Pre-operative radiomics model for prognostication in resectable pancreatic adenocarcinoma with external validation

Radiomics Beyond the Hype: A Critical Evaluation Toward Oncologic Clinical Use

Radiomics is a promising and fast-developing field within oncology that involves the mining of quantitative high-dimensional data from medical images. Radiomics has the potential to transform cancer management, whereby radiomics data can be used to aid early tumor characterization, prognosis, risk stratification, treatment planning, treatment response assessment, and surveillance. Nevertheless, certain challenges have delayed the clinical adoption and acceptability of radiomics in routine clinical practice. The objectives of this report are to (a) provide a perspective on the translational potential and potential impact of radiomics in oncology; (b) explore frequent challenges and mistakes in its derivation, encompassing study design, technical requirements, standardization, model reproducibility, transparency, data sharing, privacy concerns, quality control, as well as the complexity of multistep processes resulting in less radiologist-friendly interfaces; (c) discuss strategies to overcome these challenges and mistakes; and (d) propose measures to increase the clinical use and acceptability of radiomics, taking into account the different perspectives of patients, health care workers, and health care systems. Keywords: Radiomics, Oncology, Cancer Management, Artificial Intelligence © RSNA, 2024.

A clinical-radiomics nomogram based on dual-layer spectral detector CT to predict cancer stage in pancreatic ductal adenocarcinoma

BACKGROUND

This study aimed to evaluate the efficacy of radiomics signatures derived from polyenergetic images (PEIs) and virtual monoenergetic images (VMIs) obtained through dual-layer spectral detector CT (DLCT). Moreover, it sought to develop a clinical-radiomics nomogram based on DLCT for predicting cancer stage (early stage: stage I-II, advanced stage: stage III-IV) in pancreatic ductal adenocarcinoma (PDAC).

METHODS

A total of 173 patients histopathologically diagnosed with PDAC and who underwent contrast-enhanced DLCT were enrolled in this study. Among them, 49 were in the early stage, and 124 were in the advanced stage. Patients were randomly categorized into training (n = 122) and test (n = 51) cohorts at a 7:3 ratio. Radiomics features were extracted from PEIs and 40-keV VMIs were reconstructed at both arterial and portal venous phases. Radiomics signatures were constructed based on both PEIs and 40-keV VMIs. A radiomics nomogram was developed by integrating the 40-keV VMI-based radiomics signature with selected clinical predictors. The performance of the nomogram was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curves analysis (DCA).

RESULTS

The PEI-based radiomics signature demonstrated satisfactory diagnostic efficacy, with the areas under the ROC curves (AUCs) of 0.92 in both the training and test cohorts. The optimal radiomics signature was based on 40-keV VMIs, with AUCs of 0.96 and 0.94 in the training and test cohorts. The nomogram, which integrated a 40-keV VMI-based radiomics signature with two clinical parameters (tumour diameter and normalized iodine density at the portal venous phase), demonstrated promising calibration and discrimination in both the training and test cohorts (0.97 and 0.91, respectively). DCA indicated that the clinical-radiomics nomogram provided the most significant clinical benefit.

CONCLUSIONS

The radiomics signature derived from 40-keV VMI and the clinical-radiomics nomogram based on DLCT both exhibited exceptional performance in distinguishing early from advanced stages in PDAC, aiding clinical decision-making for patients with this condition.

Delta Radiomic Features Predict Resection Margin Status and Overall Survival in Neoadjuvant-Treated Pancreatic Cancer Patients

BACKGROUND

Neoadjuvant therapy (NAT) emerged as the standard of care for patients with pancreatic ductal adenocarcinoma (PDAC) who undergo surgery; however, surgery is morbid, and tools to predict resection margin status (RMS) and prognosis in the preoperative setting are needed. Radiomic models, specifically delta radiomic features (DRFs), may provide insight into treatment dynamics to improve preoperative predictions.

