Combined hematology, tumor morphology, and magnetic resonance imaging for predicting microvascular invasion in hepatocellular carcinoma: a clinical study

Introduction

Primary liver cancer is the sixth most common malignancy and the third leading cause of cancer-related death globally as of 2020, with hepatocellular carcinoma (HCC) accounting for approximately 90% of cases (1). HCC represents nearly half of global liver cancer incidence and mortality, posing a substantial public health threat (2). Radical surgical resection remains the primary treatment, essential for improving survival outcomes. However, due to HCC’s aggressive nature, postoperative recurrence and metastasis rates remain high and unsatisfactory, with 5-year recurrence and metastasis rates reaching 40–70%, severely affecting surgical outcomes, prognosis, and survival (3).

In recent years, microvascular invasion (MVI) has been recognized as an independent risk factor for HCC recurrence, with reported incidence rates reaching 57% (4-6). Its presence indicates a more aggressive HCC phenotype and a poorer prognosis. Studies reveal that MVI-positive patients exhibit significantly lower overall survival (OS) and recurrence-free survival (RFS) compared to MVI-negative patients (7,8). Due to tumor heterogeneity and the dissemination risk of needle biopsy, MVI status can only be determined from surgical specimens (9,10). Preoperative evaluation of MVI is essential for personalized diagnosis and optimizing treatment strategies in clinical decision-making (11-15). Thus, developing a novel, non-invasive, and highly specific diagnostic method for MVI is a pressing clinical need.

In recent years, the preoperative prediction of MVI has become a prominent focus in HCC research. Common prediction approaches include artificial intelligence (AI) techniques based on radiomics and deep learning models, imaging feature analysis, serum protein marker detection, and the identification of novel molecular markers. However, AI-based methods face high data dependency, limited generalization, interpretability issues, and poor accessibility. In contrast, serum protein markers are easily accessible but often lack adequate sensitivity and specificity, while novel molecular markers show promise yet require further validation. Some studies investigating the relationship between comprehensive factor and MVI have identified that certain serological indicators, tumor morphology, and specific imaging features exhibit potential in predicting MVI (11,16,17). Building on this background, this study aimed to develop a predictive model that integrates hematologic parameters, tumor morphology, and magnetic resonance imaging (MRI) findings for MVI prediction in HCC. This model seeks to inform more appropriate treatment strategies for HCC patients, ultimately enhancing the overall therapeutic efficacy for HCC. We present this article in accordance with the TRIPOD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-758/rc).

Methods

Study and participants

A total of 236 patients were enrolled in this study. All patients were diagnosed with HCC and underwent hepatectomy at The Affiliated Hospital of Guangdong Medical University between January 1, 2019, and December 31, 2022. Of these, 118 patients were randomly assigned to the training cohort and the remaining 118 to the validation cohort (Figure 1). The clinical data for this study were obtained from the medical records system of The Affiliated Hospital of Guangdong Medical University. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by the Ethics Committee of The Affiliated Hospital of Guangdong Medical University (No. PJKT2022-107). The requirement for informed consent was waived due to the retrospective nature of the analysis.

Figure 1 Patients meeting the criteria were randomly assigned to a training cohort and a validation cohort. The high-risk factors for MVI in the training group were statistically analyzed, and a nomogram was generated. The effectiveness and clinical benefits of the model were subsequently validated by the validation group. HCC, hepatocellular carcinoma; MRI, magnetic resonance imaging; MVI, microvascular invasion.

Inclusion criteria: (I) patients undergoing surgical treatment for HCC, including open or laparoscopic hepatectomy; (II) patients with preoperative hematological and MRI examinations, complete medical records and preoperative imaging data indicating primary liver cancer, confirmed by postoperative pathology, according to the 2022 guidelines for the diagnosis and treatment of primary liver cancer (3); (III) complete surgical specimens with pathological findings related to MVI; (IV) MRI images indicating no macrovascular invasion (including hepatic artery, portal vein, or hepatic vein), no distant intrahepatic or extrahepatic metastasis and the number of tumors was within 3, which was assessed by surgeons as resectable.

Exclusion criteria: (I) patients who did not meet the diagnostic criteria for HCC; (II) missing clinical examination or pathological data, or postoperative pathology indicating other types of liver cancer; (III) hematology and MRI examinations not performed in our hospital within 1 month before surgery; (IV) preoperative findings indicating multiple lesions (more than 4), macrovascular invasion, or distant metastasis that were unresectable as assessed by the surgeon; (V) receipt of any other anti-tumor treatment prior to surgery; (VI) significant MRI artifacts or poor image quality.

