|Year : 2021 | Volume
| Issue : 3 | Page : 652-657
A nomogram to predict microvascular invasion in early hepatocellular carcinoma
Hongguang Li1, Tao Li2, Jinhua Hu2, Jun Liu3
1 Department of Hepatobiliary Surgery, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Shandong, 250021, China
2 Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, 250021, China
3 Department of Hepatobiliary Surgery, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Department of Hepatobiliary Surgery, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, 250021, China
|Date of Submission||27-Nov-2020|
|Date of Decision||20-Jan-2021|
|Date of Acceptance||10-Feb-2021|
|Date of Web Publication||9-Jul-2021|
Department of Gastroenterology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong 250021
Source of Support: None, Conflict of Interest: None
Aim: To construct an integrated nomogram combining protein induced by vitamin K antagonist-II (PIVKA-II), alpha fetoprotein (AFP) and other clinical factors to detect microvascular invasion (MVI) in early hepatocellular carcinoma (HCC) patients with single nodule.
Methods: One hundred and eleven early HCC patients were enrolled in the present study and 43 early HCC patients were diagnosed with MVI. Serum levels of PIVKA-II, AFP and other laboratory indicators were detected. Chi-squared test, t-test and logistic regression were employed in statistic analysis. A nomogram combining independent predictors was constructed and internal validated.
Results: In early HCC patients with MVI, PIVKA-II serum level was significantly higher than those without MVI (385.97 mAU/ml vs 67.08 mAU/ml; P < 0.01), as well as AFP serum level (81.6 ng/mL vs 9.15 ng/mL P = 0.001). PIVAK-II, AFP serum levels and tumor size were independent risk factors for MVI in early HCC, which was employed to develop a logistic regression model. The area under the ROC curve (AUROC) of the model was 0.74 (95%CI 0.65 - 0.84). A nomogram combining PIVKA-II, AFP and tumor size was constructed and calibration curves showed that the model was accurate in predicting the risk of MVI in early HCC patients.
Conclusion: The present study indicates that a preoperative nomogram combining PIVKA-II, AFP and tumor size could estimate the preoperative probability of MVI in early HCC patients, which may help clinicians in choosing treatment options and prognosis evaluation.
Keywords: Early hepatocellular carcinoma, microvascular invasion, nomogram
|How to cite this article:|
Li H, Li T, Hu J, Liu J. A nomogram to predict microvascular invasion in early hepatocellular carcinoma. J Can Res Ther 2021;17:652-7
| > Introduction|| |
Hepatocellular carcinoma (HCC) is the fourth leading cause of cancer-related death in the world, and it can be treated with surgical resection, local ablation, or liver transplantation when diagnosed at early stage. However, the long-term survival outcome is still unsatisfactory due to local recurrence and multicentric occurrence. Microvascular invasion (MVI) plays a critical role in the prediction of its postoperative recurrence., Moreover, it has been recognized as an independent predictor for overall and disease-free survival of HCC patients.,,
The incidence of MVI varied between 15% and 57.1% in surgical specimens obtained after liver resection and transplantation, which may be due to the different tumor stages and diagnostic criteria of MVI. For instance, in a recent study on early HCC, 29.55% patients (13 in 44 early HCC patients) were found MVI positive, while in a Chinese cohort, 83 patients (18.1%) among 458 patients with single small HCC (≤3 cm) were identified with MVI. Yamashita, et al. demonstrated that the positive rate of MVI was 40.6% among early HCC patients with single nodule ≤3 cm (168/414 patients). Unlike macrovascular invasion, which can be diagnosed with medical imaging, MVI can only be detected by pathological examination on resected surgical specimens currently. Currently, a series of studies on preoperative evaluation of MVI have been carried out among HCC patients. For instance, some investigators focus their study on radiological features of HCC patients with MVI before surgery., Tumor biomarkers such as alpha-fetoprotein (AFP) and prothrombin induced by Vitamin K absence-II (PIVKA-II) have also been investigated in the predication of MVI among HCC patients.
