Clinical and DCE-CT signs in predicting microvascular invasion in cHCC-ICC Authors: Zhong-Jian Liao, Lun Lu, Yi-Ping Liu, Geng-geng Qin, Cun-geng Fan, Yan-Ping Liu, Ning-yang Jia, Ling Zhang Published: 2023 ABSTRACT AbstractBackgroundTo predict the microvascular invasion (MVI) in patients with cHCC-ICC.MethodsA retrospective analysis was conducted on 119 patients who underwent CT enhancement scanning (from September 2006 to August 2022). They were divided into MVI-positive and MVI-negative groups.ResultsThe proportion of patients with CEA elevation was higher in the MVI-positive group than in the MVI-negative group, with a statistically significant difference (P = 0.02). The MVI-positive group had a higher rate of peritumoral enhancement in the arterial phase (P = 0.01) whereas the MVI-negative group had more oval and lobulated masses (P = 0.04). According to the multivariate analysis, the increase in CEA (OR = 10.15, 95% CI: 1.11, 92.48, p = 0.04), hepatic capsular withdrawal (OR = 4.55, 95% CI: 1.44, 14.34, p = 0.01) and peritumoral enhancement (OR = 6.34, 95% CI: 2.18, 18.40, p < 0.01) are independent risk factors for predicting MVI. When these three imaging signs are combined, the specificity of MVI prediction was 70.59% (series connection), and the sensitivity was 100% (parallel connection).ConclusionsOur multivariate analysis found that CEA elevation, liver capsule depression, and arterial phase peritumoral enhancement were independent risk factors for predicting MVI in cHCC-ICC.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40644-023-00621-3.Keywords: Combined hepatocellular intrahepatic cholangiocarcinoma, Microvascular invasion, Computerized tomography Background Combined hepatocellular intrahepatic cholangiocarcinoma (cHCC-ICC) is a rare subtype of primary liver cancer with an incidence of 0.4 − 14.2% [1]. The tumor contains components of hepatocellular carcinoma and cholangiocarcinoma, which may originate from the same kind of liver progenitor cells and differentiate in different directions [2]. However, some studies have shown that mature hepatocytes can dedifferentiate into progenitor cells under certain inducements, thus transforming into cHCC-ICC [3]. Prognostic factors affecting cHCC-ICC are similar to those affecting HCC, including tumor size, multiple lesions, portal vein invasion, microvascular invasion [4], stage of BCLC, and treatment options such as operation type, TARE, and efficacy of systemic immunotherapy [4–8]. MVI is a part of the pathological results and is difficult to be identified by imaging inspection. The imaging signs that can prompt MVI have been analyzed and reported in numerous studies. It is hoped that MVI can be predicted before surgery to help clinicians develop reasonable and informative surgical plans. However, MVI studies mainly are focused on hepatocellular carcinoma (HCC), a more common subtype of liver cancer [9–12]. Only Wang et al. have reported MRI signs for predicting MVI in cHCC-ICC [13, 14]. No study has been found to predict MVI in cHCC-ICC with CT signs. Due to the complexity of the imaging signs of cHCC-ICC, we intended to make a preliminary exploration in this regard. Data and methods Clinical data Data of 151 cases of cHCC-ICC from patients who had received dynamic enhanced CT scanning in three hospitals from September 2006 to August 2022 were retrospectively collected. Inclusion criteria: (1) Age 1880 years old; (2) CT indicates liver tumor 1 cm; (3) Completed the surgical resection of liver tumor, which had been confirmed as cHCC-ICC by postoperative pathology. Exclusion criteria: (1) Age 0 18 years old or over 80 years old; (2) Poor quality of CT scan image or the phase was incomplete; (3) Those who did not receive resection of the liver tumor but only underwent puncture; (4) Incomplete laboratory test data; (5) Those who had received other treatments before the operation. Data of levels of alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), carbohydrate antigen 199 (CA-199), alanine aminotransferase (ALT), aspartate aminotransferase (AST), PLT, total bilirubin (TBIL), and γ-glutamyl transpeptidase (GGT), as well as quantitative analysis results of hepatitis B virus were acquired. This was a retrospective study approved by the hospital ethics committee. All images and clinical data were acquired from the PACS system; thus, informed consent was exempted. CT examination equipment 1 Siemens 64 