Clinical data

Patient data on gender, age, and preoperative hematological tests were collected, including immunological markers for hepatitis B, alpha-fetoprotein (AFP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), γ-glutamyl transpeptidase (GGT), alkaline phosphatase (ALP), total bilirubin (TBil), direct bilirubin (DBil), albumin (Alb), platelet count (PLT), prothrombin time (PT), white blood cell count, lymphocyte count, neutrophil count, and neutrophil-to-lymphocyte ratio (NLR).

Image analysis

All images were stored in the picture archiving and communication system (PACS), which are reviewed and reported by two radiologists with more than 15 years of work experience, who are mainly responsible for abdominal imaging diagnosis. The analysis focused on tumor size, number, capsule integrity, arterial phase peritumoral enhancement, intratumoral hemorrhage, and tumor morphology smoothness (Figure 2). These features were defined according to the 2018 version of the Liver Imaging Reporting and Data System (LI-RADS) (18). The diameter of the tumor is the distance between the maximum outer edges of the lesion, measured in the phase where the boundary is most clearly visible. In cases of significant disagreement between the two radiologists, a consensus report was produced following repeated reviews and discussions. During this process, the radiologists were blinded to the general condition, serological test results, and final MVI status of all patients.

Figure 2 Representative images of three liver cancer patients with positive MVI. Incomplete capsule around the tumor in arterial phase (A), portal venous phase (B), and T2 phase (C). Enhancement around the tumor in arterial phase (D), portal venous phase (E), and T2 phase (F). Intratumoral vascular enhancement and intratumoral hemorrhage in arterial phase (G), portal venous phase (H), and T2 phase (I). MVI, microvascular invasion.

Pathological feature

Pathological sampling methods and diagnostic criteria followed the 2022 edition of the Guidelines for the Diagnosis and Treatment of Primary Liver Cancer (3). Pathologists collected and numbered liver cancer tissue specimens using the “7-point sampling method” and subsequently stained them with hematoxylin-eosin (HE) and CD34. MVI pathological features are presented in Figure 3. Pathological reports were issued based on these findings. The reports included the Edmonson-Steiner grade, tumor size, and MVI status. Since MVI cannot be determined preoperatively, this study did not include tumor differentiation as one of the final preoperative predictors of MVI. MVI diagnosis for all specimens in this study was conducted by two pathologists with intermediate or senior professional titles who had worked for more than 15 years and were mainly responsible for the diagnosis of abdominal digestive diseases, and verified by one chief pathologist who had worked for more than 20 years and whose main research direction was liver cancer. In cases of disagreement, the three pathologists reached a consensus after discussion. Throughout the process, the pathologists were blinded to patients’ medical records, examination results, and imaging data, and were solely responsible for specimen collection, slide preparation, and MVI assessment through microscopic examination. It is important to note that the sampling protocols and MVI grading system outlined in this guideline are specific to Chinese patients, and their applicability to European and American populations remains unvalidated.

Figure 3 Representative images of MVI pathological features. Under the microscope, nests of cancer cells are observed within the vascular lumens lined with endothelial cells. Hematoxylin-eosin staining; magnification, 200×. MVI, microvascular invasion.

Statistical analysis

All continuous data were tested for normality. Normally distributed data were presented as mean ± standard deviation and analyzed using the t-test, while skewed data were presented as median (interquartile range) and analyzed using the Wilcoxon rank-sum test. Chi-squared test or Fisher’s exact test was used to compare categorical variables. After univariate analysis, the Benjamini-Hochberg method was applied for false discovery rate (FDR) correction of the P values of all variables (with a preset FDR threshold α=0.05). Using an adjusted P value <0.05 as the selection criterion, variables meeting this standard were incorporated into subsequent multivariate analysis to further identify the independent risk factors for MVI. The identified risk factors were incorporated to construct a nomogram model, and the model’s goodness of fit was assessed using the Hosmer-Lemeshow test. To quantify the discriminatory capacity for predicting MVI within the training set, the ROC curve was plotted, and the area under the curve (AUC) was calculated. Furthermore, clinical benefit was evaluated using decision curve analysis (DCA) and calibration curve. To further evaluate the stability of the predictive model and address the issue that features selected from a single small cohort tend not to generalize well to external datasets, we performed k-fold cross-validation (k=5) on the entire dataset. All statistical analyses were conducted through SPSS 26.0 and R software. A two-sided P value of less than 0.05 was considered statistically significant.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work the authors used ChatGPT (GPT-4o) in order to improve readability and language of the work. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Results