Both PIVKA-II and AFP are important laboratory parameters used in clinical to detect early HCC. Recent studies demonstrated that PIVKA-II had a high specificity and sensitivity of 90% and 56% for early HCC, respectively., Ryu, et al. demonstrated that AFP >95 ng/mL and PIVKA-II >55 mAU/mL were independent predictors of MVI among HCC patients with single or multiple nodules. The predicting efficacy of PIVKA-II and AFP for MVI in early HCC patients has also been reported. Yamashita, et al. found that serum levels of AFP ≥200 ng/mL and PIVKA-II ≥40 mAU/mL were independent predictors for MVI among patients with single small HCC ≤3 cm. However, Yamashita, et al. demonstrated that PIVKA-II level (^100 mAU/ml) but not AFP was at risk for MVI for patients with HCC ≤2 cm. Moreover, in a study carried out in France, according to the result of PIVKA-II immunostainings of the surgical specimen in HCC patients, n considering a single marker, PIVKA showed high accuracy to predict MVI (AUC = 0.754, 95% confidence interval [CI]: 0.602–0.906).
Based on the above findings, we hypothesized that a preoperative nomogram combining PIVKA-II and AFP could be helpful to evaluate MVI in early HCC patients. The purpose of this study is to investigate the value of PIVKA-II and AFP serum levels for MVI detection and construct a preoperative nomogram to evaluate the MVI in early HCC patients.
| > Patients and Methods|| |
A total of 111 consecutive patients with early HCC were enrolled in this study between October 2017 and April 2020. Histopathological examination was performed on each resected liver specimen to diagnose HCC and detect MVI. Early stage of HCC was diagnosed with the Barcelona liver clinic (BCLC) staging system, which was defined as a single lesion between 2 and 5 cm, or ≤3 lesions each ≤3 cm., Clinicopathological and laboratory data were obtained from each subject after written informed consents were obtained. MVI is defined as “a cancer cell nest with >50 cells in the endothelial vascular lumen under microscopy”, according to a guideline proposed by the Chinese Society of Pathology.
The protocol of the current study is consistent with the ethical guidelines of the 1975 Declaration of Helsinki and it was approved by the Ethical Committee of Shandong Provincial Hospital affiliated to Shandong University. Written informed consent was obtained from each subject.
Serum sample collection and assays
Before surgeries, peripheral blood samples were obtained from each subject. PIVKA-II serum levels were examined with Lumipulse G PIVKA-II reagent kits (FUJIREBIO Inc., Japan) on a LUMI-PULSE g1200 automatic immune analyzer according to the manufacturer's manual. MVI was defined as a tumor within a vascular space lined by endothelium, which was visible only with microscopy.
Development of logistic regression model
Baseline variables showing univariate relationship in MVI detection or considered clinically with MVI were employed to develop a logistic regression model.
A formula for the detection of MVI in early HCC patients was developed based on the subjects of the analysis group. The standard logistic regression formula is as follows:
Logit (StexP) =β0+β1 × 1+β2 × 2+…… +βnXn. “P” is the estimated probability of MVI in early HCC cohort, while “n” is the number of influence factors, “β” represents the influence coefficient, “X” is the influence factor, and “β0” is a constant.
Chi-square test was employed to compare categorical variables, while continuous variables with normal distribution were compared using Student's t-test. Logistic regression analysis is performed for independent variables of risk factors for MVI detection in early HCC patients. The accuracy of the model was evaluated by receiver operating characteristic (ROC) curve analysis for MVI detection in early HCC patients. Statistical analysis was carried out with SPSS version 21.0 (SPSS, Chicago, Illinois), Prism6 (GraphPad Software, La Jolla, California, USA), and the regression modeling strategy (rms) in the R software package version 3.5.3 (http://www.r-project.org/). Based on the results of multivariate logistic regression analysis, a visual nomogram was established using the rms package of R. C-indexes and calibration with 1000 bootstrap samples to decrease bias were used to measure the performance of the nomogram. Statistical analysis was tested on two-sided settings and P < 0.05 was considered as statistically significant.
| > Results|| |
Baseline characteristics of all subjects
A total of 111 early HCC patients with hepatectomy were enrolled in the present study. Among them, 43 (38.74%) early HCC patients were detected with MVI confirmed by histopathological examination. [Table 1] summarizes the baseline characteristics of all patients. The average age was 57.11 years, and 89.19% of the subjects were male. All HCC patients were BCLC stage 0-A3 with single nodule.
