Computed Tomography Radiomics-Based Prediction of Microvascular Invasion in Hepatocellular Carcinoma

Application of texture signatures based on multiparameter-magnetic resonance imaging for predicting microvascular invasion in hepatocellular carcinoma: Retrospective study

BACKGROUND

Despite continuous changes in treatment methods, the survival rate for advanced hepatocellular carcinoma (HCC) patients remains low, highlighting the importance of diagnostic methods for HCC.

AIM

To explore the efficacy of texture analysis based on multi-parametric magnetic resonance (MR) imaging (MRI) in predicting microvascular invasion (MVI) in preoperative HCC.

METHODS

This study included 105 patients with pathologically confirmed HCC, categorized into MVI-positive and MVI-negative groups. We employed Original Data Analysis, Principal Component Analysis, Linear Discriminant Analysis (LDA), and Non-LDA (NDA) for texture analysis using multi-parametric MR images to predict preoperative MVI. The effectiveness of texture analysis was determined using the B11 program of the MaZda4.6 software, with results expressed as the misjudgment rate (MCR).

RESULTS

Texture analysis using multi-parametric MRI, particularly the MI + PA + F dimensionality reduction method combined with NDA discrimination, demonstrated the most effective prediction of MVI in HCC. Prediction accuracy in the pulse and equilibrium phases was 83.81%. MCRs for the combination of T2-weighted imaging (T2WI), arterial phase, portal venous phase, and equilibrium phase were 22.86%, 16.19%, 20.95%, and 20.95%, respectively. The area under the curve for predicting MVI positivity was 0.844, with a sensitivity of 77.19% and specificity of 91.67%.

CONCLUSION

Texture analysis of arterial phase images demonstrated superior predictive efficacy for MVI in HCC compared to T2WI, portal venous, and equilibrium phases. This study provides an objective, non-invasive method for preoperative prediction of MVI, offering a theoretical foundation for the selection of clinical therapy.

[Prediction of microvascular invasion in hepatocellular carcinoma based on multi-phase dynamic enhanced CT radiomics feature and multi-classifier hierarchical fusion model]

OBJECTIVE

To predict microvascular invasion (MVI) in hepatocellular carcinoma (HCC) using a model based on multiphase dynamic-enhanced CT (DCE-CT) radiomics feature and hierarchical fusion of multiple classifiers.

METHODS

We retrospectively collected preoperative DCE-CT images from 111 patients with pathologically confirmed HCC in Guangzhou First People's Hospital between January, 2016 and April, 2020. The volume of interest was outlined in the early arterial phase, late arterial phase, portal venous phase and equilibrium phase, and radiomics features of these 4 phases were extracted. Seven classifiers based on different algorithms were trained using the filtered feature subsets to obtain multiple base classifiers under each phase. According to the hierarchical fusion strategy, a multi-criteria decision-making-based weight assignment algorithm was used for fusing each base classifier under the same phase with the model after extracting the phase information to obtain the prediction model. The proposed model was evaluated using a 5-fold cross-validation and assessed for area under the ROC curve (AUC), accuracy, sensitivity, and specificity. The prediction model was also compared with the fusion models using a single phase or multiple phases, models based on a single phase with a single classifier, models with different base classifier diversities, and 8 classifier models based on other ensemble methods.

RESULTS

The experimental results showed that the performance of the proposed model for predicting HCCMVI was optimal after incorporating the 4 phases and 7 classifiers, with AUC, accuracy, sensitivity, and specificity of 0.828, 0.766, 0.877, and 0.648, respectively. Comparative experiments showed that this prediction model outperformed the models based on a single phase with a single classifier and other ensemble models.

CONCLUSION

The proposed prediction model is effective for predicting MVI in HCC with superior performance to other models.

Prediction of preoperative microvascular invasion by dynamic radiomic analysis based on contrast-enhanced computed tomography

PURPOSE

Microvascular invasion (MVI) is a common complication of hepatocellular carcinoma (HCC) surgery, which is an important predictor of reduced surgical prognosis. This study aimed to develop a fully automated diagnostic model to predict pre-surgical MVI based on four-phase dynamic CT images.

