Prediction of Tumor Progression After Microwave Ablation of Hepatocellular Carcinoma by Contrast-Enhanced Ultrasound Combined with Immunohistochemical Markers.

Introduction
Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related deaths worldwide and its incidence continues to increase.1 Thermal ablation therapy has been widely used in the local treatment of HCC and is included in the local treatment guidelines for liver cancer,2 with satisfactory results for its treatment efficiency. For patients with HCC treated with thermal ablation, survival rate and tumor recurrence are the leading indicators for evaluating the therapeutic effect after the treatment.3 Microwave ablation (MWA) is an efficient and safe thermal ablation technology that has been used to treat HCC for more than 20 years4 and can accurately destroy the target tissue with little impact on the surrounding tissues. Clinical studies and animal experiments have shown that MWA not only inactivates tumor cells but also further activates the immune response and enhances the killing of tumor cells.5 Compared with surgical resection (SR), patients with HCC have a better long-term survival rate after MWA but also have a higher rate of tumor recurrence.6–8
At present, common factors associated with recurrence after thermal ablation in HCC are tumor size, tumor number, tumor markers, high albumin-bilirubin (ALBI) grade histologic grade, and micro/macrovascular invasion.9–11 Factors that are particularly relevant to prognosis after MWA are the liver disease stage and ablation necrosis area.12
Therefore, relying solely on imaging modalities for predicting tumor progression (TP) has inherent limitations. Although radiological evaluations effectively characterize morphological features, they inadequately capture the biological aggressiveness of the tumor. Furthermore, micrometastatic residual tumors at the ablation margins and early recurrent foci may be detected by conventional imaging protocols. Notably, serological biomarkers reflecting tumor biological behavior, particularly alpha-fetoprotein (AFP) levels, have a critical prognostic value in hepatocellular carcinoma (HCC). Our team’s previous studies analyzed the factors affecting the prognosis of HCC and showed that the preoperative tumor marker AFP level is an important factor affecting the ablation effect and TP. Protein induced by vitamin K absence or antagonist II (PIVKA-II) is also a common serological marker for HCC, and elevated levels often indicate high malignancy and poor prognosis.13 A significant decline in PIVKA-II levels is associated with good treatment outcomes, whereas a persistent rise or lack of a significant decline suggests the possible presence of residual tumors or early recurrence.14 In addition, indicators of liver function, such as albumin (ALB) and total bilirubin (TBIL), and blood routine indicators, such as platelet counts (PLT) and age-platelet index (API), also reflect liver reserve function and tumor microenvironment, which are associated with HCC prognosis.15
Immunohistochemistry (IHC) results are important for the diagnosis, treatment, and prognosis of tumors. Tumor staging can be determined according to HCC IHC indexes,16 which can be found in a timely and accurate manner, and the diagnosis rate can be significantly improved to provide appropriate treatment for HCC patients. IHC can not only distinguish the pathological subtypes of tumors17 and provide targets for targeted therapy18 but also predict and judge the prognosis of tumors after treatment.19
Recent studies have shown that some immune markers can be used to predict the early recurrence of HCC after treatment, including markers of hepatocyte differentiation, such as hepatocyte paraffin 1 (HepPar-1) and glypican-3 (GPC-3),20,21 and markers of malignant hepatocytes, such as arginase-1 (Arg-1) and CD34.
HepPar-1 and GPC-3 are routine pathological diagnostic biomarkers for HCC. HepPar-1 is a monoclonal antibody with a sensitivity > 70% and high specificity for HCC diagnosis.22 This antibody is limited by its low sensitivity for the diagnosis of poorly differentiated HCC and scirrhous HCC.23
GPC-3 is a membrane-associated proteoglycan that is specifically upregulated in HCC but is rarely or not expressed in normal liver tissues, and it has almost similar sensitivity to AFP in the diagnosis of HCC, while its specificity is lower than that of AFP.24 Studies have shown that GPC-3 plays a crucial role in HCC cell proliferation and metastasis. It mediates tumorigenesis via signaling pathways in the process of malignant transformation of hepatocytes, and its expression is increased in dysplastic and cancerous tissues.25
Arg-1 is highly expressed in HCC and is highly sensitive and specific, even in poorly differentiated HCC,23,26 which also has higher sensitivity and specificity than HepPar-1 for HCC diagnosis. Furthermore, the combined use of Arg-1 and HepPar-1 can more efficiently distinguish HCC from metastatic carcinoma and intrahepatic cholangiocarcinoma.27,28
The sensitivity of AFP for the diagnosis of HCC is not very high, but it is a useful indicator for differential diagnosis. Serum AFP levels is a known independent risk factor for recurrence after thermal ablation,9 and which has been used as a tumor marker to predict poor prognosis and recurrence of HCC.
Heat shock protein 70 (HSP 70) may be an important gene in the regulation of tumor cell proliferation and apoptosis and can be used as an important index for the auxiliary diagnosis and prognosis of liver cancer.
The endothelial cell marker CD34 is a specific marker for vascular tumors. CD34 was positively expressed in HCC as a diffuse, long, or branched microvessel pattern, whereas it was negative in normal hepatic sinusoids, which may help to confirm the invasion previously observed on routine histology.29 Among the numerous diagnostic markers studied, HepPar-1 and CD34 have been found to be valuable in distinguishing HCC from metastatic neoplasms of extrahepatic sites.30
Although the prognostic significance of HepPar-1, GPC-3, Arg-1, AFP, HSP 70 and CD34 in patients with HCC has been emphasized in previous studies, it is also vital to investigate recurrence and related factors to choose the correct IHC markers, establish postoperative management, and predict prognosis based on HCC IHC markers. At present, there are few studies on whether immune markers of HCC are related to MWA prognosis. Therefore, the combination of contrast-enhanced ultrasound(CEUS) features with laboratory tests (such as tumor markers, liver function indicators, etc.) can effectively improve the accuracy of the prediction model. Based on this, this study aims to construct a prediction model by integrating CEUS characteristics, laboratory markers, and immune markers, providing new ideas and methods for individualized treatment and prognosis assessment of HCC patients. Based on this, the purpose of this study was to explore the correlation between HCC IHC markers and prognosis after MWA to improve the survival rate of patients with HCC. We summarized the MWA data of 91 patients with HCC and analyzed the clinical significance of HepPar-1, GPC-3,Arg-1, and other markers in predicting postoperative HCC recurrence.

