Deep-learning-based analysis of preoperative MRI predicts microvascular invasion and outcome in hepatocellular carcinoma

Preoperative Contrast-Enhanced Ultrasound Predicts Microvascular Invasion in Hepatocellular Carcinoma as Accurately as Contrast-Enhanced MR

OBJECTIVES

Both contrast-enhanced ultrasound (CEUS) and contrast-enhanced magnetic resonance (CEMR) are important imaging methods for hepatocellular carcinoma (HCC). This study aimed to establish a model using preoperative CEUS parameters to predict microvascular invasion (MVI) in HCC, and compare its predictive efficiency with that of CEMR model.

METHODS

A total of 93 patients with HCC (39 cases in MVI positive group and 54 cases in MVI negative group) who underwent surgery in our hospital from January 2020 to June 2021 were retrospectively analyzed. Their clinical and imaging data were collected to establish CEUS and CEMR models for predicting MVI. The predictive efficiencies of both models were compared.

RESULTS

By the univariate and multivariate regression analyses of patients' clinical information, preoperative CEUS static and dynamic images, we found that serrated edge and time to peak were independent predictors of MVI. The CEUS prediction model achieved a sensitivity of 92.3%, a specificity of 83.3%, and an accuracy of 84.6% (Az: 0.934). By analyzing the clinical and CEMR information, we found that tumor morphology, fast-in and fast-out, peritumoral enhancement, and capsule were independent predictors of MVI. The CEMR prediction model achieved a sensitivity of 97.4%, a specificity of 77.8%, and an accuracy of 83.2% (Az: 0.900). The combination of the two models achieved a sensitivity of 84.6%, a specificity of 87.0%, and an accuracy of 86.2% (Az: 0.884). There was no significant statistical difference in the areas under the ROC curve of the three models.

CONCLUSION

The CEUS model and the CEMR model have similar predictive efficiencies for MVI of HCC. CEUS is also an effective method to predict MVI before operation.

Extracellular contrast agent-enhanced MRI is as effective as gadoxetate disodium-enhanced MRI for predicting microvascular invasion in HCC

PURPOSE

To compare the performances of gadoxetate disodium-enhanced MRI (EOB-MRI) and extracellular contrast agent-enhanced MRI (ECA-MRI) for predicting microvascular invasion (MVI) in HCC.

MATERIALS AND METHODS

From November 2009 to December 2021, consecutive HCC patients who underwent preoperative contrast-enhanced MRI were retrospectively enrolled into either an ECA-MRI or EOB-MRI cohort. In the ECA-MRI cohort, a preoperative MVI score was constructed in the training dataset using a logistic regression model that evaluated pathological type. In a propensity score-matched testing dataset of the ECA-MRI cohort, the MVI score was validated and compared with a previously proposed EOB-MRI-based MVI score calculated in the EOB-MRI cohort. Time-to-early recurrence survival was evaluated by the Kaplan-Meier method with the log-rank test.

RESULTS

A total of 536 patients were included (478 men; 53 years, interquartile range, 46-62 years), 322 (60.1 %) with pathologically confirmed MVI. Based on the training dataset, independent variables associated with MVI included serum alpha-fetoprotein > 400 ng/ml (odds ratio [OR] = 2.3), infiltrative appearance (OR = 4.9), internal artery (OR = 2.5) and nodule-in-nodule architecture (OR = 2.4), which were incorporated into the ECA-MRI-based MVI score. The testing dataset AUC of the ECA-MRI score was 0.720, which was comparable to that of the EOB-MRI-based MVI score (AUC = 0.721; P =.99). Patients from either the ECA-MRI or the EOB-MRI cohort with model-predicted MVI had significantly shorter time-to-early recurrence than those without MVI (P <.001).

CONCLUSION

Based on the preoperative serum alpha-fetoprotein and three MRI features, ECA-MRI demonstrated comparable performance to EOB-MRI for predicting MVI in HCC.

Clinical application of dual-layer spectral CT multi-parameter feature to predict microvascular invasion in hepatocellular carcinoma

OBJECTIVE

This study aimed to investigate the feasibility of using dual-layer spectral CT multi-parameter feature to predict microvascular invasion of hepatocellular carcinoma.

METHODS

This retrospective study enrolled 50 HCC patients who underwent multiphase contrast-enhanced spectral CT studies preoperatively. Combined clinical data, radiological features with spectral CT quantitative parameter were constructed to predict MVI. ROC was applied to identify potential predictors of MVI. The CT values obtained by simulating the conventional CT scans with 70 keV images were compared with those obtained with 40 keV images.