METHODS

A total of 140 patients with HCC from two centers were retrospectively included (training set, n = 98; testing set, n = 42). All CT phases were aligned to the portal venous phase, and were then used to train a deep-learning model for liver tumor segmentation. Radiomics features were extracted from the tumor areas of original CT phases and pairwise subtraction images, as well as peritumoral features. Lastly, linear discriminant analysis (LDA) models were trained based on clinical features, radiomics features, and hybrid features, respectively. Models were evaluated by area under curve (AUC), accuracy, sensitivity, specificity, positive and negative predictive values (PPV and NPV).

RESULTS

Overall, 86 and 54 patients with MVI- (age, 55.92 ± 9.62 years; 68 men) and MVI+ (age, 53.59 ± 11.47 years; 43 men) were included. Average dice coefficients of liver tumor segmentation were 0.89 and 0.82 in training and testing sets, respectively. The model based on radiomics (AUC = 0.865, 95% CI: 0.725-0.951) showed slightly better performance than that based on clinical features (AUC = 0.841, 95% CI: 0.696-0.936). The classification model based on hybrid features achieved better performance in both training (AUC = 0.955, 95% CI: 0.893-0.987) and testing sets (AUC = 0.913, 95% CI: 0.785-0.978), compared with models based on clinical and radiomics features (p-value < 0.05). Moreover, the hybrid model also provided the best accuracy (0.857), sensitivity (0.875), and NPV (0.917).

CONCLUSION

The classification model based on multimodal intra- and peri-tumoral radiomics features can well predict HCC patients with MVI.

Dynamic contrast-enhanced ultrasonography with sonazoid predicts microvascular invasion in early-stage hepatocellular carcinoma

OBJECTIVE

Microvascular invasion (MVI) is an independent risk factor for the early recurrence and poor survival of hepatocellular carcinoma (HCC). This study aims to investigate the potential clinical value of dynamic contrast-enhanced ultrasound (DCE-ultrasound)-Sonazoid in pre-operatively assessing MVI in HCC.

METHODS AND MATERIALS

This single centre prospective study included 140 patients with histopathologically confirmed single HCC lesions. Patients were classified according to the post-operative pathological information presence of MVI: MVI+ group (n = 32) and MVI- group (n = 108). All patients underwent DCE-ultrasound within 1 week before surgery. The quantitative perfusion parameters of HCC lesions, margins of HCC lesions, and distal liver parenchyma were obtained and analyzed.

RESULTS

Clinicopathological (serum alpha-fetoprotein, Des-gamma-carboxyprothrombin, and pathological grade) and grayscale imaging features (tumor size) were significantly different between the MVI+ and MVI- groups (p < 0.05). Further quantitative analysis showed that when comparing the MVI+ and MVI- groups, half-decrease time and wash-out rate of HCC lesions and peak enhancement in the arterial phase of difference between the margin area of HCC and distal liver parenchyma were significantly different (p = 0.045, p = 0.035, and p = 0.023, respectively). Combining the above three quantitative parameters, the accuracy, sensitivity, specificity, positive-predictive value, and negative-predictive value were 69.3% (97/140), 37.8% (17/45), 84.3% (80/95), 53.1% (17/32), 74.1% (80/108), respectively.

CONCLUSION

DCE-ultrasound with quantitative perfusion analysis has the potential to predict MVI in HCC lesions.

ADVANCES IN KNOWLEDGE

DCE-ultrasound with quantitative perfusion analysis has the potential to predict MVI in HCC lesions.

Treatment Outcomes in Patients with Advanced Fibrolamellar Hepatocellular Carcinoma Under Systemic Treatment: Analysis of Clinical Characteristics, Management, and Radiomics

Purpose

Fibrolamellar hepatocellular carcinoma (FLHCC) is a rare primary liver malignancy often diagnosed at advanced stages. While there are limited data on the efficacy of specific agents, we aim to report outcomes of patients treated with systemic therapies and explore prognostic factors.