Materials and Methods

Patients
A total of 91 patients with single HCC ≤5cm who underwent MWA in our institution from January 2016 to March 2024 were retrospectively collected, and their biopsy tissue samples were obtained before MWA and stained by IHC.
The inclusion criteria were as follows: (1) age ≥18 years; (2) tumor clinical stage of BCLC stage 0 or A; (3)Child-Pugh grade A or B; (4) diameter ≤5 cm; (5) no history of other malignancies; (6) confirmed by puncture pathology and immunohistochemical staining; (7) the target lesion did not receive other anti-cancer therapies, such as RFA, TACE, and targeted drugs; and (8) MWA was technically successful after treatment.
The exclusion criteria were as follows: (1) age < 18 years; (2) tumor clinical stage not BCLC stage 0 or A; (3)Child-Pugh grade C; (4) no pathology or IHC; and (5) vascular infiltration or metastasis.
According to the selection criteria (Figure 1), cases with more than three nodules, lesion diameter > 5 cm, non-HCC, vascular invasion, extrahepatic metastasis, no pathological or immunohistochemical results, and other treatment options were excluded, and the remaining 91 patients were included in the study for prognostic prediction. The retrospective study protocol was approved by the Ethics Committee Board (2022A-180) and all participants were conducted in accordance with the World Medical Association Declaration of Helsinki. All patients signed the informed consent form before previous CEUS and MWA. Due to the retrospective nature of this study, informed consent for the study was waived.