METHODS

We retrospectively collected clinical, pathological, and surgical data (patients with resectable, borderline, locally advanced, and metastatic disease), and pre/post-NAT contrast-enhanced computed tomography (CT) scans from PDAC patients at the University of Texas Southwestern Medical Center (UTSW; discovery) and Humanitas Hospital (validation cohort). Gross tumor volume was contoured from CT scans, and 257 radiomics features were extracted. DRFs were calculated by direct subtraction of pre/post-NAT radiomic features. Cox proportional models and binary prediction models, including/excluding clinical variables, were constructed to predict overall survival (OS), disease-free survival (DFS), and RMS.

RESULTS

The discovery and validation cohorts comprised 58 and 31 patients, respectively. Both cohorts had similar clinical characteristics, apart from differences in NAT (FOLFIRINOX vs. gemcitabine/nab-paclitaxel; p < 0.05) and type of surgery resections (pancreatoduodenectomy, distal or total pancreatectomy; p < 0.05). The model that combined clinical variables (pre-NAT carbohydrate antigen (CA) 19-9, the change in CA19-9 after NAT (∆CA19-9), and resectability status) and DRFs outperformed the clinical feature-based models and other radiomics feature-based models in predicting OS (UTSW: 0.73; Humanitas: 0.66), DFS (UTSW: 0.75; Humanitas: 0.64), and RMS (UTSW 0.73; Humanitas: 0.69).

CONCLUSIONS

Our externally validated, predictive/prognostic delta-radiomics models, which incorporate clinical variables, show promise in predicting the risk of predicting RMS in NAT-treated PDAC patients and their OS or DFS.

Development of Clinical Radiomics-Based Models to Predict Survival Outcome in Pancreatic Ductal Adenocarcinoma: A Multicenter Retrospective Study

PURPOSE

This multicenter retrospective study aims to identify reliable clinical and radiomic features to build machine learning models that predict progression-free survival (PFS) and overall survival (OS) in pancreatic ductal adenocarcinoma (PDAC) patients.

METHODS

Between 2010 and 2020 pre-treatment contrast-enhanced CT scans of 287 pathology-confirmed PDAC patients from two sites of the Hopital Universitaire de Bruxelles (HUB) and from 47 hospitals within the HUB network were retrospectively analysed. Demographic, clinical, and survival data were also collected. Gross tumour volume (GTV) and non-tumoral pancreas (RPV) were semi-manually segmented and radiomics features were extracted. Patients from two HUB sites comprised the training dataset, while those from the remaining 47 hospitals of the HUB network constituted the testing dataset. A three-step method was used for feature selection. Based on the GradientBoostingSurvivalAnalysis classifier, different machine learning models were trained and tested to predict OS and PFS. Model performances were assessed using the C-index and Kaplan-Meier curves. SHAP analysis was applied to allow for post hoc interpretability.

RESULTS

A total of 107 radiomics features were extracted from each of the GTV and RPV. Fourteen subgroups of features were selected: clinical, GTV, RPV, clinical & GTV, clinical & GTV & RPV, GTV-volume and RPV-volume both for OS and PFS. Subsequently, 14 Gradient Boosting Survival Analysis models were trained and tested. In the testing dataset, the clinical & GTV model demonstrated the highest performance for OS (C-index: 0.72) among all other models, while for PFS, the clinical model exhibited a superior performance (C-index: 0.70).

CONCLUSIONS

An integrated approach, combining clinical and radiomics features, excels in predicting OS, whereas clinical features demonstrate strong performance in PFS prediction.

An Integrated Clinical and Computerized Tomography-Based Radiomic Feature Model to Separate Benign from Malignant Pleural Effusion

INTRODUCTION

Distinguishing between malignant pleural effusion (MPE) and benign pleural effusion (BPE) poses a challenge in clinical practice. We aimed to construct and validate a combined model integrating radiomic features and clinical factors using computerized tomography (CT) images to differentiate between MPE and BPE.

METHODS

A retrospective inclusion of 315 patients with pleural effusion (PE) was conducted in this study (training cohort: n = 220; test cohort: n = 95). Radiomic features were extracted from CT images, and the dimensionality reduction and selection processes were carried out to obtain the optimal radiomic features. Logistic regression (LR), support vector machine (SVM), and random forest were employed to construct radiomic models. LR analyses were utilized to identify independent clinical risk factors to develop a clinical model. The combined model was created by integrating the optimal radiomic features with the independent clinical predictive factors. The discriminative ability of each model was assessed by receiver operating characteristic curves, calibration curves, and decision curve analysis (DCA).