Patients and Methods

Medical records of patients treated between 2010 and 2022 were reviewed. Treatments were defined after multidisciplinary assessment. Descriptive statistics were used for baseline demographics. Time-to-event outcomes were estimated using the Kaplan-Meier method, compared by log-rank and adjusted by a regression model. Radiomic features (including size, shape, and texture) of the primary lesion were extracted and dimensionality reduced. An unsupervised Gaussian Mixture Model (GMM) clustering was performed, and survival was compared between clusters.

Results

We identified 23 patients: 12 males, with a median age of 23.6 years. At diagnosis, 82.6% had metastases, most frequently to the lungs (39.1%), lymph nodes (39.1%), and peritoneum (21.7%). Patients received a median of three lines (1-8) of treatment, including different regimens. Sorafenib (39.1%), capecitabine (30.4%), and capecitabine/interferon (13%) were the most used first-line regimens. The median time-to-failure was 3.8 months (95% CI: 3.2-8.7). Capecitabine + interferon (42.1%) and platinum combinations (39.1%) were the most used second-line regimens, with a time-to-failure of 3.5 months (95% CI: 1.5-11.6). Median overall survival was 26.7 months (95% CI: 15.1-40.4). A high baseline neutrophil-to-lymphocyte ratio (NLR) was associated with worse survival (p=0.02). Radiomic features identified three clusters, with one cluster (n=6) having better survival (40.4 vs 22.6 months, p=0.039). Tumor sphericity in the arterial phase was the most relevant characteristic associated with a better prognosis (accuracy=0.93).

Conclusion

FLHCC has unique features compared to conventional HCC, including young onset, gender balance, and absence of hepatopathy. Systemic therapies can provide encouraging survival, but lack of uniformity precludes defining a preferable regimen. Radiomics and NLR were suggested to correlate with prognosis and warrant further validation.

Construction and prognostic value of enhanced CT image omics model for noninvasive prediction of HRG in bladder cancer based on logistic regression and support vector machine algorithm

Background

Urothelial Carcinoma of the bladder (BLCA) is the most prevalent cancer of the urinary system. In cancer patients, HRG fusion is linked to a poor prognosis. The prediction of HRG expression by imaging omics in BLCA has not yet been fully investigated.

Methods

HRG expression in BLCA and healthy adjoining tissues was primarily identified utilizing data sourced from The Cancer Genome Atlas (TCGA). Using Kaplan-Meier survival curves and Landmark analysis, the relationship between HRG expression, clinicopathological parameters, and overall survival (OS) was investigated. Additionally, gene set variation analysis (GSVA) was conducted and CIBERSORTx was used to investigate the relationship between HRG expression and immune cell infiltration. The Cancer Imaging Archive (TCIA) provided CT images that were used for prognostic analysis, radiomic feature extraction, and construction of the model, respectively. The HRG expression levels were predicted using the constructed and evaluated LR and SMV models.

Results

HRG expression was shown to be substantially lower in BLCA tumors as opposed to that observed in normal tissues (p < 0.05). HRG expression had a close positive relationship with Eosinophils and a close negative relationship with B cells naive. The findings of the Landmark analysis illustrated that higher HRG was associated with improved patient survival at an early stage (P=0.048). The predictive performance of the two models, based on logistic regression analysis and support vector machine, was outstanding in the training and validation sets, yielding AUCs of 0.722 and 0.708, respectively, in the SVM model, and 0.727 and 0.662, respectively, in the LR.The models have good predictive efficiency.

Conclusion

HRG expression levels can have a significant impact on BLCA patients' prognoses. The radiomic characteristics can successfully predict the pre-surgical HRG expression levels, based on CT- Image omics.

New insights into a microvascular invasion prediction model in hepatocellular carcinoma: A retrospective study from the SEER database and China

Background and Aims

The prognosis of liver cancer is strongly influenced by microvascular infiltration (MVI). Accurate preoperative MVI prediction can aid clinicians in the selection of suitable treatment options. In this study, we constructed a novel, reliable, and adaptable nomogram for predicting MVI.