Comparison of baseline characteristics

A total of 236 patients were enrolled based on the inclusion and exclusion criteria, including 64 (27.1%) MVI-positive and 172 (72.9%) MVI-negative patients, with a mean age of 56.96±11.16 years. Patients were randomly assigned to the training cohort (n=118) and validation cohort (n=118), with 31 (26.3%) MVI-positive patients in the training cohort and 33 (28.0%) in the validation cohort. No significant differences in baseline characteristics, including gender, age, serological markers, liver function tests, and tumor staging, were found between MVI-positive and MVI-negative patients in the training cohort (FDR-adjusted P>0.05), as shown in Table 1.

Comparison of AFP and MRI features

In the training cohort, comparisons of AFP and MRI features showed statistically significant differences between the MVI-positive and MVI-negative groups in AFP >800 ng/mL, tumor diameter >5 cm, intratumoral vascular enhancement, arterial phase edge enhancement, incomplete tumor capsule, irregular tumor morphology, intratumoral hemorrhage, and postoperative pathological grade (FDR-adjusted P<0.05), all of which were identified as potential risk factors for MVI occurrence. However, no statistically significant difference was observed in the presence of multiple tumor (FDR-adjusted P>0.05), as shown in Table 2. MRI features are presented in Figure 2. Since pathological grade can only be determined postoperatively, it was excluded as a preoperative MVI predictor in this study. After excluding pathological grade, the seven factors above were entered into a binary logistic regression analysis. The results indicated that AFP >800 ng/mL [OR =3.407, 95% confidence interval (CI): 1.09–10.62, P=0.035], intratumoral vascular enhancement (OR =6.300, 95% CI: 1.09–10.62, P=0.022), incomplete tumor capsule (OR =4.513, 95% CI: 1.38–14.81, P=0.013), and arterial phase edge enhancement (OR =4.610, 95% CI: 1.42–14.92, P=0.011) were independent risk factors for MVI in HCC patients. Detailed data are presented in Table 3.

ROC curve and nomogram

ROC curve analysis revealed that AFP >800 ng/mL, intratumoral vascular enhancement, incomplete tumor capsule, and arterial phase edge enhancement exhibited high diagnostic efficacy in predicting HCC with MVI, with respective AUCs of 0.648, 0.650, 0.725, and 0.715 (Figure 4A). In both the training and validation cohorts, the AUC for the combined prediction model was 0.871 and 0.878, respectively (Table 4, Figure 4A,4B). Based on these results, the nomogram prediction model constructed by these four factors is illustrated in Figure 5. Furthermore, the results of the Hosmer-Lemeshow test for both prediction models yielded P=0.375 and P=0.169, respectively, exceeding 0.05, indicating high goodness of fit with no significant difference between predicted and actual outcomes. The calibration curves also demonstrated that the predicted nomogram closely aligned with the actual probability of MVI in both cohorts (Figure 6A,6B). Moreover, DCA showed that the predicted nomogram curves for all MVI patients surpassed the default line across all reasonable threshold probabilities (Figure 6C,6D).

Figure 4 The ROC curves of the MVI predicting model in the training and validation cohorts. (A) The ROC curves of the MVI predicting model in the training cohort (AUC: 0.871). (B) The ROC curves of the MVI predicting model in the validation cohort (AUC: 0.878). AFP, alpha-fetoprotein; AUC, alanine aminotransferase; CI, confidence interval; MVI, microvascular invasion; ROC, receiver operating characteristic.

Figure 5 Nomogram for predicting MVI of liver cancer based on clinical features and MRI features. *, P<0.05; **, P<0.01. AFP, alpha-fetoprotein; MRI, magnetic resonance imaging; MVI, microvascular invasion.

Figure 6 Calibration curve and DCA curve of nomogram in the training cohort and validation cohort. (A) Calibration curve of nomogram in the training cohort. (B) Calibration curve of nomogram in the validation cohort. (C) DCA curve of nomogram in the training cohort. (D) DCA curve of nomogram in the validation cohort. DCA, decision curve analysis.