Higher prothrombin induced by Vitamin K absence-II and alpha-fetoprotein serum levels were detected in early hepatocellular carcinoma patients with microvascular invasion than in controls
Elevated PIVKA-II and AFP serum levels were detected in early HCC patients with MVI compared with early HCC patients without MVI, as shown in [Figure 1] after logarithmic transformation. Serum level of PIVKA-II among early HCC patients with MVI was 537.13 (40.97–36587.2) mAU/ml, statistically higher than that in controls 67.08 (4.4–22344.02) mAU/ml (P < 0.01) [Table 1]. Serum level of AFP among early HCC patients with MVI was 99.2 (0.9–8720.20) ng/ml, which was statistically elevated than that in early HCC patients without MVI 9.15 (1.00–6637.5) ng/ml (P < 0.01) [Table 1].
Univariate analysis demonstrated that serum PIVKA-II level (odds ratio [OR] = 3.25, 95% CI: 1.80–5.90) and serum AFP level (OR = 2.04, 95% CI: 1.36–3.04) were risk factors for detection of MVI after logarithmic transformation in early HCC patients [Table 2].
|Figure 1: (a) Difference of PIVKA plasma levels after logarithmic transformation between early HCC patients with and without MVI. Serum PIVKA-II among early HCC patients with MVI was 537.13 (40.97–36587.2) mAU/ml, higher than that of patients without MVI 67.08 (4.40–22344.02) mAU/ml (P < 0.01). (b) Difference of AFP plasma levels after logarithmic transformation between early HCC patients with and without MVI. The serum level of PIVKA-II among early HCC patients with MVI was 99.2 (0.9–8720.20) ng/ml, higher than that of patients without MVI 9.15 (1.00–6637.5) ng/ml (P < 0.01). HCC = Hepatocellular carcinoma, PIVKA = Prothrombin induced by Vitamin K absence, MVI = Microvascular invasion, AFP = Alpha-fetoprotein|
Click here to view
|Table 2: Univariate and multivariate analyses of characteristics in all subjects|
Click here to view
According to the ROC curve, PIVKA-II presented a similar diagnostic value with AFP for MVI detection in early HCC patients with the area under the receiver operating characteristic curve (AUROC) of 0.76 (95% CI 0.68–0.85) and 0.72 (95% CI 0.62–0.81), respectively. Moreover, multivariate analysis also showed that tumor size was an independent risk factor (OR = 1.73, 95% CI 1.09–2.74) for MVI detection in early HCC patients, with the AUROC of 0.74 (95% CI 0.65–0.83) derived from ROC curve [Figure 2]. In addition, at a cutoff value of 75.25 mAU/ml, the sensitivity and specificity of PIVKA-II for MVI detection in early HCC were 55.7% and 90.2%. At a cutoff value of 43.35 ng/ml, the sensitivity and specificity of AFP for MVI detection in early HCC were 57.4% and 75.0%. At a cutoff value of 2.65 cm, the sensitivity and specificity of tumor size for MVI detection in early HCC were 56.0% and 97.2%.
|Figure 2: Diagnostic values of PIVKA-II (blue), AFP (green), tumor size (yellow), and the logistic regression model (red) for MVI detection in early HCC patients. PIVKA = Prothrombin induced by Vitamin K absence, MVI = Microvascular invasion, AFP = Alpha-fetoprotein, HCC = Hepatocellular carcinoma|
Click here to view
Development of a logistic regression model to detect microvascular invasion in early hepatocellular carcinoma patients
The baseline variables that considered to be clinically relevant and risk factors for MVI were employed to construct multivariate logistic regression model. Factors with P < 0.2 in the logistic regression model were also employed to conduct a nomogram. The factors integrated in logistic regression were PIVKA and AFP serum levels after logarithmic transformation and tumor size (cm). The final logistic regression model to detect MVI in early HCC was: Logit (P) = (−4.758) + 0.687 (PIVKA) + 0.505 (AFP) + 0.546 (tumor size), the performance of the model was good with AUROC of 0.81 (95% CI: 0.73–0.89) [Figure 3]. At a cutoff probability of 0.376 (Logit [P] = −0.507), the sensitivity and specificity of the model were 75.00% and 88.9%, respectively.
