty-five recipients (32.4%) demonstrated pancreatic atrophy, and 11 (20%) of them recovered their pancreatic thickness. While recipients without pancreatic atrophy gradually recovered their BW (P < 0.001), those with atrophy did not (P = 0.12). Moderate and severe chronic GVHD tended to be slightly more common in the atrophy group (47.3% vs 38.3%), whereas the pancreatic thickness tended to recover in these recipients (30.8% vs 10.3%). HCT from a female donor to a male recipient showed superior pancreatic recovery compared to other donor and recipient sex combinations. Pancreatic atrophy treated as a significantly associated with inferior survival (HR 4.91, P < 0.001) and an increased risk of non-relapse mortality (HR 8.75, P < 0.001).

CONCLUSIONS

These results suggest that it is important to monitor pancreatic thickness after HCT. Further prospective investigations are warranted to clarify the significance of pancreatic atrophy on clinical outcomes.

Current update on molecular cytogenetics, diagnosis and management of gastrointestinal stromal tumors

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal (GI) tract and are thought to arise from precursors of the interstitial cells of Cajal. GISTs can arise anywhere in the GI tract, but most commonly originate from the stomach and small intestine. The majority of GISTs occur as a result of activating mutations in two receptor protein tyrosine kinases: KIT and/or platelet-derived growth factor receptor-α. Mutational analyses allow for predicting patient prognosis and treatment response. Clinical presentations can vary from no symptoms, typical in the case of small incidentally found tumors, to GI bleeding, abdominal discomfort, and ulcer-related symptoms when the tumor is enlarged. Imaging plays a critical role in the diagnosis and management of these tumors with multiphasic computed tomography serving as the imaging modality of choice. Magnetic resonance imaging and positron emission tomography-computed tomography can serve as imaging adjuncts in lesion characterization, especially with liver metastases, and subsequent staging and assessment for treatment response or recurrence. Surgical resection is the preferred management for small GISTs, while tyrosine kinase inhibitors − imatinib mesylate and sunitinib malate − serve as crucial molecular-targeted therapies for locally advanced and metastatic GISTs. This review article highlights the clinical presentation, pathology and molecular cytogenetics, imaging features, and current management of GISTs.

Two-stage deep learning model for fully automated pancreas segmentation on computed tomography: Comparison with intra-reader and inter-reader reliability at full and reduced radiation dose on an external dataset

PURPOSE

To develop a two-stage three-dimensional (3D) convolutional neural networks (CNNs) for fully automated volumetric segmentation of pancreas on computed tomography (CT) and to further evaluate its performance in the context of intra-reader and inter-reader reliability at full dose and reduced radiation dose CTs on a public dataset.

METHODS

A dataset of 1994 abdomen CT scans (portal venous phase, slice thickness ≤ 3.75-mm, multiple CT vendors) was curated by two radiologists (R1 and R2) to exclude cases with pancreatic pathology, suboptimal image quality, and image artifacts (n = 77). Remaining 1917 CTs were equally allocated between R1 and R2 for volumetric pancreas segmentation [ground truth (GT)]. This internal dataset was randomly divided into training (n = 1380), validation (n = 248), and test (n = 289) sets for the development of a two-stage 3D CNN model based on a modified U-net architecture for automated volumetric pancreas segmentation. Model's performance for pancreas segmentation and the differences in model-predicted pancreatic volumes vs GT volumes were compared on the test set. Subsequently, an external dataset from The Cancer Imaging Archive (TCIA) that had CT scans acquired at standard radiation dose and same scans reconstructed at a simulated 25% radiation dose was curated (n = 41). Volumetric pancreas segmentation was done on this TCIA dataset by R1 and R2 independently on the full dose and then at the reduced radiation dose CT images. Intra-reader and inter-reader reliability, model's segmentation performance, and reliability between model-predicted pancreatic volumes at full vs reduced dose were measured. Finally, model's performance was tested on the benchmarking National Institute of Health (NIH)-Pancreas CT (PCT) dataset.