RESULTS

Out of the total 1,834 radiomic features extracted, 15 optimal radiomic features explicitly related to MPE were picked to develop the radiomic model. Among the radiomic models, the SVM model demonstrated the highest predictive performance [area under the curve (AUC), training cohort: 0.876, test cohort: 0.774]. Six clinically independent predictive factors, including age, effusion laterality, procalcitonin, carcinoembryonic antigen, carbohydrate antigen 125 (CA125), and neuron-specific enolase (NSE), were selected for constructing the clinical model. The combined model (AUC: 0.932, 0.870) exhibited superior discriminative performance in the training and test cohorts compared to the clinical model (AUC: 0.850, 0.820) and the radiomic model (AUC: 0.876, 0.774). The calibration curves and DCA further confirmed the practicality of the combined model.

CONCLUSION

This study presented the development and validation of a combined model for distinguishing MPE and BPE. The combined model was a powerful tool for assisting in the clinical diagnosis of PE patients.

An automatic texture feature analysis framework of renal tumor: surgical, pathological, and molecular evaluation based on multi-phase abdominal CT

OBJECTIVES

To determine whether the texture feature analysis of multi-phase abdominal CT can provide a robust prediction of benign and malignant, histological subtype, pathological stage, nephrectomy risk, pathological grade, and Ki67 index in renal tumor.

METHODS

A total of 1051 participants with renal tumor were split into the internal cohort (850 patients from four different hospitals) and the external testing cohort (201 patients from another local hospital). The proposed framework comprised a 3D-kidney and tumor segmentation model by 3D-UNet, a feature extractor for the regions of interest based on radiomics and image dimension reduction, and the six classifiers by XGBoost. A quantitative model interpretation method called SHAP was used to explore the contribution of each feature.

RESULTS

The proposed multi-phase abdominal CT model provides robust prediction for benign and malignant, histological subtype, pathological stage, nephrectomy risk, pathological grade, and Ki67 index in the internal validation set, with the AUROC values of 0.88 ± 0.1, 0.90 ± 0.1, 0.91 ± 0.1, 0.89 ± 0.1, 0.84 ± 0.1, and 0.88 ± 0.1, respectively. The external testing set also showed impressive results, with AUROC values of 0.83 ± 0.1, 0.83 ± 0.1, 0.85 ± 0.1, 0.81 ± 0.1, 0.79 ± 0.1, and 0.81 ± 0.1, respectively. The radiomics feature including the first-order statistics, the tumor size-related morphology, and the shape-related tumor features contributed most to the model predictions.

CONCLUSIONS

Automatic texture feature analysis of abdominal multi-phase CT provides reliable predictions for multi-tasks, suggesting the potential usage of clinical application.

CLINICAL RELEVANCE STATEMENT

The automatic texture feature analysis framework, based on multi-phase abdominal CT, provides robust and reliable predictions for multi-tasks. These valuable insights can serve as a guiding tool for clinical diagnosis and treatment, making medical imaging an essential component in the process.

KEY POINTS

• The automatic texture feature analysis framework based on multi-phase abdominal CT can provide more accurate prediction of benign and malignant, histological subtype, pathological stage, nephrectomy risk, pathological grade, and Ki67 index in renal tumor. • The quantitative decomposition of the prediction model was conducted to explore the contribution of the extracted feature. • The study involving 1051 patients from 5 medical centers, along with a heterogeneous external data testing strategy, can be seamlessly transferred to various tasks involving new datasets.