US and MWA Techniques
B-mode ultrasound and contrast-enhanced ultrasound(CEUS) with the contrast agent of SonoVue (Bracco Imaging, Milan, Italy) were examined in all patients using a Philips EPIQ 7 and ACUSON Redwood ultrasound diagnostic apparatus with convex array transducer C5-1 (1–5 MHz). Based on the 2D US findings, tumors are classified into high-risk and low-risk location according to their proximity to major blood vessels (with an intravascular diameter>5 mm) and adjacent structures (diaphragm, surrounding organs like the gallbladder and gastrointestinal tract). High-risk location are defined as those with distances between tumors and these anatomical structures <5 mm.
All patients were treated with MWA (a water-cooled ablation system, Canyon Medical, Nanjing, China) under ultrasound-guided intravenous conscious analgesia sedation by the same senior doctor with rich clinical experience, and the power and time of ablation were set based on the tumor diameter and location according to the manufacturer’s instructions. The microwave output power is set to 55–65 W for 5–10 min at a time, which can be adjusted according to the tumor location and/or the patient’s pain tolerance during surgery. The operation should be terminated when the ablative gas echo covers the tumor for at least 10 mm in each section.

CT and MRI Images
All patients underwent CEUS within 2 week before or after MWA for assessment of the therapeutic effect. Contrast-enhanced CT(CECT) or CEMRI was performed when suspicious enhanced peripheral nodule was present in the arterial phase of CEUS. Irregular nodular enhancement, which suggested the presence of a residual unablated lesion, reablation was then considered if the patient still met the criteria for MWA. Technical success was considered to have been achieved. When the ablation zone completely covered the tumor and if there was no irregular enhancement at the ablation margin. CEUS or CECT/CEMRI of the abdomen was performed 1, 2, and 3 months after MWA.

Immunohistochemistry
HCC paraffin sections were deparaffinized in xylene and then hydrated with alcohol of decreasing concentration followed by rinsing with PBS for 5–10 minutes. Antigen repair was performed, and the cells were treated with citrate buffer (pH 6.0) for 10 min in a microwave oven and cooled for 10 min before IHC staining. Endogenous peroxidase was then inactivated and the slides were incubated in 3% hydrogen peroxide for 10 min and washed in buffer solution. The primary antibody was diluted with PBS, and the tissue was incubated with the diluted primary antibody for 1 h at room temperature to allow adequate binding of the primary antibody to the antigen. The diluted secondary antibody was dropped onto tissue sections that were incubated with the primary antibody, washed, and incubated for 30 min at room temperature. The antibody response was then detected using an avidin-biotin detection kit with diaminobenzidine (DAB) as the chromophore for 5 min and DAKO auto hematoxylin as the counterstain for 15 min. The slices were sealed for observation. IHC markers such as HepPar-1, GPC-3, Arg-1, AFP, HSP 70 and CD34 were detected using the corresponding antibodies. Primary antibodies from Zeta (Sierra Madre, USA) were diluted as follows: mouse monoclonal antibody against HepPar-1 (OCH1E5 1:200 dilution, cytoplasmic staining), mouse monoclonal antibody against Glypican-3 (1G12, 1:200 dilution, cytoplasmic staining), mouse monoclonal antibody against Arg-1 (EP261, 1:100 dilution, cytoplasmic staining), mouse monoclonal antibody against AFP (EP209, 1:100 dilution, cytoplasmic staining), mouse monoclonal antibody against CD34 (QBEnd/10, 1:200 dilution, cytoplasmic staining), and mouse monoclonal antibody against HSP 70 (W27, 1:200 dilution, nuclear staining). IHC analysis was performed using a 2-step EnVision procedure. Appropriate positive and negative controls were used. Staining results were assessed as negative (<5% of tumor cells stained) or positive. Immunostained slides were independently evaluated by two experienced pathologists.

Laboratory Indicators
The preoperative and postoperative laboratory parameters of the patients, including ALB, TBIL, PLT, AFP, and PIVKA-II levels, were recorded.
Age-platelet index (API) = age + platelet count. Age (years): <30=0; 30~39=1 point; 40~49=2 points; 50~59=3 points; 60~69=4 points; ≥70=5 points. Platelet technology count (109/L): ≥225=0 points; 200~224=1 score; 175~199=2 points; 150~174=3 points; 125~149=4 points; <125=5 points. API scores ranged from 0 to 10 points.