Cross-validation

The results of the 5-fold cross-validation are shown in Table 5. The mean values of AUC, sensitivity, specificity, and accuracy were 0.500, 0.901, 0.860, and 0.792, respectively. This indicates that the model has stable overall discriminative ability for positive and negative samples, good predictive accuracy, and relatively strong ability to identify negative samples, but insufficient ability to identify positive samples.

Discussion

In this study, the combined predictive model yielded results with an AUC of 0.871, a sensitivity of 77.8%, and a specificity of 80.0%, as shown in Table 4. A prior meta-analysis indicated that using the apparent diffusion coefficient (ADC) alone to predict MVI yielded an AUC of 0.78, with a sensitivity and specificity of 73% and 70%, respectively (19). The predictive model developed in this study exhibited superior diagnostic efficacy compared to that model.

Among all laboratory tests, only AFP >800 ng/mL was identified as an independent risk factor for MVI. AFP is clinically used in screening, diagnosis, assessing surgical effectiveness, and prognosis in HCC. High AFP levels not only indicate HCC but also reflect high tumor invasiveness, correlating with worse treatment outcome and prognosis (20). The findings across studies vary due to differences in AFP threshold levels. Ryu et al. (21) identified AFP >95 ng/mL as an independent predictor of MVI, while Deng et al. (22) found that AFP >155 ng/mL also served as an independent predictor of MVI. AFP levels may also rise in hepatitis, cirrhosis, other tumors, and some benign conditions. Consequently, the specific AFP threshold for predicting MVI necessitates further clinical investigation.

While a larger tumor diameter is widely acknowledged as a risk factor for MVI, the threshold for predicting MVI based on tumor size remains controversial (23). Deng et al. (22) reported a 25% risk of MVI for tumors <3 cm, 40% for tumors measuring 3–5 cm, and 63% for tumors measuring 5–6.5 cm. Pawlik et al. identified a tumor diameter >5 cm as an independent predictor of MVI (24). In this study, the incidence of MVI was 20.5% for tumors with a diameter <5 cm, compared to 57.5% for those >5 cm. However, a tumor diameter >5 cm did not emerge as an independent risk factor, differing from previous studies.

Gadoxetic acid-enhanced MRI (Gd-EOB-DTPA MRI) offers superior soft tissue resolution and more comprehensive diagnostic information than computed tomography (CT) improving diagnostic accuracy. This study used Gd-EOB-DTPA MRI for MVI prediction. The results showed that the imaging features closely associated with MVI occurrence included incomplete capsule, arterial phase edge enhancement and intratumoral vascular enhancement.

Some studies have shown that the capsule is a crucial structure to prevent tumor invasion; however, others indicate no relationship between metastasis and capsule integrity (25). A study found an incomplete tumor capsule predictive of MVI, whereas Zhu et al. contended that capsule integrity was not a predictive factor for MVI (26,27). In this study, an incomplete capsule emerged as an independent risk factor for MVI prediction. Some perspectives suggest that the capsule serves as both a protective factor and a risk factor. It isolates the tumor from normal extrahepatic tissue, partially hindering tumor invasion; however, the tumor can readily invade the blood vessels within the capsule, leading to MVI. This dual effect may explain the differing views on capsule integrity’s relevance to MVI.

It has been reported that arterial phase peritumoral enhancement of HCC on contrast-enhanced CT and Gd-EOB-DTPA-enhanced MRI is a significant predictor of MVI (28). Arterial phase peritumoral enhancement may reflect cellular proliferation, tumor tissue distortion induced by MVI, inflammation associated with extracellular matrix remodeling, and changes in hemodynamic perfusion related to MVI in HCC (29). Studies have suggested that the mechanism behind peritumoral arterial enhancement in MVI-positive HCC may involve obstruction of portal vein branches by small peritumoral emboli, resulting in reduced or absent portal venous flow and subsequent regional compensatory arterial hyperperfusion (30). Furthermore, intratumoral artery enhancement may also be a sign of compensatory arterial hyperperfusion resulting from microvascular tumor thrombus obstructing small hepatic vein branches. Segal identified that intratumoral arterial enhancement is highly predictive of MVI based on genomic and imaging characteristics (31). In this study, both arterial phase peritumoral enhancement and intratumoral vascular enhancement were independent risk factors for MVI, consistent with previous research findings.