RESULTS

Three-dimensional CNN had mean (SD) Dice similarity coefficient (DSC): 0.91 (0.03) and average Hausdorff distance of 0.15 (0.09) mm on the test set. Model's performance was equivalent between males and females (P = 0.08) and across different CT slice thicknesses (P > 0.05) based on noninferiority statistical testing. There was no difference in model-predicted and GT pancreatic volumes [mean predicted volume 99 cc (31cc); GT volume 101 cc (33 cc), P = 0.33]. Mean pancreatic volume difference was -2.7 cc (percent difference: -2.4% of GT volume) with excellent correlation between model-predicted and GT volumes [concordance correlation coefficient (CCC)=0.97]. In the external TCIA dataset, the model had higher reliability than R1 and R2 on full vs reduced dose CT scans [model mean (SD) DSC: 0.96 (0.02), CCC = 0.995 vs R1 DSC: 0.83 (0.07), CCC = 0.89, and R2 DSC:0.87 (0.04), CCC = 0.97]. The DSC and volume concordance correlations for R1 vs R2 (inter-reader reliability) were 0.85 (0.07), CCC = 0.90 at full dose and 0.83 (0.07), CCC = 0.96 at reduced dose datasets. There was good reliability between model and R1 at both full and reduced dose CT [full dose: DSC: 0.81 (0.07), CCC = 0.83 and reduced dose DSC:0.81 (0.08), CCC = 0.87]. Likewise, there was good reliability between model and R2 at both full and reduced dose CT [full dose: DSC: 0.84 (0.05), CCC = 0.89 and reduced dose DSC:0.83(0.06), CCC = 0.89]. There was no difference in model-predicted and GT pancreatic volume in TCIA dataset (mean predicted volume 96 cc (33); GT pancreatic volume 89 cc (30), p = 0.31). Model had mean (SD) DSC: 0.89 (0.04) (minimum-maximum DSC: 0.79 -0.96) on the NIH-PCT dataset.

CONCLUSION

A 3D CNN developed on the largest dataset of CTs is accurate for fully automated volumetric pancreas segmentation and is generalizable across a wide range of CT slice thicknesses, radiation dose, and patient gender. This 3D CNN offers a scalable tool to leverage biomarkers from pancreas morphometrics and radiomics for pancreatic diseases including for early pancreatic cancer detection.

Imatinib-induced pancreatic hypertrophy in patients with gastrointestinal stromal tumor: Association with overall survival

OBJECTIVES

To investigate the frequency of imatinib-induced pancreatic complications and determine whether these are survival prognostic factors in patients with gastrointestinal stromal tumor (GIST).

METHODS

This retrospective multicenter study included patients with histopathologically diagnosed GIST treated with imatinib who underwent computed tomography (CT) within 100 days before (pretreatment CT) and 500 days after (post-treatment CT) imatinib initiation (January 2004-December 2019). Forty-eight patients (63.0 ± 12.1 years, 30 men) were included. Two blinded radiologists independently measured pancreatic volumes. Pancreatic volume on pretreatment CT was compared with that of the control (within 1 year prior to pretreatment CT) and the first two post-treatment CTs using paired t-tests. Thresholds for pancreatic hypertrophy and atrophy were defined using a log-rank test. The prognostic importance of pancreatic hypertrophy was further analyzed using multivariate Cox proportional hazard regression models.

RESULTS

Pancreatic volume was significantly higher for the first post-treatment CT than pretreatment CT (71.5 cm³ vs. 67.4 cm³, P = .027), whereas no significant difference was observed between the pretreatment and control CTs. Optimal thresholds for pancreatic hypertrophy and atrophy were defined as an 22% increase and 30% decrease and found in 20 and three patients, respectively. Pancreatic hypertrophy was significantly associated with reduced survival [hazard ratio = 2.9 (95% confidence interval, 1.3-6.5), P = .0088]. No patients showed serum lipase elevation, nor were they suspected of having acute pancreatitis.

CONCLUSION

There was frequent asymptomatic pancreatic swelling in patients with GIST after imatinib treatment, and a ≥22% increase in pancreatic volume was a predictor of reduced survival.

Development of a volumetric pancreas segmentation CT dataset for AI applications through trained technologists: a study during the COVID 19 containment phase

PURPOSE

To evaluate the performance of trained technologists vis-à-vis radiologists for volumetric pancreas segmentation and to assess the impact of supplementary training on their performance.

METHODS

In this IRB-approved study, 22 technologists were trained in pancreas segmentation on portal venous phase CT through radiologist-led interactive videoconferencing sessions based on an image-rich curriculum. Technologists segmented pancreas in 188 CTs using freehand tools on custom image-viewing software. Subsequent supplementary training included multimedia videos focused on common errors, which were followed by second batch of 159 segmentations. Two radiologists reviewed all cases and corrected inaccurate segmentations. Technologists' segmentations were compared against radiologists' segmentations using Dice-Sorenson coefficient (DSC), Jaccard coefficient (JC), and Bland-Altman analysis.