Follow Up Protocol
All patients were regularly followed up. Liver function, serum tumor markers, and abdominal CEUS were performed at 1, 2, and 3 months after MWA, and then every 6 months until recurrence or withdrawal, up to 48 months. When metastasis or recurrence is suspected on CEUS or follow-up with tumor markers, further evaluation can be performed using CECT or CEMRI, and if necessary, ultrasound-guided biopsy can be used to confirm the diagnosis.

Quality Control
To ensure model accuracy, multiple quality control measures were implemented, including standardized contrast agent injection protocols, unified image acquisition parameters, standardized ablation technique operations, standardized 3D CEUS model construction, standardized image analysis criteria, and double-blind measurement methods. The model’s accuracy was validated through comparison of postoperative pathological results with follow-up imaging examinations.

Statistical Analysis
SPSS 26.0 statistical software was used for analysis. Measurement data are expressed as the mean ± standard deviation, and categorical variables are expressed as the number of cases and percentage (n(%)). T-tests were used for continuous variables. For categorical variables, the chi-square test or Fisher’s exact test was used. Statistically significant variables obtained from univariate analysis were included in the multivariate logistic regression (forward, step-by-step analysis) analysis, and ROC analysis was performed based on the optimal variables in the multivariate logistic regression analysis to evaluate the performance of the model. Statistical significance was set at p < 0.05.

Results

Analysis of Correlation Factors Between TP and Baseline Characteristics
A total of 91 patients with single HCC with ultrasound-guided percutaneous MWA were included in this study, including 24 females and 67 males, with an average age of 56.84±10.48 years old. The tumor diameter ranged from 0.5 cm to 5 cm (2.34±0.84 cm), of which 68 cases were ≤3 cm and 23 cases ranged from 3 cm to 5 cm. Of the total group of patients, 20 (21.98%) patients received MWA along with systemic targeting/immunotherapy, and 69 (75.82%) patients received no other therapy. Nineteen patients (20.88%) had previously undergone surgical resection of HCC, and ablation was performed on new lesions located far from the previous surgical area. The patients were followed up for 18 months (12–54 months), and TP was present in 12 patients (13.2%), including 5 cases of LTP and 7 cases of new intrahepatic lesions. Patients were divided into progressive and non-progressive groups according to the presence of TP. The baseline data of the two groups were compared and analyzed, as shown in Table 1. In this group, 87 patients had a hepatitis background, including 80 patients with hepatitis B and 7 patients with hepatitis C. A statistically significant difference was observed between hepatitis and TP after ablation (p < 0.001). Cirrhosis background, BCLC stage, previous surgical history and systemic treatment during ablation had no significant effects on TP after ablation, with no statistical significance (all p > 0.05).

Analysis of Correlation Factors Between TP and Imaging Features
The comparative analysis of the basic characteristics and imaging features of tumors in the two groups is shown in Table 2. Tumor size, high-risk locationhigh-risk location, ill-defined margins, irregular morphology and contrast washout time were significantly correlated with whether HCC recurred after ablation (all p < 0.05). Larger tumor diameter, poorly defined margins, irregular morphology, and earlier contrast washout were positively associated with TP after ablation. However, there were no significant differences between tumor echogenicity, histological grade, intralesional vascularity, other parameters of contrast ultrasound and CECT/CEMRI related indexes with TP after treatment in this group (all p > 0.05).

Analysis of Correlation Factors Between TP and Laboratory Indicators
Univariate analysis of liver function indexes and TP after ablation showed that ALB, TBIL and PLT had no statistical difference for TP after ablation, while API in the two groups was significantly correlated with the occurrence of TP after ablation (p = 0.014). Univariate analysis of tumor serological markers and TP after ablation in the two groups showed that abnormal increases in AFP and PIVKA-II before and after tumor ablation were significantly correlated with TP after HCC ablation (both p < 0.05), as shown in Table 3.

Analysis of Correlation Factors Between TP and Tumor IHC Makers
The univariate analysis of tumor immunohistochemical indexes and TP after ablation in the two groups is shown in Table 4 and (Figures 2 and 3). The positive expression of GPC-3, Arg-1, HSP 70 and higher Ki-67 index were significantly correlated with the progression of HCC after ablation (all p < 0.05).