A retrospective study of 323 postoperative liver cancer patients identified a tumor count exceeding two as an independent risk factor for MVI (32). Increased tumor cell count may indicate higher invasiveness, as tumors invade microvessels and metastasize intrahepatically via the portal vein. However, other studies have reported no association between tumor count and MVI in HCC patients (33). This study suggested that tumor count was not an independent risk factor for MVI.

There are several limitations in this study: (I) being a single-center retrospective study, it is subject to inherent selection bias, necessitating multicenter and prospective studies for validation. (II) The small sample size and lack of external validation require an expanded sample and external validation to establish accuracy. (III) The MVI prediction model developed in this study may be more applicable to Asian populations with hepatitis B virus-related HCC. Its validity in non-viral HCC or among European and American populations requires further verification. (IV) The feature extraction of the model in this study relies on manual selection from magnetic resonance images, which is vulnerable to interference from the operator’s subjective experience and may lead to annotation biases. In large-scale research, such reliance on manual work may further result in low implementability of the method, poor reproducibility of results, and limited extensibility of the study. To address this limitation, our team will focus on exploring two key directions in subsequent work: first, introducing multimodal radiomics features; second, integrating the deep learning image reconstruction (DLIR) algorithm to optimize the quality of input images.

Conclusions

In this study, the model developed using AFP, intratumoral vascular enhancement, arterial phase peritumoral enhancement, and incomplete tumor capsule can predict the presence of MVI in HCC patients to a certain extent. Furthermore, these four factors are easily accessible in clinical practice, enhancing the model’s practicality and accuracy. This model may offer valuable insights for formulating treatment decisions in HCC patients. However, its diagnostic efficacy requires further validation in large-sample, multicenter studies.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-758/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-758/dss