Radiomics in CT and MR imaging of the liver and pancreas: tools with potential for clinical application

Radiomics allows the extraction of quantitative imaging features from clinical magnetic resonance imaging (MRI) and computerized tomography (CT) studies. The advantages of radiomics have primarily been exploited in oncological applications, including better characterization and staging of oncological lesions and prediction of patient outcomes and treatment response. The potential introduction of radiomics in the clinical setting requires the establishment of a standardized radiomics pipeline and a quality assurance program. Radiomics and texture analysis of the liver have improved the differentiation of hypervascular lesions such as adenomas, focal nodular hyperplasia, and hepatocellular carcinoma (HCC) during the arterial phase, and in the pretreatment determination of HCC prognostic factors (e.g., tumor grade, microvascular invasion, Ki-67 proliferation index). Radiomics of pancreatic CT and MR images has enhanced pancreatic ductal adenocarcinoma detection and its differentiation from pancreatic neuroendocrine tumors, mass-forming chronic pancreatitis, or autoimmune pancreatitis. Radiomics can further help to better characterize incidental pancreatic cystic lesions, accurately discriminating benign from malignant intrapancreatic mucinous neoplasms. Nonetheless, despite their encouraging results and exciting potential, these tools have yet to be implemented in the clinical setting. This non-systematic review will describe the essential steps in the implementation of the radiomics and feature extraction workflow from liver and pancreas CT and MRI studies for their potential clinical application. A succinct overview of reported radiomics applications in the liver and pancreas and the challenges and limitations of their implementation in the clinical setting is also discussed, concluding with a brief exploration of the future perspectives of radiomics in the gastroenterology field.

Performances and variability of CT radiomics for the prediction of microvascular invasion and survival in patients with HCC: a matter of chance or standardisation?

OBJECTIVES

To measure the performance and variability of a radiomics-based model for the prediction of microvascular invasion (MVI) and survival in patients with resected hepatocellular carcinoma (HCC), simulating its sequential development and application.

METHODS

This study included 230 patients with 242 surgically resected HCCs who underwent preoperative CT, of which 73/230 (31.7%) were scanned in external centres. The study cohort was split into training set (158 patients, 165 HCCs) and held-out test set (72 patients, 77 HCCs), stratified by random partitioning, which was repeated 100 times, and by a temporal partitioning to simulate the sequential development and clinical use of the radiomics model. A machine learning model for the prediction of MVI was developed with least absolute shrinkage and selection operator (LASSO). The concordance index (C-index) was used to assess the value to predict the recurrence-free (RFS) and overall survivals (OS).

RESULTS

In the 100-repetition random partitioning cohorts, the radiomics model demonstrated a mean AUC of 0.54 (range 0.44-0.68) for the prediction of MVI, mean C-index of 0.59 (range 0.44-0.73) for RFS, and 0.65 (range 0.46-0.86) for OS in the held-out test set. In the temporal partitioning cohort, the radiomics model yielded an AUC of 0.50 for the prediction of MVI, a C-index of 0.61 for RFS, and 0.61 for OS, in the held-out test set.

CONCLUSIONS

The radiomics models had a poor performance for the prediction of MVI with a large variability in the model performance depending on the random partitioning. Radiomics models demonstrated good performance in the prediction of patient outcomes.

CLINICAL RELEVANCE STATEMENT

Patient selection within the training set strongly influenced the performance of the radiomics models for predicting microvascular invasion; therefore, a random approach to partitioning a retrospective cohort into a training set and a held-out set seems inappropriate.

KEY POINTS

• The performance of the radiomics models for the prediction of microvascular invasion and survival widely ranged (AUC range 0.44-0.68) in the randomly partitioned cohorts. • The radiomics model for the prediction of microvascular invasion was unsatisfying when trying to simulate its sequential development and clinical use in a temporal partitioned cohort imaged with a variety of CT scanners. • The performance of the radiomics models for the prediction of survival was good with similar performances in the 100-repetition random partitioning and temporal partitioning cohorts.

An automated surgical decision-making framework for partial or radical nephrectomy based on 3D-CT multi-level anatomical features in renal cell carcinoma

OBJECTIVES

To determine whether 3D-CT multi-level anatomical features can provide a more accurate prediction of surgical decision-making for partial or radical nephrectomy in renal cell carcinoma.

