Imaging features of gadoxetic acid-enhanced MR imaging for evaluation of tumor-infiltrating CD8 cells and PD-L1 expression in hepatocellular carcinoma

Hepatobiliary and Pancreatic Neoplasms: Essential Predictive Imaging Features for Personalized Therapy

Tumor biologic characteristics encompassing histopathologic, immune microenvironmental, genetic, and molecular aspects are becoming indispensable factors to be considered in treatment strategies for patients with cancer. Innovations in oncologic treatment have broadened the range of therapeutic approaches and now hold promise for treatments personalized according to tumor biologic characteristics. Particularly for hepatobiliary and pancreatic neoplasms, the advent of cytostatic agents such as molecularly targeted agents and immune checkpoint inhibitors, which differ markedly from conventional cytotoxic agents, has contributed to advances in clinical practice. These cytostatic agents increase the potential for curative-intent treatment of unresectable cancers by reducing tumor volume. Radiologic examinations are of more interest than ever to noninvasively obtain information about tumor biologic features. Radiomics represents an invaluable research method for elucidating associations between tumor biologic characteristics and radiologic imaging findings, but their applicability in daily clinical practice remains challenging. Various radiologic predictive findings for tumor biologic characteristics have already been proposed for hepatobiliary and pancreatic neoplasms. Radiologists must gain familiarity with these findings and the roles they have in predicting the clinical prognosis and treatment efficacy. In addition, radiologists should explore the potential applications of these imaging findings to current treatment strategies for the coming era of personalized medicine. The authors describe predictive findings using CT and MRI for diagnosis of hepatocellular carcinoma, colorectal liver metastases, intrahepatic cholangiocarcinoma, and pancreatic adenocarcinoma, with correlations to pathologic, immunologic, molecular, and genetic background factors. ©RSNA, 2025 Supplemental material is available for this article. See the invited commentary by Ronot in this issue.

Application of Intravoxel Incoherent Motion in the Prediction of Intra-Tumoral Tertiary Lymphoid Structures in Hepatocellular Carcinoma

Objective

To explore the value of intravoxel incoherent motion (IVIM) sequences in predicting intra-tumoral tertiary lymphoid structures (TLSs).

Materials and Methods

This prospective study pre-operatively enrolled hepatocellular carcinoma (HCC) patients who underwent magnetic resonance imaging including IVIM sequences, between January 2019 and April 2021. Intra-tumoral TLSs presence was assessed on pathological slide images. Clinical and radiological characteristics were collected. IVIM quantitative parameters and radiomics features were obtained based on the whole delineated tumor volume. By using feature selection techniques, 22 radiomics features, clinical-radiological features (lymphocyte count and satellite nodules), and IVIM parameters (apparent diffusion coefficient (ADC_90Percentile), perfusion fraction (f_Maximum)) were selected. The logistic regression algorithm was used to construct the prediction model based on the combination of these features. The diagnostic performance was assessed using the area under the receiver operating characteristic (AUC). The recurrence-free survival (RFS) was evaluated with Kaplan-Meier curves.

Results

A total of 168 patients were divided into training (n=128) and testing (n=40) cohorts (mean age: 56.83±14.43 years; 149 [88.69%] males; 130 TLSs+). In testing cohort, the model combining multimodal features demonstrated a good performance (AUC: 0.86) and significantly outperformed models based on single-modality features. The model based on radiomics features (AUC: 0.80) had better performance than other features, including IVIM parameter maps (ADC_90Percentile and f_Maximum, AUC: 0.72) and clinical-radiological characteristics (satellite nodules and lymphocyte counts, AUC: 0.59). TLSs+ patients had higher RFS than TSLs- patients (all p <0.05).

Conclusion

The nomogram based on the proposed model can be used as a pre-operative predictive biomarker of TLSs.

Critical Relevance Statement

The nomogram incorporating IVIM sequences may serve as a pre-operative predictive biomarker of intra-tumoral tertiary lymphoid structure (TLS) status.