RESULTS

50 hepatocellular carcinomas were detected with 30 lesions (Group A) with microvascular invasion and 20 (Group B) without. There were significant differences in AFP,tumer size, IC, NIC,slope and effective atomic number in AP and ICrr in VP between Group A ((1000(10.875,1000),4.360±0.3105, 1.7750 (1.5350,1.8825) mg/ml, 0.1785 (0.1621,0.2124), 2.0362±0.2108,8.0960±0.1043,0.2830±0.0777) and Group B (4.750(3.325,20.425),3.190±0.2979,1.4700 (1.4500,1.5775) mg/ml, 0.1441 (0.1373,0.1490),1.8601±0.1595, 7.8105±0.7830 and 0.2228±0.0612) (all p < 0.05). Using 0.1586 as the threshold for NIC, one could obtain an area-under-curve (AUC) of 0.875 in ROC to differentiate between tumours with and without microvascular invasion. AUC was 0.625 with CT value at 70 keV and improved to 0.843 at 40 keV.

CONCLUSION

Dual-layer spectral CT provides additional quantitative parameters than conventional CT to enhance the differentiation between hepatocellular carcinoma with and without microvascular invasion. Especially, the normalized iodine concentration (NIC) in arterial phase has the greatest potential application value in determining whether microvascular invasion exists, and can offer an important reference for clinical treatment plan and prognosis assessment.

Prediction of Microvascular Invasion in Solitary AFP-Negative Hepatocellular Carcinoma ≤ 5 cm Using a Combination of Imaging Features and Quantitative Dual-Layer Spectral-Detector CT Parameters

RATIONALE AND OBJECTIVES

AFP-negative hepatocellular carcinoma (AFPN-HCC) within 5 cm is a special subgroup of HCC. This study aimed to investigate the value of dual-layer spectral-detector CT (DLCT) and construct a scoring model based on imaging features as well as DLCT for predicting microvascular invasion (MVI) in AFPN-HCC within 5 cm.

METHODS

This retrospective study enrolled 104 HCC patients who underwent multiphase contrast-enhanced DLCT studies preoperatively. Combined radiological features (CR) and combined DLCT quantitative parameter (CDLCT) were constructed to predict MVI. Multivariable logistic regression was applied to identify potential predictors of MVI. Based on the coefficient of the regression model, a scoring model was developed. The predictive efficacy was assessed through ROC analysis.

RESULTS

Microvascular invasion (MVI) was found in 28 (26.9%) AFPN-HCC patients. Among single parameters, the effective atomic number in arterial phase demonstrated the best predictive efficiency for MVI with an area under the curve (AUC) of 0.792. CR and CDLCT showed predictive performance with AUCs of 0.848 and 0.849, respectively. A risk score (RS) was calculated using the independent predictors of MVI as follows: RS = 2 × (mosaic architecture) + 2 × (corona enhancement) + 2 × (incomplete tumor capsule) + 2 × (2-trait predictor of venous invasion [TTPVI]) + 3 × (CDLCT > -1.229). Delong's test demonstrated this scoring system could significantly improve the AUC to 0.929 compared with CR (p = 0.016) and CDLCT (p = 0.034).

CONCLUSION

The scoring model combining radiological features with DLCT provides a promising tool for predicting MVI in solitary AFPN-HCC within 5 cm preoperatively.

Development and validation of a nomogram to predict the recurrence of hepatocellular carcinoma patients with dynamic changes in AFP undergoing locoregional treatments

Background

Serum alpha-fetoprotein (AFP) is an important clinical indicator for screening, diagnosis, and prognosis of primary hepatocellular carcinoma (HCC). Our team's previous study showed that patients with negative AFP at baseline and positive AFP at relapse had a worse prognosis (N-P). Therefore, the aim of our study was to develop and validate a nomogram for this group of patients.

Methods

A total of 513 patients with HCC who received locoregional treatments at Beijing You'an Hospital, Capital Medical University, from January 2012 to December 2019 were prospectively enrolled. Patients admitted from 2012 to 2015 were assigned to the training cohort (n = 335), while 2016 to 2019 were in the validation cohort (n =183). The clinical and pathological features of patients were collected, and independent risk factors were identified using univariate and multivariate Cox regression analysis as a basis for developing a nomogram. The performance of the nomogram was evaluated by C-index, receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) curves in the training and validation cohorts.