Logistic Regression Analysis of Factors Related to Tumor Progression
Univariate logistic regression analysis showed: TP was significantly correlated with tumor size, high-risk location, boundary, morphology, contrast washout time, serum AFP levels (> 200 ng/mL) before surgery, continuous increase of AFP and PIVKA-II levels after surgery, positive expression of GPC-3, Arg-1, HSP 70 and Ki-67 index. Multivariate logistic regression analysis showed that tumor morphology, positive expression of serum AFP levels (> 200 ng/mL) before surgery, GPC-3 and HSP 70 were independent predictors of tumor recurrence after ablation (Table 5).

ROC Curve Evaluation Model for Predicting TP
The predictive performance of the model was evaluated using the ROC curve (Figure 4). In predicting outcomes following HCC ablation, the model incorporating GPC-3 expression, tumor imaging morphology, and pre-MWA serum AFP levels demonstrated near-perfect discriminatory performance (AUC = 0.998), significantly outperforming any single predictor (p < 0.001). The model achieved a sensitivity of 100% and a specificity of 98.7%, indicating its ability to identify nearly all patients with a favorable treatment response while correctly excluding the vast majority of non-responders. This integrated tool, which combines pathological, imaging, and non-invasive biomarkers, holds significant promise for enabling individualized post-operative management and advancing precision medicine in HCC care.