METHODS

This is a retrospective study based on multi-center cohorts. A total of 473 participants with pathologically proved renal cell carcinoma were split into the internal training and the external testing set. The training set contains 412 cases from five open-source cohorts and two local hospitals. The external testing set includes 61 participants from another local hospital. The proposed automatic analytic framework contains the following modules: a 3D kidney and tumor segmentation model constructed by 3D-UNet, a multi-level feature extractor based on the region of interest, and a partial or radical nephrectomy prediction classifier by XGBoost. The fivefold cross-validation strategy was used to get a robust model. A quantitative model interpretation method called the Shapley Additive Explanations was conducted to explore the contribution of each feature.

RESULTS

In the prediction of partial versus radical nephrectomy, the combination of multi-level features achieved better performance than any single-level feature. For the internal validation, the AUROC was 0.93 ± 0.1, 0.94 ± 0.1, 0.93 ± 0.1, 0.93 ± 0.1, and 0.93 ± 0.1, respectively, as determined by the fivefold cross-validation. The AUROC from the optimal model was 0.82 ± 0.1 in the external testing set. The tumor shape Maximum 3D Diameter plays the most vital role in the model decision.

CONCLUSIONS

The automated surgical decision framework for partial or radical nephrectomy based on 3D-CT multi-level anatomical features exhibits robust performance in renal cell carcinoma. The framework points the way towards guiding surgery through medical images and machine learning.

CLINICAL RELEVANCE STATEMENT

We proposed an automated analytic framework that can assist surgeons in partial or radical nephrectomy decision-making. The framework points the way towards guiding surgery through medical images and machine learning.

KEY POINTS

• The 3D-CT multi-level anatomical features provide a more accurate prediction of surgical decision-making for partial or radical nephrectomy in renal cell carcinoma. • The data from multicenter study and a strict fivefold cross-validation strategy, both internal validation set and external testing set, can be easily transferred to different tasks of new datasets. • The quantitative decomposition of the prediction model was conducted to explore the contribution of each extracted feature.

A primer on artificial intelligence in pancreatic imaging

Artificial Intelligence (AI) is set to transform medical imaging by leveraging the vast data contained in medical images. Deep learning and radiomics are the two main AI methods currently being applied within radiology. Deep learning uses a layered set of self-correcting algorithms to develop a mathematical model that best fits the data. Radiomics converts imaging data into mineable features such as signal intensity, shape, texture, and higher-order features. Both methods have the potential to improve disease detection, characterization, and prognostication. This article reviews the current status of artificial intelligence in pancreatic imaging and critically appraises the quality of existing evidence using the radiomics quality score.

An Automated Prognostic Model for Pancreatic Ductal Adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) constitutes a leading cause of cancer-related mortality despite advances in detection and treatment methods. While computed tomography (CT) serves as the current gold standard for initial evaluation of PDAC, its prognostic value remains limited, as it relies on diagnostic stage parameters encompassing tumor size, lymph node involvement, and metastasis. Radiomics have recently shown promise in predicting postoperative survival of PDAC patients; however, they rely on manual pancreas and tumor delineation by clinicians. In this study, we collected a dataset of pre-operative CT scans from a cohort of 40 PDAC patients to evaluate a fully automated pipeline for survival prediction. Employing nnU-Net trained on an external dataset, we generated automated pancreas and tumor segmentations. Subsequently, we extracted 854 radiomic features from each segmentation, which we narrowed down to 29 via feature selection. We then combined these features with the Tumor, Node, Metastasis (TNM) system staging parameters, as well as the patient's age. We trained a random survival forest model to perform an overall survival prediction over time, as well as a random forest classifier for the binary classification of two-year survival, using repeated cross-validation for evaluation. Our results exhibited promise, with a mean C-index of 0.731 for survival modeling and a mean accuracy of 0.76 in two-year survival prediction, providing evidence of the feasibility and potential efficacy of a fully automated pipeline for PDAC prognostication. By eliminating the labor-intensive manual segmentation process, our streamlined pipeline demonstrates an efficient and accurate prognostication process, laying the foundation for future research endeavors.

Is the radiomics-clinical combined model helpful in distinguishing between pancreatic cancer and mass-forming pancreatitis?

PURPOSE

To develop CT-based radiomics models for distinguishing between resectable PDAC and mass-forming pancreatitis (MFP) and to provide a non-invasive tool for cases of equivocal imaging findings with EUS-FNA needed.