Hypointense Findings on Hepatobiliary Phase MR Images

Hepatobiliary (HB) contrast agents are increasingly valuable diagnostic tools in MRI, offering a wider range of applications as their clinical use expands. Normal hepatocytes take up HB contrast agents, which are subsequently excreted in bile. This property creates a distinct HB phase providing valuable insights into liver function and biliary anatomy. HB contrast agents can assist in diagnosing a broad spectrum of HB diseases ranging from diffuse liver disease to focal hepatic lesions and can delineate anatomic details of the biliary tree. Understanding the pharmacodynamics of HB contrast agents is paramount to their appropriate clinical application and troubleshooting. HB phase hypointensity can arise from various diffuse and focal abnormalities that may or may not be associated with biliary excretion. Hypointensity during the HB phase can be broadly grouped into diffuse hypointensity, regional hypointensity, and focal lesions for better evaluation of the underlying cause. Abnormalities may arise from hepatic parenchymal, biliary, or vascular causes, or a combination thereof in each of the broad groups. Recognition of a suboptimal hypointense HB phase is important in the evaluation of focal lesions in patients with cirrhosis of the liver and particularly in those with hepatocellular carcinoma. Furthermore, hypointensity can also suggest the aggressiveness of malignancies such as hepatocellular carcinoma or colorectal metastases, which may affect the prognosis. It is essential to consider all imaging findings relative to the clinical context and the complete set of the MRI sequences performed for diagnosis of liver abnormalities. This comprehensive approach minimizes the risk of misinterpretation or pitfalls. The authors aim to equip radiologists with key insights for accurately understanding hypointensity in the HB phase, ultimately leading to more accurate diagnoses. ©RSNA, 2025 Supplemental material is available for this article.

Prediction of PD-L1 expression in unresectable hepatocellular carcinoma with gadoxetic acid-enhanced MRI

OBJECTIVES

To develop a model to predict programmed death-ligand 1 (PD-L1) expression in unresectable hepatocellular carcinoma (HCC) based on gadoxetic acid-enhanced magnetic resonance imaging (MRI) findings and clinical characteristics.

MATERIALS AND METHODS

We enrolled patients with unresectable HCC who underwent gadoxetic acid-enhanced MRI between January 2021 and May 2023. Immunohistochemical staining of PD-L1 was performed on a biopsy specimen. Patients with a history of any prior treatment for HCC or those lacking an MRI scan within 30 days of the biopsy date were excluded. Using the clinical and MRI findings, we developed a PD-L1 prediction score using logistic regression.

RESULTS

This study included 49 patients with HCC (median age, 64 years; interquartile range, 57-73 years; 44 men). Among these, 15 (31 %) were positive for PD-L1 expression. The PD-L1 prediction score was defined as the sum of arterial phase hypoenhancement (score 1), necrosis (score 1), and AFP >4000 ng/mL (score 2). The AUC value of the PD-L1 prediction score was 0.838 (95 % confidence interval [CI], 0.715-0.962). When the PD-L1 prediction score was ≥3, the sensitivity, specificity, and positive predictive value of PD-L1 positivity were 67 %, 91 %, and 77 %, respectively.

CONCLUSION

We developed a PD-L1 prediction score for unresectable HCC with high specificity that could potentially contribute to the identification of effective candidates for immune checkpoint inhibitors.

Prediction of Efficacy for Atezolizumab/Bevacizumab in Unresectable Hepatocellular Carcinoma with Hepatobiliary-Phase Gadolinium Ethoxybenzyl-Diethylenetriaminepentaacetic Acid MRI

BACKGROUND

This study aimed to examine whether the coefficient of variation (CV) in the hepatobiliary-phase (HBP) of Gd-EOB-DTPA-MRI could be an independent predictive factor for tumor progression.

METHODS

Patients who underwent Gd-EOB-DTPA-MRI before Atezolizumab/bevacizumab therapy at six affiliated institutions between 2018 and 2022 were included. CV for each patient was calculated as the mean value for up to five tumors larger than 10 mm, and CV of the whole tumor was calculated using LIFEx software. The tumor response was evaluated within 6-10 weeks. The primary endpoint was to investigate the predictive factors, including CV, related to tumor progression using logistic regression analysis. The secondary endpoints were tumor response rate and progression-free survival (PFS) based on CV.

RESULTS

Of the 46 enrolled patients, 13 (28.3%) underwent early progressive disease. Multivariate analysis revealed that a high CV (≥0.22) was an independent predictive factor for tumor progression (p = 0.043). Patients with a high CV had significantly frequent PD than those with a low CV (43.5 vs. 13.0%, p = 0.047). Patients with a high CV tended to have shorter PFS than those with a low CV (3.5 vs. 6.7 months, p = 0.071).