RESULTS

Corrections were made in 71 (38%) cases from first batch [26 (37%) oversegmentations and 45 (63%) undersegmentations] and in 77 (48%) cases from second batch [12 (16%) oversegmentations and 65 (84%) undersegmentations]. DSC, JC, false positive (FP), and false negative (FN) [mean (SD)] in first versus second batches were 0.63 (0.15) versus 0.63 (0.16), 0.48 (0.15) versus 0.48 (0.15), 0.29 (0.21) versus 0.21 (0.10), and 0.36 (0.20) versus 0.43 (0.19), respectively. Differences were not significant (p > 0.05). However, range of mean pancreatic volume difference reduced in the second batch [- 2.74 cc (min - 92.96 cc, max 87.47 cc) versus - 23.57 cc (min - 77.32, max 30.19)].

CONCLUSION

Trained technologists could perform volumetric pancreas segmentation with reasonable accuracy despite its complexity. Supplementary training further reduced range of volume difference in segmentations. Investment into training technologists could augment and accelerate development of body imaging datasets for AI applications.

A Review of the Mechanisms and Clinical Implications of Precision Cancer Therapy-Related Toxicity: A Primer for the Radiologist

OBJECTIVE. The purpose of this review is to elucidate the mechanisms, types, and clinical significance of molecular targeted therapy (MTT) and immune checkpoint inhibitors (ICIs) and their related toxicity, emphasizing the radiologic manifestations. CONCLUSION. The related toxicities of MTT and ICIs can have acute, recurrent, chronic, and delayed presentations. These toxicities may serve as markers of response and survival. By understanding the clinical significance of drug toxicities, radiologists can play an important role in personalized cancer therapy.

Investigation of morphological and functional changes in the liver and pancreas during bevacizumab treatment

Objectives: Recently, there have been reports regarding the atrophy of various organs caused by molecular targeted drugs. We investigated morphological and clinical changes in the liver and pancreas caused by treatment with bevacizumab.Methods: We investigated 30 patients with colorectal cancer who received bevacizumab-containing chemotherapy (study group) and 11 patients with colorectal cancer who received chemotherapy without bevacizumab (control group) from 2010 to 2014. We obtained computed tomography data of the liver and pancreas and performed three-dimensional image analysis and volumetry. Laboratory data before and after chemotherapy were analyzed.Results: There was no significant difference in liver volume before and after bevacizumab-containing chemotherapy, but the pancreatic volume was found to be significantly reduced after bevacizumab-containing chemotherapy (57.9 ± 16 mL versus 47.4 ± 15.3 mL; p = .005). The liver and pancreatic volume did not change statistically in the control group. With regard to complete blood cell counts and laboratory data, no significant differences were observed in the leukocyte count and hemoglobin, hemoglobin A1c, triglyceride, albumin, and C-reactive protein levels. In contrast, there was a significant decrease in the platelet count, total cholesterol level and a significant increase in the amylase level. A chemotherapy regimen that included bevacizumab reduced pancreatic volume and significantly altered the morphology of the pancreas.Conclusions: Although bevacizumab caused atrophy of the pancreas and reduced pancreatic volume, pancreatic endocrine function showed no change. Future studies should investigate the survival rate and functional changes caused by bevacizumab treatment.

Radiological and Pathological Assessment of the 2017 Revised International Association of Pancreatology Consensus Guidelines for Intraductal Papillary Mucinous Neoplasm, With an Emphasis on the Gastric Pyloric Gland Type

OBJECTIVES

This study aimed to assess the pitfalls of the current International Association of Pancreatology guidelines (IAPCG2017) for pancreatic intraductal papillary mucinous neoplasm (IPMN) and identify the criteria for future guidelines.

METHODS

Eighty surgically resected, consecutive IPMN cases were analyzed. Data including tumor site, IPMN duct type, and surgery type were collected. Based on radiological data, cases were retrospectively classified as high-risk stigmata (HRS) and non-HRS. Pathological grades and histological subtypes of IPMN cases were determined. Severe stromal sclerosis of the IPMN septa/marked parenchymal atrophy in the upstream pancreas was investigated pathologically. Positive/negative predictive values of the IAPCG2017 were calculated. Clinicopathological features of HRS-benign cases (pathologically benign IPMN cases meeting the HRS criteria) were extracted.