METHODS

A total of 201 patients with resectable PDAC and 54 patients with MFP were included. Development cohort: patients without preoperative EUS-FNA (175 PDAC cases, 38 MFP cases); validation cohort: patients with EUS-FNA (26 PDAC cases, 16 MFP cases). Two radiomic signatures (LASSOscore, PCAscore) were developed based on the LASSO model and principal component analysis. LASSOCli and PCACli prediction models were established by combining clinical features with CT radiomic features. ROC analysis and decision curve analysis (DCA) were performed to evaluate the utility of the model versus EUS-FNA in the validation cohort.

RESULTS

In the validation cohort, the radiomic signatures (LASSOscore, PCAscore) were both effective in distinguishing between resectable PDAC and MFP (AUC_LASSO = 0.743, 95% CI: 0.590-0.896; AUC_PCA = 0.788, 95% CI: 0.639-0.938) and improved the diagnostic accuracy of the baseline onlyCli model (AUC_onlyCli = 0.760, 95% CI: 0.614-0.960) after combination with variables including age, CA19-9, and the double-duct sign (AUC_PCACli = 0.880, 95% CI: 0.776-0.983; AUC_LASSOCli = 0.825, 95% CI: 0.694-0.955). The PCACli model showed comparable performance to FNA (AUC_FNA = 0.810, 95% CI: 0.685-0.935). In DCA, the net benefit of the PCACli model was superior to that of EUS-FNA, avoiding biopsies in 70 per 1000 patients at a risk threshold of 35%.

CONCLUSIONS

The PCACli model showed comparable performance with EUS-FNA in discriminating resectable PDAC from MFP.

Role of radiomics in the diagnosis and treatment of gastrointestinal cancer

Gastrointestinal cancer (GIC) has high morbidity and mortality as one of the main causes of cancer death. Preoperative risk stratification is critical to guide patient management, but traditional imaging studies have difficulty predicting its biological behavior. The emerging field of radiomics allows the conversion of potential pathophysiological information in existing medical images that cannot be visually recognized into high-dimensional quantitative image features. Tumor lesion characterization, therapeutic response evaluation, and survival prediction can be achieved by analyzing the relationships between these features and clinical and genetic data. In recent years, the clinical application of radiomics to GIC has increased dramatically. In this editorial, we describe the latest progress in the application of radiomics to GIC and discuss the value of its potential clinical applications, as well as its limitations and future directions.

Pancreatic cancer, radiomics and artificial intelligence

Patients with pancreatic ductal adenocarcinoma (PDAC) are generally classified into four categories based on contrast-enhanced CT at diagnosis: resectable, borderline resectable, unresectable, and metastatic disease. In the initial grading and staging of PDAC, structured radiological templates are useful but limited, as there is a need to define the aggressiveness and microscopic disease stage of these tumours to ensure adequate treatment allocation. Quantitative imaging analysis allows radiomics and dynamic imaging features to provide information of clinical outcomes, and to construct clinical models based on radiomics signatures or imaging phenotypes. These quantitative features may be used as prognostic and predictive biomarkers in clinical decision-making, enabling personalised management of advanced PDAC. Deep learning and convolutional neural networks also provide high level bioinformatics tools that can help define features associated with a given aspect of PDAC biology and aggressiveness, paving the way to define outcomes based on these features. Thus, the prediction of tumour phenotype, treatment response and patient prognosis may be feasible by using such comprehensive and integrated radiomics models. Despite these promising results, quantitative imaging is not ready for clinical implementation in PDAC. Limitations include the instability of metrics and lack of external validation. Large properly annotated datasets, including relevant semantic features (demographics, blood markers, genomics), image harmonisation, robust radiomics analysis, clinically significant tasks as outputs, comparisons with gold-standards (such as TNM or pretreatment classifications) and fully independent validation cohorts, will be required for the development of trustworthy radiomics and artificial intelligence solutions to predict PDAC aggressiveness in a clinical setting.