CONCLUSION

Quantitative analysis using CV in the HBP of Gd-EOB-DTPA-MRI may be useful for predicting tumor progression for atezolizumab/bevacizumab therapy.

Hepatobiliary-phase gadolinium ethoxybenzyl-diethylenetriaminepentaacetic acid MRI for pretreatment prediction of efficacy-to-standard-therapies based on Barcelona Clinic Liver Cancer algorithm: an up-to-date review

Recent advances in systemic therapy have had major impacts on treatment strategies for hepatocellular carcinoma (HCC). The 2022 Barcelona Clinic Liver Cancer (BCLC) guidelines incorporate a new section on clinical decision-making for personalized medicine, although the first treatment suggested by the BCLC guidelines is based on solid scientific evidence. More than ever before, the appropriate treatment strategy must be selected prior to the initiation of therapy for HCC. Gadolinium ethoxybenzyl-diethylenetriaminepentaacetic acid magnetic resonance imaging (Gd-EOB-DTPA-MRI) is essential for liver imaging and the hepatobiliary phase (HBP) of EOB-MRI reflects the expression of organic anion transporting polypeptide (OATP) transporters. Molecules associated with OATP expression are relevant in the molecular classification of HCC subclasses, and EOB-MRI is becoming increasingly important with advances in the molecular and genetic understanding of HCC. In this review, we describe imaging findings for the pretreatment prediction of response to standard therapies for HCC based on the BCLC algorithm using the HBP of EOB-MRI, with specific attention to the molecular background of OATPs. A more complete understanding of these findings will help radiologists suggest appropriate treatments and clinical follow-ups and could lead to the development of more personalized treatment strategies in the future. CLINICAL RELEVANCE STATEMENT: In the coming era of personalized medicine, HBP of EOB-MRI reflecting molecular and pathological factors could play a predictive role in the therapeutic efficacy of HCC and contribute to treatment selection. KEY POINTS: • Imaging features of hepatobiliary phase predict treatment efficacy prior to therapy and contribute to treatment choice. • Wnt/β-catenin activation associated with organic anion transporting polypeptide expression is involved in the tumor immune microenvironment and chemo-responsiveness. • Peritumoral hypointensity of hepatobiliary phase reflecting microvascular invasion affects the therapeutic efficacy of locoregional to systemic therapy.

Reproducibility and usefulness of quantitative apparent diffusion coefficient measurements for predicting program death-ligand 1 expression in nasopharyngeal carcinoma

BACKGROUND

Accurate assessment of programmed death-ligand 1 (PD-L1) expression status in nasopharyngeal carcinoma (NPC) before immunotherapy is crucial. We aimed to explore the reproducibility and usefulness of the quantitative apparent diffusion coefficient (ADC) measurements for predicting PD-L1expression status in NPC.

METHODS

We retrospectively recruited 134 NPC patients who underwent MRI scans and PD-L1 detection. A PD-L1 combined positive score (CPS) ≥ 20 was identified as high expression status. Patients were divide into two cohorts based on the MRI scanning devices, including a 1.5-T MRI cohort (n = 85, 44 PD-L1 high expression) and a 3.0-T MRI cohort (n = 49, 24 PD-L1 high expression). The mean ADC (ADCmean), minimum ADC (ADCmin) and maximal ADC (ADCmax) values were independently measured by two observers. The ADC measurement reproducibility was assessed by interclass correlation coefficients (ICC). The correlations between ADC parameters and CPS were analyzed by spearman's correlation coefficient (r), and the performance for PD-L1expression status prediction was assessed by the area under receiver operating characteristic curve (AUC).