RESULTS

The positive/negative predictive values were 72.7%/64.0%, 70.0%/34.6%, and 54.0%/63.3% for IAPCG2017, HRS-main pancreatic duct, and HRS-nodule criteria, respectively. The 15 HRS-benign cases (18.8%) included 13 pancreatoduodenectomies and 10 cases of gastric pyloric (GP) gland subtype. Severe upstream atrophy was significantly related to IPMN malignancy, unlike the severe sclerosis of IPMN septa.

CONCLUSIONS

Benign IPMNs of GP subtype are sometimes categorized as HRS with the IAPCG2017. Collecting data on the natural course of GP-IPMN is necessary. To evaluate upstream atrophy may be of value to predict IPMN malignancy.

Imaging of Metastatic Germ Cell Tumors in Male Patients From Initial Diagnosis to Treatment-Related Toxicities: A Primer for Radiologists

OBJECTIVE. This review describes the influence of histology and metastatic sites on prognosis in male patients with metastatic germ cell tumors (GCTs) and explains the role imaging in assessing therapeutic response, residual disease, recurrence, sand treatment-related toxicities. CONCLUSION. Seminomatous and nonseminomatous GCTs differ in imaging appearance, pattern of spread, and prognosis, and an organ-based approach is helpful in prognostication. Multimodality imaging aids in accurate staging, prognostication, characterization of treatment response, and identification of therapy-related toxicity.

Frequency and imaging features of abdominal immune-related adverse events in metastatic lung cancer patients treated with PD-1 inhibitor

PURPOSE

To investigate the frequency and imaging features of radiographically evident abdominal immune-related adverse events (irAEs) in patients with metastatic non-small-cell lung cancer (NSCLC) treated with PD-1 inhibitors.

METHODS

This retrospective study included 137 patients with metastatic NSCLC treated with PD-1 inhibitor nivolumab monotherapy (75 women; median age: 65 years), who had a baseline CT and at least one follow-up abdomen CT during therapy. Baseline and all follow-up abdominal CTs performed for monitoring of nivolumab therapy were reviewed to identify the organ-specific abdominal irAEs including colitis/enteritis, hepatitis, biliary toxicity, pancreatitis, nephritis, sarcoid-like reaction, and pancreatic and adrenal atrophy. Their frequency and imaging features were described.

RESULTS

Eighteen (13%) patients had radiologically identified abdominal irAEs (median 2.1 months after starting nivolumab; interquartile range 1.17-5.83 months); 16 patients developed enteritis/colitis (12 pancolitis, two segmental colitis, one enterocolitis, one enteritis), two hepatitis, one adrenalitis. One patient with hepatitis also developed colitis/enteritis. Radiographic abdominal irAE occurred after nivolumab therapy was discontinued in six patients before any subsequent therapy was started. IrAEs prompted nivolumab interruption and treatment with steroids in four patients (three colitis/enteritis, one hepatitis). Most common CT features of colitis/enteritis included mesenteric hyperemia (n = 15), bowel wall thickening (n = 13), mucosal hyperenhancement (n = 10), and fluid-filled colon (n = 9).

CONCLUSION

Abdominal irAEs were detected on CT in 13% of NSCLC patients treated with nivolumab, and colitis, in the pancolitis form, was the most common irAE. Given the expanding role of immunotherapy, radiologists should be aware of the frequency and imaging manifestations of abdominal irAEs and the impact on patient management.

Sorafenib-induced Prostate Volume Reduction, a New Adverse Effect Detected by Imaging: A Pilot Study

Background

Sorafenib has been used in the treatment of advanced hepatocellular carcinoma (HCC) and renal cell carcinoma (RCC). Sorafenib-associated organ reduction have been reported on imaging, such as thyroid, pancreas and muscle, but there has been no research on prostate volume reduction (PVR).

Methods

We retrospectively analyzed 26 patients (twenty with HCC and six patients with RCC) who underwent sorafenib therapy for 31 to 1225 days (median, 100 days). PVR was estimated by two independent readers using CT volumetry.