Funding: This study was supported by the Natural Science Foundation of Guangdong Province (No. 2024A1515220108); Affiliated Hospital of Guangdong Medical University 2023 In-house Clinical Research Project (No. LCYJ2023A001); Affiliated Hospital of Guangdong Medical University 2021 In-house Clinical Research Project (No. LCYJ2021B002); Beijing Science and Technology Innovation Medical Development Foundation (No. KC2023-JX-0186-FQ036).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-758/coif). Y.G.Y. was supported by the Natural Science Foundation of Guangdong Province (No. 2024A1515220108); Affiliated Hospital of Guangdong Medical University 2023 In-house Clinical Research Project (No. LCYJ2023A001); Affiliated Hospital of Guangdong Medical University 2021 In-house Clinical Research Project (No. LCYJ2021B002); Beijing Science and Technology Innovation Medical Development Foundation (No. KC2023-JX-0186-FQ036). The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Affiliated Hospital of Guangdong Medical University (No. PJKT2022-107). Individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Iguchi T, Shirabe K, Aishima S, Wang H, Fujita N, Ninomiya M, Yamashita Y, Ikegami T, Uchiyama H, Yoshizumi T, Oda Y, Maehara Y. New Pathologic Stratification of Microvascular Invasion in Hepatocellular Carcinoma: Predicting Prognosis After Living-donor Liver Transplantation. Transplantation 2015;99:1236-42. [Crossref] [PubMed]
3. Zhou J, Sun H, Wang Z, Cong W, Zeng M, Zhou W, et al. Guidelines for the Diagnosis and Treatment of Primary Liver Cancer (2022 Edition). Liver Cancer 2023;12:405-44. [Crossref] [PubMed]
4. Rodríguez-Perálvarez M, Luong TV, Andreana L, Meyer T, Dhillon AP, Burroughs AK. A systematic review of microvascular invasion in hepatocellular carcinoma: diagnostic and prognostic variability. Ann Surg Oncol 2013;20:325-39. [Crossref] [PubMed]
5. Zhang X, Li J, Shen F, Lau WY. Significance of presence of microvascular invasion in specimens obtained after surgical treatment of hepatocellular carcinoma. J Gastroenterol Hepatol 2018;33:347-54. [Crossref] [PubMed]
6. Lim KC, Chow PK, Allen JC, Chia GS, Lim M, Cheow PC, Chung AY, Ooi LL, Tan SB. Microvascular invasion is a better predictor of tumor recurrence and overall survival following surgical resection for hepatocellular carcinoma compared to the Milan criteria. Ann Surg 2011;254:108-13. [Crossref] [PubMed]
7. Chen ZH, Zhang XP, Wang H, Chai ZT, Sun JX, Guo WX, Shi J, Cheng SQ. Effect of microvascular invasion on the postoperative long-term prognosis of solitary small HCC: a systematic review and meta-analysis. HPB (Oxford) 2019;21:935-44. [Crossref] [PubMed]
8. Erstad DJ, Tanabe KK. Prognostic and Therapeutic Implications of Microvascular Invasion in Hepatocellular Carcinoma. Ann Surg Oncol 2019;26:1474-93. [Crossref] [PubMed]
9. Court CM, Harlander-Locke MP, Markovic D, French SW, Naini BV, Lu DS, Raman SS, Kaldas FM, Zarrinpar A, Farmer DG, Finn RS, Sadeghi S, Tomlinson JS, Busuttil RW, Agopian VG. Determination of hepatocellular carcinoma grade by needle biopsy is unreliable for liver transplant candidate selection. Liver Transpl 2017;23:1123-32. [Crossref] [PubMed]
10. Szpakowski JL, Drasin TE, Lyon LL. Rate of seeding with biopsies and ablations of hepatocellular carcinoma: A retrospective cohort study. Hepatol Commun 2017;1:841-51. [Crossref] [PubMed]
11. Yamashita Y, Tsuijita E, Takeishi K, Fujiwara M, Kira S, Mori M, Aishima S, Taketomi A, Shirabe K, Ishida T, Maehara Y. Predictors for microinvasion of small hepatocellular carcinoma ≤ 2 cm. Ann Surg Oncol 2012;19:2027-34. [Crossref] [PubMed]
12. Hirokawa F, Hayashi M, Miyamoto Y, Asakuma M, Shimizu T, Komeda K, Inoue Y, Uchiyama K. Outcomes and predictors of microvascular invasion of solitary hepatocellular carcinoma. Hepatol Res 2014;44:846-53. [Crossref] [PubMed]
13. Mazzaferro V, Llovet JM, Miceli R, Bhoori S, Schiavo M, Mariani L, et al. Predicting survival after liver transplantation in patients with hepatocellular carcinoma beyond the Milan criteria: a retrospective, exploratory analysis. Lancet Oncol 2009;10:35-43. [Crossref] [PubMed]
14. Ueno M, Hayami S, Shigekawa Y, Kawai M, Hirono S, Okada K, Tamai H, Shingaki N, Mori Y, Ichinose M, Yamaue H. Prognostic impact of surgery and radiofrequency ablation on single nodular HCC ≤5 cm: Cohort study based on serum HCC markers. J Hepatol 2015;63:1352-9. [Crossref] [PubMed]
15. Lee S, Kang TW, Song KD, Lee MW, Rhim H, Lim HK, Kim SY, Sinn DH, Kim JM, Kim K, Ha SY. Effect of Microvascular Invasion Risk on Early Recurrence of Hepatocellular Carcinoma After Surgery and Radiofrequency Ablation. Ann Surg 2021;273:564-71. [Crossref] [PubMed]