RESULTS

The measurement reproducibility of ADCmean, ADCmin and ADCmax was good in the 1.5-T MRI cohort (ICC: 0.843-0.930) and 3.0-T MRI cohort (ICC: 0.929-0.960). The ADCmean, ADCmin, and ADCmax tended to inversely correlate with the CPS (r:-0.37 - -0.52 in the 1.5-T MRI cohort, and - 0.52 - -0.60 in the 3.0-T MRI cohort; P all < 0.01). The ADCmean, ADCmin and ADCmax yielded the AUC of 0.756 (95% CI: 0.651, 0.861), 0.689 (95% CI: 0.576, 0.802), and 0.733 (95%CI: 0.626, 0.839) in the 1.5-T MRI cohort and 0.820 (95%CI: 0.703, 0.937), 0.755 (95% CI: 0.616, 0.894), and 0.760 (95%CI: 0.627, 0.893) in the 3.0-T MRI cohort for predicting PD-L1 high expression status, respectively.

CONCLUSION

ADC measurements may act as a reproducible and feasible method to predict PD-L1 expression status in NPC.

Response Assessment of Primary Liver Tumors to Novel Therapies: an Imaging Perspective

The latest developments in cancer immunotherapy, namely the introduction of immune checkpoint inhibitors, have led to a fundamental change in advanced cancer treatments. Imaging is crucial to identify tumor response accurately and delineate prognosis in immunotherapy-treated patients. Simultaneously, advances in image acquisition techniques, notably functional and molecular imaging, have facilitated more accurate pretreatment evaluation, assessment of response to therapy, and monitoring for tumor recurrence. Traditional approaches to assessing tumor progression, such as RECIST, rely on changes in tumor size, while new strategies for evaluating tumor response to therapy, such as the mRECIST and the EASL, rely on tumor enhancement. Moreover, the assessment of tumor volume, enhancement, cellularity, and perfusion are some novel techniques that have been investigated. Validation of these novel approaches should rely on comparing their results with those of standard evaluation methods (EASL, mRECIST) while considering the ultimate outcome, which is patient survival. More recently, immunotherapy has been used in the management of primary liver tumors. However, little is known about its efficacy. This article reviews imaging modalities and techniques for assessing tumor response and survival in immunotherapy-treated patients with primary hepatic malignancies.

The Era of Immunotherapy in Hepatocellular Carcinoma: The New Mission and Challenges of Magnetic Resonance Imaging

In recent years, significant advancements in immunotherapy for hepatocellular carcinoma (HCC) have shown the potential to further improve the prognosis of patients with advanced HCC. However, in clinical practice, there is still a lack of effective biomarkers for identifying the patient who would benefit from immunotherapy and predicting the tumor response to immunotherapy. The immune microenvironment of HCC plays a crucial role in tumor development and drug responses. However, due to the complexity of immune microenvironment, currently, no single pathological or molecular biomarker can effectively predict tumor responses to immunotherapy. Magnetic resonance imaging (MRI) images provide rich biological information; existing studies suggest the feasibility of using MRI to assess the immune microenvironment of HCC and predict tumor responses to immunotherapy. Nevertheless, there are limitations, such as the suboptimal performance of conventional MRI sequences, incomplete feature extraction in previous deep learning methods, and limited interpretability. Further study needs to combine qualitative features, quantitative parameters, multi-omics characteristics related to the HCC immune microenvironment, and various deep learning techniques in multi-center research cohorts. Subsequently, efforts should also be undertaken to construct and validate a visual predictive tool of tumor response, and assess its predictive value for patient survival benefits. Additionally, future research endeavors must aim to provide an accurate, efficient, non-invasive, and highly interpretable method for predicting the effectiveness of immune therapy.

Prediction of therapeutic response of advanced hepatocellular carcinoma to combined targeted immunotherapy by MRI

PURPOSE

To assess the value of pre-treatment MRI in predicting treatment response to combined targeted immunotherapy in advanced hepatocellular carcinoma (HCC).

METHODS

Totally 35 HCC participants who underwent pre-treatment contrast-enhanced MRI and received combined tyrosine kinase inhibitor (TKI) and anti-PD-1 antibody treatment were enrolled. Univariable and multivariable logistic regression analyses were carried out for comparing clinical and MRI characteristics between patients with therapeutic response and those without. A predictive model based on MRI data and the corresponding nomogram were developed using data generated by multivariate analysis, and the diagnostic performance was evaluated. A cutoff for the combined index was measured by receiver operating characteristic curve analysis, and progression-free survival (PFS) rates were compared between cases with high and low combined index values.

