Contributing to the prediction of prognosis for treated hepatocellular carcinoma: Imaging aspects that sculpt the future

Abstract

A novel nomogram model to predict the prognosis of hepatocellular carcinoma (HCC) treated with radiofrequency ablation and transarterial chemoembolization was recently published in the World Journal of Gastrointestinal Surgery. This model includes clinical and laboratory factors, but emerging imaging aspects, particularly from magnetic resonance imaging (MRI) and radiomics, could enhance the predictive accuracy thereof. Multiparametric MRI and deep learning radiomics models significantly improve prognostic predictions for the treatment of HCC. Incorporating advanced imaging features, such as peritumoral hypointensity and radiomics scores, alongside clinical factors, can refine prognostic models, aiding in personalized treatment and better predicting outcomes. This letter underscores the importance of integrating novel imaging techniques into prognostic tools to better manage and treat HCC.

Key Words: Hepatocellular carcinoma; Radiofrequency ablation; Transcatheter arterial chemoembolization; Prognosis; Magnetic resonance imaging; Radiomics; Artificial intelligence

Core Tip: Emerging imaging techniques, particularly multiparametric magnetic resonance imaging and radiomics, enhance prognostic models for hepatocellular carcinoma (HCC) treated with radiofrequency ablation and transarterial chemoembolization. Incorporating advanced imaging features alongside clinical factors refines prediction accuracy, aiding personalized treatment and improving patient outcomes. This integration is crucial for advancing HCC prognosis and therapy.

TO THE EDITOR

I read with great interest the article titled, “Nomogram predicting the prognosis of primary liver cancer after radiofrequency ablation combined with transcatheter arterial chemoembolization,” recently published by Shen et al[1] in World Journal of Gastrointestinal Surgery. This article develops a novel nomogram model to predict the prognosis of patients with hepatocellular carcinoma (HCC) treated with radiofrequency ablation (RFA) and transarterial chemoembolization (TACE), which is based on factors such as the presence of portal vein tumor thrombosis, vascular invasion, liver cirrhosis, as well as levels of alpha-fetoprotein, α-L-fucosidase, and the prognostic nutritional index.

I commend the authors for this interesting report, which provides a new tool to improve prognosis prediction for patients with HCC receiving RFA and TACE. However, I intend to contribute to the article by highlighting the importance of emergent imaging aspects that could enhance the diagnostic features of the nomogram prediction model, finely sculpting the ability to predict the future prognosis of treated HCC.

At present, HCC is considered the most common type of primary liver cancer and the third leading cause of cancer-related death worldwide[2]. Its complex multimodality treatment depends on the patient’s intrinsic liver function, tumor burden, and treatment intention, ranging widely from locoregional therapies to liver transplantation[3].

TACE and RFA are two widely used locoregional treatments for HCC due to their high efficacy in local disease control and low adverse effects[4,5]. However, the increased rate of tumor recurrence after treatment and the lack of non-invasive features that allow us to accurately predict the future prognosis for each patient remains an unsolved challenge[6-8].

According to Shen et al[1], several clinical and laboratory factors are independent prognostic factors for patients with HCC treated with RFA and TACE. However, compelling evidence has shown the development of novel imaging-based features that could play a crucial role in predicting the prognosis in patients with treated HCC, since they may improve the diagnostic accuracy of the predictive model[9,10].

In recent decades, multimodality imaging has played a central role in evaluating therapeutic responses to HCC. In particular, magnetic resonance imaging (MRI) has demonstrated promising advances due to its high tumor-to-liver contrast and good depiction of intrahepatic vascular and biliary structures[11,12]. Additionally, the emergence of artificial intelligence and the growing accumulation of liver MRI data have made radiomics analyses increasingly precise in evaluating tumor prognosis after ablative therapies[13,14].

A pioneering study has developed a novel deep learning radiomics model based on multiparametric MRI (DLRMM), which demonstrated the ability to predict tumor progression (LTP) after thermal ablation in patients with HCC[15]. The DLRMM operates by extracting high-dimensional features from multiparametric MRI sequences, including T1-weighted, T2-weighted, diffusion-weighted, and contrast-enhanced images[16,17].

These features are then fed into a convolutional neural network architecture that has been specifically trained on large datasets of HCC images. The network learns to identify subtle imaging biomarkers that correlate with tumor behavior and treatment response[18]. Implementing this model in clinical practice would require the integration of these MRI sequences into a standardized workflow, coupled with high-performance computational resources to handle the model’s processing demands[19].

Similarly, Kang et al[20] proposed a prediction model based on MRI features before RFA, such as peritumoral hypointensity on hepatobiliary phase images, tumor margin, and tumor size. This model could robustly identify HCC patients at high risk for LTP after treatment.

Another study exploring the feasibility of MRI radiomics to predict HCC recurrence after RFA proposed an interesting nomogram that combines radiomics scores with clinical features to individualize the prediction of tumor recurrence. Notably, this radiomics-based nomogram shows high diagnostic accuracy for predicting early recurrence in HCC after RFA[21-23].

Several authors have developed comprehensive models that incorporate radiomic features and clinical factors to predict treatment responses and survival in HCC patients after TACE. In this line, a recent study conducted by Shi et al[24], investigated the value of integrating contrast-enhanced computed tomography (CT) information with clinical factors to predict treatment response and long-term outcomes in HCC patients receiving TACE as a first-line treatment, elaborating upon a radiomics-based prognostic nomogram for all-stage HCC patients after TACE, successfully predicting survival and treatment response.

Notably, these findings were consistent with a model proposed by Sun et al[25], which used clinical factors and quantitative radiomic features from contrast-enhanced CT to develop a highly accurate model to predicting responses to TACE, which can be utilized as a potent tool for precision treatment. A recently published systematic review and meta-analysis demonstrated that the combined radiomics-clinical model provided the best performance in predicting a therapeutic response and survival of HCC patients treated with TACE[26]. In this sense, the growing use of radiomics as a complement to clinical elements in evaluating HCC patients treated with TACE has enabled significant advances in personalizing treatment and predicting long-term prognosis post-treatment. It has also laid the foundation for short-term preoperative prediction of a therapy response based on intratumoral and peritumoral tissue characteristics[27-29].

Taken together, all these data highlight the essential contribution of imaging aspects in HCC after treatment. Despite these advances, future research should focus on addressing several key challenges that remain, including the need for large, multicentric datasets to train and validate these models, which can improve their robustness and generalizability across different patient populations and imaging protocols, as well as integrating these imaging models with genomic and proteomic data to develop even more precise and individualized predictive tools. This compels us to continue to investigate the numerous future opportunities that arise from combining clinical aspects with advanced imaging to refine the future of patient prognostic prediction models.