Results

The content of the nomogram includes gender, tumor number, tumor size, lymphocyte, direct bilirubin (DBIL), gamma-glutamyl transferase (GGT), and prealbumin. The C-index (0.717 and 0.752) and 1-, 3-, and 5-year AUCs (0.721, 0.825, 0.845, and 0.740, 0.868, 0.837) of the training and validation cohorts proved the good predictive performance of the nomogram. Calibration curves and DCA curves suggested accuracy and net clinical benefit rates. The nomogram enabled to classify of patients with dynamic changes in AFP into three groups according to the risk of recurrence: low risk, intermediate risk, and high risk. There was a statistically significant difference in RFS between the three groups in the training and validation cohorts (P<0.001).

Conclusion

The nomogram developed and validated in this study had good predictive power for patients with dynamic changes in AFP.

Liver Transplant in Patients with Hepatocarcinoma: Imaging Guidelines and Future Perspectives Using Artificial Intelligence

Hepatocellular carcinoma is the most common primary malignant hepatic tumor and occurs most often in the setting of chronic liver disease. Liver transplantation is a curative treatment option and is an ideal solution because it solves the chronic underlying liver disorder while removing the malignant lesion. However, due to organ shortages, this treatment can only be applied to carefully selected patients according to clinical guidelines. Artificial intelligence is an emerging technology with multiple applications in medicine with a predilection for domains that work with medical imaging, like radiology. With the help of these technologies, laborious tasks can be automated, and new lesion imaging criteria can be developed based on pixel-level analysis. Our objectives are to review the developing AI applications that could be implemented to better stratify liver transplant candidates. The papers analysed applied AI for liver segmentation, evaluation of steatosis, sarcopenia assessment, lesion detection, segmentation, and characterization. A liver transplant is an optimal treatment for patients with hepatocellular carcinoma in the setting of chronic liver disease. Furthermore, AI could provide solutions for improving the management of liver transplant candidates to improve survival.

Radiomics: A radiological evidence-based artificial intelligence technique to facilitate personalized precision medicine in hepatocellular carcinoma

The high postoperative recurrence rates in hepatocellular carcinoma (HCC) remain a major hurdle in its management. Appropriate staging and treatment selection may alleviate the extent of fatal recurrence. However, effective methods to preoperatively evaluate pathophysiologic and molecular characteristics of HCC are lacking. Imaging plays a central role in HCC diagnosis and stratification due to the non-invasive diagnostic criteria. Vast and crucial information is hidden within image data. Other than providing a morphological sketch for lesion diagnosis, imaging could provide new insights to describe the pathophysiological and genetic landscape of HCC. Radiomics aims to facilitate diagnosis and prognosis of HCC using artificial intelligence techniques to harness the immense information contained in medical images. Radiomics produces a set of archetypal and robust imaging features that are correlated to key pathological or molecular biomarkers to preoperatively risk-stratify HCC patients. Inferred with outcome data, comprehensive combination of radiomic, clinical and/or multi-omics data could also improve direct prediction of response to treatment and prognosis. The evolution of radiomics is changing our understanding of personalized precision medicine in HCC management. Herein, we review the key techniques and clinical applications in HCC radiomics and discuss current limitations and future opportunities to improve clinical decision making.

Lenvatinib plus anti-PD-1 therapy represents a feasible conversion resection strategy for patients with initially unresectable hepatocellular carcinoma: A retrospective study

Purpose

We aimed to investigate the feasibility of lenvatinib plus anti-PD-1 therapy as a conversion therapy for initially unresectable hepatocellular carcinoma (HCC).

Methods

Patients with initially unresectable HCC who received combined lenvatinib and anti-PD-1 antibody between May 2020 and Jan 2022 in Zhongshan Hospital were retrospectively analyzed. Tumor response and resectability were assessed by imaging every two months according to RECIST version 1.1 and modified RECIST (mRECIST) criteria.

Results

A total of 107 patients were enrolled. 30 (28%) of them received conversion surgery within 90.5 (range: 53-456) days after the initiation of lenvatinib plus anti-PD-1 therapy. At baseline, the median largest tumor diameter of these 30 patients was 9.2 cm (range: 3.5-15.0 cm), 26 patients had Barcelona Clinic Liver Cancer stage B-C, and 4 had stage A. Prior to surgery, all cases displayed tumor regression and 15 patients achieved objective response. Pathological complete response (pCR) was observed in 10 patients. No severe drug-related adverse events or surgical complications were observed. After a median follow-up of 16.5 months, 28 patients survived and 11 developed tumor recurrence. Survival analysis showed patients achieving tumor response before surgery or pCR had a longer tumor-free survival. Notably, patients with microvascular invasion (MVI) had significantly higher recurrence rate and poorer overall survival than patients without.

Conclusions

Lenvatinib combined with anti-PD-1 therapy represents a feasible conversion strategy for patients with initially unresectable HCC. Patients achieving tumor responses are more likely to benefit from conversion resection to access a longer term of tumor-free survival.