RESULTS

Fifteen (42.86%) cases achieved overall response during treatment. Multivariable analysis revealed that homogeneous signal (odds ratio [OR] = 13.51, P = 0.010) and no arterial peritumoral enhancement (APE; OR = 10.29, P = 0.024) independently predicted treatment response. The combined model including both significant MRI parameters showed a satisfactory predictive performance with the largest area under the curve of 0.837 (95%CI 0.673-0.939), and both sensitivity and specificity of 80.0%. HCCs with high-combined index had higher PFS rate compared with those showing a low value (P = 0.034).

CONCLUSION

The combination of pre-treatment MRI features of homogeneous signal and no APE could be used for predicting treatment response to combined targeted immunotherapy in advanced HCC.

Novel Computed-Tomography-Based Transformer Models for the Noninvasive Prediction of PD-1 in Pre-Operative Settings

The expression status of programmed cell death protein 1 (PD-1) in patients with hepatocellular carcinoma (HCC) is associated with the checkpoint blockade treatment responses of PD-1/PD-L1. Thus, accurately and preoperatively identifying the status of PD-1 has great clinical implications for constructing personalized treatment strategies. To investigate the preoperative predictive value of the transformer-based model for identifying the status of PD-1 expression, 93 HCC patients with 75 training cohorts (2859 images) and 18 testing cohorts (670 images) were included. We propose a transformer-based network architecture, ResTransNet, that efficiently employs convolutional neural networks (CNNs) and self-attention mechanisms to automatically acquire a persuasive feature to obtain a prediction score using a nonlinear classifier. The area under the curve, receiver operating characteristic curve, and decision curves were applied to evaluate the prediction model's performance. Then, Kaplan-Meier survival analyses were applied to evaluate the overall survival (OS) and recurrence-free survival (RFS) in PD-1-positive and PD-1-negative patients. The proposed transformer-based model obtained an accuracy of 88.2% with a sensitivity of 88.5%, a specificity of 88.9%, and an area under the curve of 91.1% in the testing cohort.

Systemic therapies in hepatocellular carcinoma: Existing and emerging biomarkers for treatment response

Hepatocellular carcinoma (HCC) is the fifth most common malignancy and the third most common cause of cancer-related death worldwide. Due to asymptomatic patients in the early stage, most patients are diagnosed at an advanced stage and lose the opportunity for radical resection. In addition, for patients who underwent procedures with curative intent for early-stage HCC, up to 70% of patients may have disease recurrence within 5 years. With the advent of an increasing number of systemic therapy medications, we now have more options for the treatment of HCC. However, data from clinical studies show that with different combinations of regimens, the objective response rate is approximately 40%, and most patients will not respond to treatment. In this setting, biomarkers for predicting treatment response are of great significance for precise treatment, reducing drug side effects and saving medical resources. In this review, we summarized the existing and emerging biomarkers in the literature, with special emphasis on the pathways and mechanism underlying the prediction value of those biomarkers for systemic treatment response.

Preoperative prediction of glypican-3 positive expression in solitary hepatocellular carcinoma on gadoxetate-disodium enhanced magnetic resonance imaging

Purpose

As a coreceptor in Wnt and HGF signaling, glypican-3 (GPC-3) promotes the progression of tumor and is associated with a poor prognosis in hepatocellular carcinoma (HCC). GPC-3 has evolved as a target molecule in various immunotherapies, including chimeric antigen receptor T cell. However, its evaluation still relies on invasive histopathologic examination. Therefore, we aimed to develop an easy-to-use and noninvasive risk score integrating preoperative gadoxetic acid–enhanced magnetic resonance imaging (EOB-MRI) and clinical indicators to predict positive GPC-3 expression in HCC.

Methods and materials

Consecutive patients with surgically-confirmed solitary HCC who underwent preoperative EOB-MRI between January 2016 and November 2021 were retrospectively included. EOB-MRI features were independently evaluated by two masked abdominal radiologists and the expression of GPC-3 was determined by two liver pathologists. On the training dataset, a predictive scoring system for GPC-3 was developed against pathology via logistical regression analysis. Model performances were characterized by computing areas under the receiver operating characteristic curve (AUCs).