Results

The sum of all prostate volumes measured by reader 1 was 24.2 ± 13.8 cm³ on the baseline CT and 20.4 ± 10.6 cm³ on the follow-up CT (p < 0.001), and that measured by reader 2 was 22.3 ± 13.9 cm³ on the baseline CT and 19.2 ± 10.6 cm³ on the follow-up CT (p < 0.001). The concordance correlation coefficient for the prostate volume measured by the two readers was 0.95 on the baseline CT scans and 0.94 on the follow-up CT scans. Sorafenib-associated PVR demonstrated slight dependence to the exposure time (r = -0.23). One patient with benign prostatic hyperplasia (BPH) showed PVR (from 80.4 to 61.5 cm³ [reader 1]; 83.4 to 61.6 cm³ [reader 2]) after sorafenib administration. Sorafenib-associated PVR occurred in patients both with and without underlying liver dysfunction with relative prostate volume changes of 86.7 ± 12.0% and 85.0 ± 9.0%, respectively.

Conclusion

Our study demonstrated significant PVR with sorafenib treatment in patients regardless of the presence of BPH and underlying liver dysfunction.

Unintended Consequences of Systemic and Ablative Oncologic Therapy in the Abdomen and Pelvis

Human cancers are genetically complex and diverse. Although advances in oncologic therapy aim to define and target unique steps in carcinogenesis, oncologists often rely on less discriminate anticancer therapies that have consequences for normal tissues. Even many of the so-called targeted therapies currently employed can adversely affect normal cells, leading to complications that necessitate dose reductions or cessation of specific therapies. This article explores the unintended consequences of currently employed systemic and ablative anticancer therapies that might manifest at imaging examinations of the abdomen and pelvis, including cytotoxic, molecular targeted, and immunologic agents; ablation; and hematopoietic stem cell transplant. Each of these treatments can have both major and minor unintended effects in the targeted organ(s), in local or adjacent structures, or at distant sites. Timely detection and reporting of adverse consequences of anticancer therapies by the astute imager can result in critical treatment modifications and/or lifesaving interventions; therefore, knowledge of these unintended effects is paramount for radiologists interpreting the results of imaging examinations in cancer patients. ^©RSNA, 2018.

Unravelling the pharmacologic opportunities and future directions for targeted therapies in gastro-intestinal cancers part 2: Neuroendocrine tumours, hepatocellular carcinoma, and gastro-intestinal stromal tumours

Until the 1990s, cytotoxic chemotherapy has been the cornerstone of medical therapy for gastrointestinal (GI) cancers. Better understanding of the cancer cell molecular biology has led to the therapeutic revolution of targeted therapies, i.e. monoclonal antibodies or small molecule inhibitors directed against proteins that are specifically overexpressed or mutated in cancer cells. These agents, being more specific to cancer cells, were expected to be less toxic than conventional cytotoxic agents. However, their effects have sometimes been disappointing, due to intrinsic or acquired resistance mechanisms, or to an activity restricted to some tumour settings, illustrating the importance of patient selection and early identification of predictive biomarkers of response to these therapies. Targeted agents have provided clinical benefit in many GI cancer types. Particularly, some GI tumours are considered chemoresistant and targeted therapies have offered a new therapeutic base for their management. Hence, somatostatin receptor-directed strategies, sorafenib, and imatinib have revolutioned the management of neuroendocrine tumours (NET), hepatocellular carcinoma (HCC), and gastrointestinal stromal tumours (GIST), respectively, and are now used as first-line treatment in many patients affected by these tumours. However, these agents face problems of resistances and identification of predictive biomarkers from imaging and/or biology. We propose a comprehensive two-part review providing a panoramic approach of the successes and failures of targeted agents in GI cancers to unravel the pharmacologic opportunities and future directions for these agents in GI oncology. In this second part, we will focus on NET, HCC, and GIST, whose treatment relies primarily on targeted therapies.

Biliary and pancreatic complications of molecular targeted therapies in cancer imaging

The purpose of this review is to familiarize radiologists with the different imaging manifestations of biliary and pancreatic toxicity of molecular targeted therapies. The advent of molecular targeted therapies for cancer treatment has prompted radiologists to be familiar with these new molecules, their patterns of response, and their class-specific toxicities. While liver and bowel toxicities have been extensively reported in literature, less is known about the pathogenesis and imaging of toxicity involving the pancreatobiliary system. Biliary and pancreatic toxicity of molecular targeted therapies present with variable manifestations and varying degrees of severity, from asymptomatic liver function tests elevation to acute pancreatitis or cholecystitis. Management of these conditions depends on the clinical scenario and the severity of the findings. In this article, we will (1) present the various classes of molecular targeted therapies most commonly associated with biliary and pancreatic toxicity; (2) illustrate imaging findings of drug-associated biliary and pancreatic injuries and their possible differential diagnosis; and (3) provide a guide for management of these conditions.