Discussion
Previous studies have shown that the invasiveness, vascular invasion tendency and recurrence risk of HCC are closely related to tumor marker levels, tumor size, number, location, morphology and boundary, histological grade, and large vessel infiltration.8,31,32
The results of this study indicate that the viral hepatitis status may be an important factor affecting the TP of MWA in HCC patients. The chronic inflammatory and sustained hepatocyte injury induced by hepatitis virus infection likely promote tumorigenesis and disease progression. Therefore, hepatitis etiology should be incorporated as a critical prognostic factor in therapeutic decision-making and outcome assessment for HCC.
Tumor size and high-risk location are conventional influencing factors known from previous studies; this study found that the margin characteristics, morphology, enhancement range and contrast washout time of the tumor were significantly correlated with the post-ablation recurrence in HCC, which was consistent with previous studies.32 Specifically, the larger the tumor, the less clear its borders, the more irregular its morphology, the greater the enhancement area compared to two-dimensional images, and the earlier the contrast agent washout during imaging, all showed a positive correlation with TP after ablation. However, there were no significant differences in tumor echogenicity, histological grade, intralesional vascularity, other CEUS parameters compared to post-treatment TP (all p > 0.05). These findings fully confirm the significant value of tumor morphology and imaging characteristics in predicting post-ablation progression of HCC. The larger the tumor diameter, the more technically demanding thermal ablation becomes, often requiring adjustments to the ablation strategy (ablation mode, applicator positioning and treatment cycles, combined with TACE or molecular-targeted agents) based on tumor size to achieve better therapeutic outcomes.33 The unclear tumor margins and irregular morphology may reflect the invasive growth pattern of the tumor, which may lead to an increased risk of TP after ablation.34
By integrating the information of patient age and liver fibrosis severity, API provides a simple and effective tool for risk stratification, prognosis evaluation and treatment optimization of liver cancer. However, its implementation in clinical practice often requires standardized calculation and combined with other indicators for comprehensive judgment. The results of cases showed that there was no significant correlation between conventional liver function parameters (such as ALB, TBIL, PLT) and post-ablation TP (all p > 0.05), suggesting that although the simple liver function parameters can effectively reflect the patient’s liver function reserve, they are not enough to reflect the biological behavior of the tumor. However, there was a significant correlation between API and TP (p = 0.368), and the mechanism may be related to the composite evaluation of API integrating liver disease background and host factors. The mean serum AFP and PIVKA-II levels before and after ablation were higher than normal levels, and the analysis showed that there was a significant correlation between TP after HCC ablation (all p < 0.05), which was consistent with previous studies.35 The changes in the levels of AFP and PIVKA-II, which are serum markers of HCC, can accurately reflect the biological behavior and therapeutic response of the tumor,36,37 Therefore, it has important reference value in the prognosis evaluation of HCC. The elevation of AFP level is usually associated with the proliferative activity and invasiveness of tumors, while PIVKA-II is closely related to the angiogenesis and invasiveness of tumors.38 Therefore, serial monitoring in AFP and PIVKA-II levels can provide important information for the individualized treatment and prognosis evaluation of HCC patients.
This study found that the positive expression of GPC-3, Arg-1 and HSP 70 and elevated Ki-67 proliferation index were significantly correlated with MWA postoperative TP in HCC patients. These results suggest that these biomarkers may be of great value in predicting the prognosis of HCC after ablation. GPC-3 positive expression is closely related to the invasiveness, vascular invasion and poor prognosis of HCC.39 In this study, GPC-3 positive expression was associated with TP, suggesting that it may be a potential marker to predict recurrence or progression after ablation. Arg-1 promotes tumor immune escape by inhibiting T cell function, which may lead to the proliferation and recurrence of residual tumor cells after ablation.40 Therefore, the positive expression of Arg-1 may be associated with TP after HCC ablation. HSP 70 may promote tumor cell survival, inhibit apoptosis and enhance the invasive ability of tumor cells, leading to TP after ablation.41 Arg-1 and HSP 70 are involved in tumor immune escape and stress response respectively, and their positive expression may indicate more aggressive biological behavior of tumor. In this study, HSP 70 positive expression is related to TP, which further supports its potential value in HCC prognosis evaluation. In addition, the study showed that elevated Ki-67 proliferation adhering to the electrode after radiofrequency ablation of liver tumor were independent predictors of local PFS and OS.42 The difference of tumor Ki-67 index in this group further suggested its role in the evaluation of prognosis after HCC ablation. Therefore, these IHC markers reflect the proliferative activity, differentiation status and potential invasiveness of tumors, which may provide new molecular markers for the prognosis evaluation of HCC.
The results of univariate and multivariate logistic regression analysis suggested that the morphology of tumor, liver function index, serum marker and immunohistochemical index could predict the TP after HCC ablation more accurately.
The irregularity of tumor morphology and the extended post-contrast enhancement range may reflect the invasive growth pattern of the tumor. The elevated preoperative serum AFP level suggests the proliferative activity and invasiveness of the tumor. API indicates an increased risk of transformation from liver cirrhosis to HCC, while the positive expression of GPC-3 and HSP 70 further suggests the biological behavior and treatment response of the tumor.
The ROC curve analysis demonstrated the diagnostic performance of the model, indicating that combining multiple predictive factors can significantly improve the accuracy of TP prediction following HCC treatment. The integrated assessment of CEUS parameters such as post-enhancement range, tumor morphology, GPC-3 and HSP 70 positive expression, Ki-67 index, API, and preoperative serum AFP levels not only increased the model’s sensitivity and specificity but also provides clinicians with a more comprehensive prognostic evaluation tool. This approach facilitates the development of personalized treatment strategies.
Overall, this study analyzed the clinical data from 91 patients with solitary HCC who underwent ultrasound-guided percutaneous MWA. The results demonstrated that multiple factors were significantly associated with TP following ablation, including hepatitis background, tumor size, margin status, morphological characteristics, post-enhancement features, contrast washout time, API, serum AFP levels and PIVKA-II levels, as well as immunohistochemical markers (GPC-3, Arg-1, HSP 70, and Ki-67 index). Multivariate logistic regression analysis further identified irregular tumor morphology, extended enhancement range, higher API score, preoperative serum AFP levels (> 200 ng/mL), positive expression of GPC-3 and HSP 70, and elevated Ki-67 index as independent predictors of HCC recurrence after ablation.
However, this study does have limitations. In the subsequent research, we will conduct multi-center studies to avoid sample bias issues. The settings of immunohistochemical indicators will be further verified and improved.

Conclusion
The predictive model constructed by these predictors showed high diagnostic performance in predicting the progression of HCC after ablation. These findings provide an important theoretical basis for the individualized treatment and prognosis evaluation of HCC patients.