Correlation of transcriptional subtypes with a validated CT radiomics score in resectable pancreatic ductal adenocarcinoma

OBJECTIVES

Transcriptional classifiers (Bailey, Moffitt and Collison) are key prognostic factors of pancreatic ductal adenocarcinoma (PDAC). Among these classifiers, the squamous, basal-like, and quasimesenchymal subtypes overlap and have inferior survival. Currently, only an invasive biopsy can determine these subtypes, possibly resulting in treatment delay. This study aimed to investigate the association between transcriptional subtypes and an externally validated preoperative CT-based radiomic prognostic score (Rad-score).

METHODS

We retrospectively evaluated 122 patients who underwent resection for PDAC. All treatment decisions were determined at multidisciplinary tumor boards. Tumor Rad-score values from preoperative CT were dichotomized into high or llow categories. The primary endpoint was the correlation between the transcriptional subtypes and the Rad-score using multivariable linear regression, adjusting for clinical and histopathological variables (i.e., tumor size). Prediction of overall survival (OS) was secondary endpoint.

RESULTS

The Bailey transcriptional classifier significantly associated with the Rad-score (coefficient = 0.31, 95% confidence interval [CI]: 0.13-0.44, p = 0.001). Squamous subtype was associated with high Rad-scores while non-squamous subtype was associated with low Rad-scores (adjusted p = 0.03). Squamous subtype and high Rad-score were both prognostic for OS at multivariable analysis with hazard ratios (HR) of 2.79 (95% CI: 1.12-6.92, p = 0.03) and 4.03 (95% CI: 1.42-11.39, p = 0.01), respectively.

CONCLUSIONS

In patients with resectable PDAC, an externally validated prognostic radiomic model derived from preoperative CT is associated with the Bailey transcriptional classifier. Higher Rad-scores were correlated with the squamous subtype, while lower Rad-scores were associated with the less lethal subtypes (immunogenic, ADEX, pancreatic progenitor).

KEY POINTS

• The transcriptional subtypes of PDAC have been shown to have prognostic importance but they require invasive biopsy to be assessed. • The Rad-score radiomic biomarker, which is obtained non-invasively from preoperative CT, correlates with the Bailey squamous transcriptional subtype and both are negative prognostic biomarkers. • The Rad-score is a promising non-invasive imaging biomarker for personalizing neoadjuvant approaches in patients undergoing resection for PDAC, although additional validation studies are required.

Multimodal survival prediction in advanced pancreatic cancer using machine learning

Background

Existing risk scores appear insufficient to assess the individual survival risk of patients with advanced pancreatic ductal adenocarcinoma (PDAC) and do not take advantage of the variety of parameters that are collected during clinical care.

Methods

In this retrospective study, we built a random survival forest model from clinical data of 203 patients with advanced PDAC. The parameters were assessed before initiation of systemic treatment and included age, CA19-9, C-reactive protein, metastatic status, neutrophil-to-lymphocyte ratio and total serum protein level. Separate models including imaging and molecular parameters were built for subgroups.

Results

Over the entire cohort, a model based on clinical parameters achieved a c-index of 0.71. Our approach outperformed the American Joint Committee on Cancer (AJCC) staging system and the modified Glasgow Prognostic Score (mGPS) in the identification of high- and low-risk subgroups. Inclusion of the KRAS p.G12D mutational status could further improve the prediction, whereas radiomics data of the primary tumor only showed little benefit. In an external validation cohort of PDAC patients with liver metastases, our model achieved a c-index of 0.67 (mGPS: 0.59).

Conclusions

The combination of multimodal data and machine-learning algorithms holds potential for personalized prognostication in advanced PDAC already at diagnosis.

•

We developed a machine-learning-based prediction model that outperforms the AJCC staging system and mGPS.

•

Applying our model to an external validation cohort demonstrates generalizability.

•

Explainable machine learning enables to understand the decision making of our model and identifies relevant parameters.

•

Combining clinical, imaging and genetic data holds potential for personalized prognostication in advanced PDAC.

A systematic review of prognosis predictive role of radiomics in pancreatic cancer: heterogeneity markers or statistical tricks?