|Figure 3: Nomogram to predict MVI risk in early HCC patients. To use the nomogram, find the score for each variable including PIVKA-II serum levels and AFP serum levels after logarithmic transformation on the corresponding axis, add the points for all variables, and draw a line from the total points' axis to the risk and determine the MVI risk. PIVKA = Prothrombin induced by Vitamin K absence, MVI = Microvascular invasion, AFP = Alpha-fetoprotein, HCC = Hepatocellular carcinoma|
Click here to view
Nomogram construction and internal validation
Identified independent risk factors for MVI detection by logistic regression were integrated to construct a nomogram, which is shown in [Figure 3]. It was formulated based on PIVKA and AFP serum levels after logarithmic transformation and tumor size. High MVI probability is due to high scores based on the sum of the assigned points for each factor in the nomogram. In order to quantify the discriminant performance of standard graphs, the C index of Harrell is calculated. The model showed good discrimination ability, and the C index is 0.81 and good calibration as shown in [Figure 4].
|Figure 4: Calibration curves for the nomogram in estimating the risk of MVI in early HCC patients. The x-axis is the nomogram-predicted probability of MVI in early HCC patients and the y-axis is the actual probability. MVI = Microvascular invasion, HCC = Hepatocellular carcinoma|
Click here to view
| > Discussion|| |
In this study, we sought to assess the performance of PIVKA-II and AFP in the prediction of MVI in early HCC patients. The present study indicated that serum levels of PIVKA-II were significantly higher in early patients with MVI than those without MVI. In line with our findings, a recent study in France demonstrated that high PIVKA-II tissue expression was significantly correlated with MVI development. We found that the AUC of PIVKA-II level for MVI detection in early HCC was 0.76 (95% CI: 0.68–0.85). The current study also displayed that PIVKA-II level >75.25 mAU/ml was an independent risk for MVI detection in early HCC patients. Consistent with our findings, Kaibori et al. demonstrated that PIVKA-II serum level ≥200 mAU/ml was an independent predictor of MVI in HCC patients undergoing potentially curative resection. Several recent studies reveal that elevated PIVKA-II could promote the release of angiogenic factors from HCC and vascular endothelial cell, which could lead to tumor angiogenesis in HCC., Moreover, due to the biological malignant potential of PIVKA-II, it has been described as an autologous growth factor and could integrate HCC with vascular endothelial cells, and it was found to play an important role in the process of angiogenesis. Therefore, it is not intricate that PIVKA-II levels were elevated in early patients with MVI in contrast with HCC patients without MVI and it may play an important role in the development of MVI.
According to our results, AFP serum levels in early HCC patients with MVI were also found significantly higher than those without MVI. In agreement with our results, Mitsuhashi et al. revealed that AFP-positive cases also showed a significantly higher microvessel density than AFP-negative cases, another study also demonstrated that higher VEGF tissue expression and microvessel density were detected in HCC patients with high AFP serum levels (>300 ng/mL). Moreover, we found that higher serum AFP independently correlated with MVI in early HCC patients, which was in line with the results of other researchers.
Tumor size is another independent risk for MVI detection in early HCC according to our results, which is in line with several recent studies. For instance, it was found that large size was an independent preoperative factor predicted by MVI in the hepatitis B virus-related HCC. Another study conducted in Korea also demonstrated that large tumor size (>5 cm) showed statistically significant associations with MVI in HCC patients who are undergoing liver transplantation.
Our study revealed that the AUROC of tumor size to detect MVI in early HCC was 0.74 (95% CI: 0.65–0.83), while the AUC of PIVKA-II serum level and AFP level were 0.76 (95% CI: 0.68–0.85) and 0.72 (95% CI: 0.62–0.81), respectively. We developed a nomogram to detect MVI among early HCC patients based on the above predictors. In the current study, the C-index (0.81), calibration curve, and ROC curve analysis demonstrated that our nomogram was accurate in predicting MVI in early HCC patients. Consistent with our findings, Gao et al. conduct a nomogram combining tumor size (P = 0.002), computerized tomography (CT) value in the delayed phase (P = 0.018), and peritumoral enhancement to predict MVI among BCLC 0/A HCC, and they showed that the nomogram displayed an unadjusted C-index of 0.851. In another recent study, Lin, et al. constructed a nomogram for MVI detection in HCC patients with single nodule combining serum AFP level, with or without intratumoral artery, tumor type classified by MRI, and tumor diameter, which showed good performance with an AUROC of 0.803 (95% CI, 0.746–0.860) and 0.814 (95% CI, 0.720–0.908) in the training and validation groups, respectively. Zhu et al. constructed two nomograms based on AFP and radiological features (contrast-enhanced magnetic resonance imaging), the concordance indexes of which for MVI predictions were 0.810 and 0.799, respectively. In line with the present study, subjects in the above study were also HCC patients with a single nodule (<5.0 cm at diameter). Yan et al. developed a nomogram integrated with hepatitis B virus DNA loading, portal hypertension, BCLC stage, and three CT imaging features, which indicated significant clinical usefulness. In contrast with the above nomograms, we concluded that the nomogram of the current study displayed similar discriminability of MVI prediction but easier to use clinically.