16. Nitta H, Allard MA, Sebagh M, Karam V, Ciacio O, Pittau G, Vibert E, Sa Cunha A, Cherqui D, Castaing D, Bismuth H, Guettier C, Samuel D, Baba H, Adam R. Predictive model for microvascular invasion of hepatocellular carcinoma among candidates for either hepatic resection or liver transplantation. Surgery 2019;165:1168-75. [Crossref] [PubMed]
17. Huang Z, Zhu RH, Li SS, et al. Comparison of Sonazoid-Contrast‑Enhanced Ultrasound and Gd‑EOB‑DTPA‑Enhanced MRI for Predicting Microvascular Invasion in Hepatocellular Carcinoma. Ultrasound Med Biol 2024;50:1339-45. [Crossref] [PubMed]
18. Chernyak V, Fowler KJ, Kamaya A, Kielar AZ, Elsayes KM, Bashir MR, Kono Y, Do RK, Mitchell DG, Singal AG, Tang A, Sirlin CB. Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in At-Risk Patients. Radiology 2018;289:816-30. [Crossref] [PubMed]
19. Deng Y, Li J, Xu H, Ren A, Wang Z, Yang D, Yang Z. Diagnostic Accuracy of the Apparent Diffusion Coefficient for Microvascular Invasion in Hepatocellular Carcinoma: A Meta-analysis. J Clin Transl Hepatol 2022;10:642-50. [Crossref] [PubMed]
20. Lin K, Huang Q, Zeng J, Ding Z, Wang L, Chen Z, Guo P, Zeng Y, Zhou W, Liu J. Clinical Significance of Alpha-Fetoprotein in Alpha-Fetoprotein Negative Hepatocellular Carcinoma Underwent Curative Resection. Dig Dis Sci 2021;66:4545-56. [Crossref] [PubMed]
21. Ryu T, Takami Y, Wada Y, Tateishi M, Hara T, Yoshitomi M, Momosaki S, Yasumori K, Saitsu H, Okuda K. A Clinical Scoring System for Predicting Microvascular Invasion in Patients with Hepatocellular Carcinoma Within the Milan Criteria. J Gastrointest Surg 2019;23:779-87. [Crossref] [PubMed]
22. Deng G, Yao L, Zeng F, Xiao L, Wang Z. Nomogram For Preoperative Prediction Of Microvascular Invasion Risk In Hepatocellular Carcinoma. Cancer Manag Res 2019;11:9037-45. [Crossref] [PubMed]
23. Nagano Y, Shimada H, Takeda K, Ueda M, Matsuo K, Tanaka K, Endo I, Kunisaki C, Togo S. Predictive factors of microvascular invasion in patients with hepatocellular carcinoma larger than 5 cm. World J Surg 2008;32:2218-22. [Crossref] [PubMed]
24. Pawlik TM, Delman KA, Vauthey JN, Nagorney DM, Ng IO, Ikai I, Yamaoka Y, Belghiti J, Lauwers GY, Poon RT, Abdalla EK. Tumor size predicts vascular invasion and histologic grade: Implications for selection of surgical treatment for hepatocellular carcinoma. Liver Transpl 2005;11:1086-92. [Crossref] [PubMed]
25. Song L, Li J, Luo Y. The importance of a nonsmooth tumor margin and incomplete tumor capsule in predicting HCC microvascular invasion on preoperative imaging examination: a systematic review and meta-analysis. Clin Imaging 2021;76:77-82. [Crossref] [PubMed]
26. Zhu YJ, Feng B, Wang BZ, Wang S, Ye F, Ma XH, Zhao XM. Value of gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid enhanced magnetic resonance imaging and diffusion-weighted MR imaging in predicting microvascular invasion in hepatocellular carcinoma and the prognostic significance. Zhonghua Zhong Liu Za Zhi 2021;43:312-7. [Crossref] [PubMed]
27. Zhu F, Yang F, Li J, Chen W, Yang W. Incomplete tumor capsule on preoperative imaging reveals microvascular invasion in hepatocellular carcinoma: a systematic review and meta-analysis. Abdom Radiol (NY) 2019;44:3049-57. [Crossref] [PubMed]
28. Wu G, Wu J, Wang B, Zhu X, Shi X, Ding Y. Importance of tumor size at diagnosis as a prognostic factor for hepatocellular carcinoma survival: a population-based study. Cancer Manag Res 2018;10:4401-10. [Crossref] [PubMed]
29. Deng Y, Yang D, Tan X, Xu H, Xu L, Ren A, Liu P, Yang Z. Preoperative evaluation of microvascular invasion in hepatocellular carcinoma with a radiological feature-based nomogram: a bi-centre study. BMC Med Imaging 2024;24:29. [Crossref] [PubMed]
30. Choi JY, Lee JM, Sirlin CB. CT and MR imaging diagnosis and staging of hepatocellular carcinoma: part II. Extracellular agents, hepatobiliary agents, and ancillary imaging features. Radiology 2014;273:30-50. [Crossref] [PubMed]
31. Segal E, Sirlin CB, Ooi C, Adler AS, Gollub J, Chen X, Chan BK, Matcuk GR, Barry CT, Chang HY, Kuo MD. Decoding global gene expression programs in liver cancer by noninvasive imaging. Nat Biotechnol 2007;25:675-80. [Crossref] [PubMed]
32. Zhou M, Wang H, Zeng X, Yin P, Zhu J, Chen W, et al. Mortality, morbidity, and risk factors in China and its provinces, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2019;394:1145-58. [Crossref] [PubMed]
33. Wang H, Lu Y, Liu R, Wang L, Liu Q, Han S. A Non-Invasive Nomogram for Preoperative Prediction of Microvascular Invasion Risk in Hepatocellular Carcinoma. Front Oncol 2021;11:745085. [Crossref] [PubMed]