Methods

Using the Surveillance, Epidemiology, and End Results (SEER) database, we extracted the clinical data of 1,063 patients diagnosed with hepatocellular carcinoma (HCC) and divided it into either a training (n = 739) or an internal validation cohort (n = 326). Based on multivariate analysis, the training cohort data were analyzed and a nomogram was generated for MVI prediction. This was further verified using an internal validation cohort and an external validation cohort involving 293 Chinese patients. Furthermore, to evaluate the efficacy, accuracy, and clinical use of the nomogram, we used concordance index (C-index), calibration curve, and decision curve analysis (DCA) techniques.

Results

In accordance with the multivariate analysis, tumor size, tumor number, alpha-fetoprotein (AFP), and histological grade were independently associated with MVI. The established model exhibited satisfactory performance in predicting MVI. The C-indices were 0.719, 0.704, and 0.718 in the training, internal validation, and external validation cohorts, respectively. The calibration curves showed an excellent consistency between the predictions and actual observations. Finally, DCA demonstrated that the newly developed nomogram had favorable clinical utility.

Conclusions

We established and verified a novel preoperative MVI prediction model in HCC patients. This model can be a beneficial tool for clinicians in selecting an optimal treatment plan for HCC patients.

Preoperative and Prognostic Prediction of Microvascular Invasion in Hepatocellular Carcinoma: A Review Based on Artificial Intelligence

Microvascular invasion of hepatocellular carcinoma is an important factor affecting tumor recurrence after liver resection and liver transplantation. There are many ways to classify microvascular invasion, however, an international consensus is urgently needed. Recently, artificial intelligence has emerged as an important tool for improving the clinical management of hepatocellular carcinoma. Many studies about microvascular invasion currently focus on preoperative and prognosis prediction of microvascular invasion using artificial intelligence. In this paper, we review the definition and staging of microvascular invasion, especially the diagnosis of it by using artificial intelligence. In preoperative prediction, deep learning based on multimodal data modeling of radiomics-screened features, clinical features, and medical images is currently the most effective means. In prognostic prediction, pathology is the gold standard, and the techniques used should more effectively utilize the global features of the pathology images.

Clinical features and prognostic factors in patients with microvascular infiltration of hepatocellular carcinoma: Development and validation of a nomogram and risk stratification based on the SEER database

Background

The goal is to establish and validate an innovative prognostic risk stratification and nomogram in patients of hepatocellular carcinoma (HCC) with microvascular invasion (MVI) for predicting the cancer-specific survival (CSS).

Methods

1487 qualified patients were selected from the Surveillance, Epidemiology and End Results (SEER) database and randomly assigned to the training cohort and validation cohort in a ratio of 7:3. Concordance index (C-index), area under curve (AUC) and calibration plots were adopted to evaluate the discrimination and calibration of the nomogram. Decision curve analysis (DCA) was used to quantify the net benefit of the nomogram at different threshold probabilities and compare it to the American Joint Committee on Cancer (AJCC) tumor staging system. C-index, net reclassification index (NRI) and integrated discrimination improvement (IDI) were applied to evaluate the improvement of the new model over the AJCC tumor staging system. The new risk stratifications based on the nomogram and the AJCC tumor staging system were compared.

Results

Eight prognostic factors were used to construct the nomogram for HCC patients with MVI. The C-index for the training and validation cohorts was 0.785 and 0.776 respectively. The AUC values were higher than 0.7 both in the training cohort and validation cohort. The calibration plots showed good consistency between the actual observation and the nomogram prediction. The IDI values of 1-, 3-, 5-year CSS in the training cohort were 0.17, 0.16, 0.15, and in the validation cohort were 0.17, 0.17, 0.17 (P<0.05). The NRI values of the training cohort were 0.75 at 1-year, 0.68 at 3-year and 0.67 at 5-year. The DCA curves indicated that the new model more accurately predicted 1-year, 3-year, and 5-year CSS in both training and validation cohort, because it added more net benefit than the AJCC staging system. Furthermore, the risk stratification system showed the CSS in different groups had a good regional division.

Conclusions

A comprehensive risk stratification system and nomogram were established to forecast CSS for patients of HCC with MVI.