There are some limitations in the current research. For instance, selection bias was unavoidable for that this was a retrospective study carried out in a single center. The number of subjects was relatively small; therefore, more subjects and centers are needed to verify our nomogram.
| > Conclusion|| |
The current preoperative nomogram model could identify early HCC patients at high risk for MVI development and provide assistance in deciding appropriate operative procedures for surgeons to improve clinical outcomes.
Financial support and sponsorship
This work was supported by the Natural Science Foundation of Shandong Province, China (Grant No.ZR2020QH039).
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Yang JD, Hainaut P, Gores GJ, Amadou A, Plymoth A, Roberts LR. A global view of hepatocellular carcinoma: Trends, risk, prevention and management. Nat Rev Gastroenterol Hepatol 2019;16:589-604.
Kang KJ, Ahn KS. Anatomical resection of hepatocellular carcinoma: A critical review of the procedure and its benefits on survival. World J Gastroenterol 2017;23:1139-46.
Du M, Chen L, Zhao J, Tian F, Zeng H, Tan Y, et al
. Microvascular invasion (MVI) is a poorer prognostic predictor for small hepatocellular carcinoma. BMC Cancer 2014;14:38.
Fan ST, Poon RT, Yeung C, Lam CM, Lo CM, Yuen WK, et al
. Outcome after partial hepatectomy for hepatocellular cancer within the Milan criteria. Br J Surg 2011;98:1292-300.
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.
Barreto SG, Brooke-Smith M, Dolan P, Wilson TG, Padbury RT, Chen JW. Cirrhosis and microvascular invasion predict outcomes in hepatocellular carcinoma. ANZ J Surg 2013;83:331-5.
Wang A, Du L, Jiang K, Kong Q, Zhang X, Li L. Long noncoding RNA microvascular invasion in hepatocellular carcinoma is an indicator of poor prognosis and a potential therapeutic target in gastric cancer. J Cancer Res Ther 2019;15:126-31.
Poté N, Cauchy F, Albuquerque M, Voitot H, Belghiti J, Castera L, et al
. Performance of PIVKA-II for early hepatocellular carcinoma diagnosis and prediction of microvascular invasion. J Hepatol 2015;62:848-54.
Yamashita YI, Imai K, Yusa T, Nakao Y, Kitano Y, Nakagawa S, et al
. Microvascular invasion of single small hepatocellular carcinoma ≤3 cm: Predictors and optimal treatments. Ann Gastroenterol Surg 2018;2:197-203.
Witjes CD, Willemssen FE, Verheij J, van der Veer SJ, Hansen BE, Verhoef C, et al
. Histological differentiation grade and microvascular invasion of hepatocellular carcinoma predicted by dynamic contrast-enhanced MRI. J Magn Reson Imaging 2012;36:641-7.
Kim MJ, Lee M, Choi JY, Park YN. Imaging features of small hepatocellular carcinomas with microvascular invasion on gadoxetic acid-enhanced MR imaging. Eur J Radiol 2012;81:2507-12.
Gouw AS, Balabaud C, Kusano H, Todo S, Ichida T, Kojiro M. Markers for microvascular invasion in hepatocellular carcinoma: Where do we stand? Liver Transpl 2011;17 Suppl 2:S72-80.
Marrero JA, Feng Z, Wang Y, Nguyen MH, Befeler AS, Roberts LR, et al
. Alpha-fetoprotein, des-gamma carboxyprothrombin, and lectin-bound alpha-fetoprotein in early hepatocellular carcinoma. Gastroenterology 2009;137:110-8.
Marrero JA, Su GL, Wei W, Emick D, Conjeevaram HS, Fontana RJ, et al
. Des-gamma carboxyprothrombin can differentiate hepatocellular carcinoma from nonmalignant chronic liver disease in American patients. Hepatology 2003;37:1114-21.