Results

A total of 278 patients (training set, n=156; internal validation set, n=39; external validation set, n=83) with solitary HCC (208 [75%] with positive GPC-3 expression) were included. Serum alpha-fetoprotein >10 ng/ml (AFP, odds ratio [OR]=2.3, four points) and five EOB-MR imaging features, including tumor size >3.0cm (OR=0.5, -3 points), nonperipheral “washout” (OR=3.0, five points), infiltrative appearance (OR=9.3, 10 points), marked diffusion restriction (OR=3.3, five points), and iron sparing in solid mass (OR=0.2, -7 points) were significantly associated with positive GPC-3 expression. The optimal threshold of scoring system for predicting GPC-3 positive expression was 5.5 points, with AUC 0.726 and 0.681 on the internal and external validation sets, respectively.

Conclusion

Based on serum AFP and five EOB-MRI features, we developed an easy-to-use and noninvasive risk score which could accurately predict positive GPC-3 HCC, which may help identify potential responders for GPC-3-targeted immunotherapy.

MRI-based Nomogram Predicts the Risk of Progression of Unresectable Hepatocellular Carcinoma After Combined Lenvatinib and anti-PD-1 Antibody Therapy

RATIONALE AND OBJECTIVES

Combined immune and anti-angiogenic treatment has shown promising results for unresectable hepatocellular carcinoma (HCC), but with a high risk of early progression. In this study, we aimed to investigate whether pre-treatment magnetic resonance imaging (MRI) features and MRI-based nomogram could predict the risk of disease progression of unresectable HCC after first-line lenvatinib/anti-PD-1 antibody therapy.

MATERIALS AND METHODS

Thirty-seven HCC participants with qualified pre-treatment contrast-enhanced MRI were enrolled. All patients received combined lenvatinib and anti-PD-1 antibody treatment. Progression free survival rate was analyzed using the Kaplan-Meier method. Potential clinical-radiological risk factors for progression were analyzed using the log-rank tests and Cox regression model. The performance of MRI-based nomogram was evaluated based on C-index, calibration, and decision curve analyses.

RESULTS

The 6-month and 12-month cumulative progression free survival rates were 59.5% (95% confidence interval (CI), 43.6%-75.4%) and 48.0% (95% CI, 31.7%-64.3%). On multivariate analysis, no or incomplete tumor capsule (hazard ratio (HR) = 15.215 [95% CI 2.707-85.529], p = 0.002), heterogeneous signal on T2-weighted imaging (HR = 28.179 [95% CI 2.437-325.838]; p = 0.008) and arterial contrast-to-noise ratio ≤95.45 (HR = 5.113 [95% CI 1.538-17.00]; p = 0.008) were independent risk factors for disease progression. Satisfactory predictive performance of the nomogram incorporating the three independent imaging features was obtained with a C-index value of 0.880 (95% CI 0.824-0.937), and the combined nomogram had more favorable clinical prediction performance than any single feature.

CONCLUSION

MRI features can be considered effective predictors of disease progression for unresectable HCC with first-line lenvatinib plus anti-PD-1 antibody therapy, and the combined MRI-based nomogram achieved a superior prognostic model, which may help to identify appropriate candidates for the therapy.

Assessing PD-L1 Expression Level via Preoperative MRI in HCC Based on Integrating Deep Learning and Radiomics Features

To assess if quantitative integrated deep learning and radiomics features can predict the PD-L1 expression level in preoperative MRI of hepatocellular carcinoma (HCC) patients. The data in this study consist of 103 hepatocellular carcinoma patients who received immunotherapy in a single center. These patients were divided into a high PD-L1 expression group (30 patients) and a low PD-L1 expression group (73 patients). Both radiomics and deep learning features were extracted from their MRI sequence of T2-WI, which were merged into an integrative feature space for machine learning for the prediction of PD-L1 expression. The five-fold cross-validation was adopted to validate the performance of the model, while the AUC was used to assess the predictive ability of the model. Based on the five-fold cross-validation, the integrated model achieved the best prediction performance, with an AUC score of 0.897 ± 0.084, followed by the deep learning-based model with an AUC of 0.852 ± 0.043 then the radiomics-based model with AUC of 0.794 ± 0.035. The feature set integrating radiomics and deep learning features is more effective in predicting PD-L1 expression level than only one feature type. The integrated model can achieve fast and accurate prediction of PD-L1 expression status in preoperative MRI of HCC patients.