OBJECTIVES

We aimed to systematically evaluate the prognostic prediction accuracy of radiomics features extracted from pre-treatment imaging in patients with pancreatic ductal adenocarcinoma (PDAC).

METHODS

Radiomics literature on overall survival (OS) prediction of PDAC were all included in this systematic review. A further meta-analysis was performed on the effect size of first-order entropy. Methodological quality and risk of bias of the included studies were assessed by the radiomics quality score (RQS) and prediction model risk of bias assessment tool (PROBAST).

RESULTS

Twenty-three studies were finally identified in this review. Two (8.7%) studies compared prognosis prediction ability between radiomics model and TNM staging model by C-index, and both showed a better performance of the radiomics. Twenty-one (91.3%) studies reported significant predictive values of radiomics features. Nine (39.1%) studies were included in the meta-analysis, and it showed a significant correlation between first-order entropy and OS (HR 1.66, 95%CI 1.18-2.34). RQS assessment revealed validation was only performed in 5 (21.7%) studies on internal datasets and 2 (8.7%) studies on external datasets. PROBAST showed that 22 (95.7%) studies have a high risk of bias in participants because of the retrospective study design.

CONCLUSION

First-order entropy was significantly associated with OS and might improve the accuracy of PDAC prognosis prediction. Existing studies were poorly validated, and it should be noted in future studies. Modification of PROBAST for radiomics studies is necessary since the strict requirements of prospective study design may not be applicable to the demand for a large sample size in the model construction stage.

KEY POINTS

• Radiomics based on the primary lesion holds great potential for prognosis prediction. First-order entropy was significantly associated with the overall survival of PDAC and might improve the accuracy of current PDAC prognosis prediction. • We strongly recommend that at least an internal validation should be conducted in any radiomics study. Attention should be paid to the complex relationships between radiomics features. • Due to the close relationship between radiomics and big data, the strict requirement of prospective study design in PROABST may not be appropriate for radiomics studies. A balance between study types and sample sizes for radiomics studies needs to be found in the model construction stage.

Setting the Research Agenda for Clinical Artificial Intelligence in Pancreatic Adenocarcinoma Imaging

Simple Summary

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest cancers worldwide, associated with a 98% loss of life expectancy and a 30% increase in disability-adjusted life years. Image-based artificial intelligence (AI) can help improve outcomes for PDAC given that current clinical guidelines are non-uniform and lack evidence-based consensus. However, research on image-based AI for PDAC is too scattered and lacking in sufficient quality to be incorporated into clinical workflows. In this review, an international, multi-disciplinary team of the world’s leading experts in pancreatic cancer breaks down the patient pathway and pinpoints the current clinical touchpoints in each stage. The available PDAC imaging AI literature addressing each pathway stage is then rigorously analyzed, and current performance and pitfalls are identified in a comprehensive overview. Finally, the future research agenda for clinically relevant, image-driven AI in PDAC is proposed.

Abstract

Pancreatic ductal adenocarcinoma (PDAC), estimated to become the second leading cause of cancer deaths in western societies by 2030, was flagged as a neglected cancer by the European Commission and the United States Congress. Due to lack of investment in research and development, combined with a complex and aggressive tumour biology, PDAC overall survival has not significantly improved the past decades. Cross-sectional imaging and histopathology play a crucial role throughout the patient pathway. However, current clinical guidelines for diagnostic workup, patient stratification, treatment response assessment, and follow-up are non-uniform and lack evidence-based consensus. Artificial Intelligence (AI) can leverage multimodal data to improve patient outcomes, but PDAC AI research is too scattered and lacking in quality to be incorporated into clinical workflows. This review describes the patient pathway and derives touchpoints for image-based AI research in collaboration with a multi-disciplinary, multi-institutional expert panel. The literature exploring AI to address these touchpoints is thoroughly retrieved and analysed to identify the existing trends and knowledge gaps. The results show absence of multi-institutional, well-curated datasets, an essential building block for robust AI applications. Furthermore, most research is unimodal, does not use state-of-the-art AI techniques, and lacks reliable ground truth. Based on this, the future research agenda for clinically relevant, image-driven AI in PDAC is proposed.