Ryu T, Takami Y, Wada Y, Tateishi M, Hara T, Yoshitomi M, et al
. A Clinical scoring system for predicting microvascular invasion in patients with hepatocellular carcinoma within the Milan Criteria. J Gastrointest Surg 2019;23:779-87.
Yamashita Y, Tsuijita E, Takeishi K, Fujiwara M, Kira S, Mori M, et al
. Predictors for microinvasion of small hepatocellular carcinoma 2 cm. Ann Surg Oncol 2012;19:2027-34.
Poté N, Cauchy F, Albuquerque M, Cros J, Soubrane O, Bedossa P, et al
. Contribution of virtual biopsy to the screening of microvascular invasion in hepatocellular carcinoma: A pilot study. Liver Int 2018;38:687-94.
Song DS, Bae SH. Changes of guidelines diagnosing hepatocellular carcinoma during the last ten-year period. Clin Mol Hepatol 2012;18:258-67.
Bruix J, Sherman M, American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: An update. Hepatology 2011;53:1020-2.
Cong WM, Bu H, Chen J, Dong H, Zhu YY, Feng LH, et al
. Practice guidelines for the pathological diagnosis of primary liver cancer: 2015 update. World J Gastroenterol 2016;22:9279-87.
Roayaie S, Blume IN, Thung SN, Guido M, Fiel MI, Hiotis S, et al
. A system of classifying microvascular invasion to predict outcome after resection in patients with hepatocellular carcinoma. Gastroenterology 2009;137:850-5.
Kaibori M, Ishizaki M, Matsui K, Kwon AH. Predictors of microvascular invasion before hepatectomy for hepatocellular carcinoma. J Surg Oncol 2010;102:462-8.
Zhang YS, Chu JH, Cui SX, Song ZY, Qu XJ. Des-γ-carboxy prothrombin (DCP) as a potential autologous growth factor for the development of hepatocellular carcinoma. Cell Physiol Biochem 2014;34:903-15.
Matsubara M, Shiraha H, Kataoka J, Iwamuro M, Horiguchi S, Nishina S, et al
. Des-γ-carboxyl prothrombin is associated with tumor angiogenesis in hepatocellular carcinoma. J Gastroenterol Hepatol 2012;27:1602-8.
Mitsuhashi N, Kobayashi S, Doki T, Kimura F, Shimizu H, Yoshidome H, et al
. Clinical significance of alpha-fetoprotein: Involvement in proliferation, angiogenesis, and apoptosis of hepatocellular carcinoma. J Gastroenterol Hepatol 2008;23:e189-97.
Shan YF, Huang YL, Xie YK, Tan YH, Chen BC, Zhou MT, et al
. Angiogenesis and clinicopathologic characteristics in different hepatocellular carcinoma subtypes defined by EpCAM and α-fetoprotein expression status. Med Oncol 2011;28:1012-6.
Lin S, Ye F, Rong W, Song Y, Wu F, Liu Y, et al
. Nomogram to assist in surgical plan for hepatocellular carcinoma: A prediction model for microvascular invasion. J Gastrointest Surg 2019;23:2372-82.
Lei Z, Li J, Wu D, Xia Y, Wang Q, Si A, et al
. Nomogram for preoperative estimation of microvascular invasion risk in hepatitis B virus-related hepatocellular carcinoma within the Milan Criteria. JAMA Surg 2016;151:356-63.
Ahn SY, Lee JM, Joo I, Lee ES, Lee SJ, Cheon GJ, et al
. Prediction of microvascular invasion of hepatocellular carcinoma using gadoxetic acid-enhanced MR and (18) F-FDG PET/CT. Abdom Imaging 2015;40:843-51.
Gao SX, Liao R, Wang HQ, Liu D, Luo F. A nomogram predicting microvascular invasion risk in BCLC 0/A hepatocellular carcinoma after curative resection. Biomed Res Int 2019;2019:9264137.
Zhu YJ, Feng B, Wang S, Wang LM, Wu JF, Ma XH, et al
. Model-based three-dimensional texture analysis of contrast-enhanced magnetic resonance imaging as a potential tool for preoperative prediction of microvascular invasion in hepatocellular carcinoma. Oncol Lett 2019;18:720-32.
Yan Y, Zhou Q, Zhang M, Liu H, Lin J, Liu Q, et al
. Integrated nomograms for preoperative prediction of microvascular invasion and lymph node metastasis risk in hepatocellular carcinoma patients. Ann Surg Oncol 2020;27:1361-71.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]