slice CT scanner, scanning parameters: tube voltage 120 kV, tube current automatic mA regulation technology, layer thickness 5 mm; Reconstruction parameters: pitch 1, reconstruction layer thickness 1 mm. The window width was 250 Hu, and the window level was 50 Hu. The tracking monitoring started 10 s after the contrast medium was injected. When the CT value of the monitoring layer reached the trigger threshold, the early arterial phase was acquired at 6 s delay and the late arterial phase was obtained at 8 s delay. The portal vein phase was captured 60 s after injection of the intravenous contrast medium, and the delayed image was obtained 120 s after injecting the contrast medium. The images were transmitted to PACS in DICOM format. Equipment 2: GE 64 slice CT scanner (Biograph m CT). Parameters: layer thickness 6 mm, voltage 120 kV, matrix 512 × 512. The hepatic artery phase was 20 ~ 30 s, the portal vein phase was 50 ~ 60 s, the delay phase was 180 ~ 200 s, the window width was 250 ~ 300 HU, and the window level was 40 ~ 80 HU. Analysis of CT signs Two radiologists who have been engaged in abdominal imaging diagnosis for more than 15 years analyzed the signs of all cHCC-ICC cases in the PACS system, including whether there was atrophy of the liver lobe, depression of the liver capsule, dilation of the surrounding bile duct, size, edge, shape of the focus, the presence or absence of capsule and lipid inside, CT density, tumor thrombus in the portal vein, enhancement methods in each phase of enhanced scanning, whether there was a peritumoral enhancement in the arterial phase. The evaluators had no prior knowledge of patients’ clinical data. In case of doubt, consultation would be initiated to reach an agreement. Pathological analysis Microvascular invasion (MVI) refers to the microscopic observation of cancer cell nests in the vascular lumen lined with endothelial cells, mainly portal vein or hepatic vein branches (including intrathecal vessels). Pathological grading method: M0: no MVI was found; M1 (low-risk group): ≤ 5 MVI, and it occurred in the liver tissue near the cancer; M2 (high-risk group): >5 MVI, or MVI occurring in the liver tissue near the distant cancer. After reading the images, it was found that there were only a few M2 cases in this group, and thus M1 and M2 were combined into the MVI-positive group. An experienced pathologist (who has 16-year experience in the pathological diagnosis of liver cancer) was asked to review the sections of all cases and evaluate whether cirrhosis was present. The relationship between CT findings and histopathology was jointly determined by radiologists and pathologists. Specimens were fixed with 4% neutral buffered formaldehyde and embedded in conventional paraffin µM serial sections, routine HE staining and immunohistochemical staining. DAB (3,3’-diaminobenzidine) was used in immunohistochemical (IHC) staining. Hematoxylin was applied for re-staining, prior to which the slides were hydrated. Blue flower was dehydrated and sealed. Factor VIII related antigen (F - VIII Ag) was used to display the microvessels in the lesions. Results Clinical characteristics See Table 1 for details. After screening, there were 119 cases of cHCC-ICC with complete CT plain scan and enhancement data and confirmed by surgery and pathology, which met the inclusion criteria, with a mean age of 52.41 ± 10.36, and a male-to-female ratio of 105:14. There was no significant difference between the MVI-positive group and negative group in age, sex, tumor size, lymph node metastasis, hepatitis B and cirrhosis (P > 0.05). Laboratory examination: The proportion of patients with CEA elevation was higher in the MVI-positive group than in the MVI-negative group, with a statistically significant difference (P < 0.05). There was no significant difference in AFP, CA-199, ALT, AST, PLT, TBIL, or r-GT between the MVI-positive group and the MVI-negative group of cHCC-ICC (P > 0.05). Table 1Comparison of patient characteristics according to MVICharacteristicsTotalMVI N(%)StatisticP valuemvi = 0mvi = 1Total11968(57.14)51(42.86)Age52.41 ± 10.3654.35 ± 9.1049.80 ± 11.440.820.41sex0.330.57male10559(56.19)46(43.81)..female149(64.29)5(35.71)..AFP.2.880.0905536(65.45)19(34.55)..16432(50.00)32(50.00)..CEA.5.29 0.02 010549(46.67)56(53.33)..1142(14.29)12(85.71)..CA-199.2.640.1007941(51.90)38(48.10)..14027(67.50)13(32.50)..Hepatitis B.1.110.2903422(64.71)12(35.29)..18546(54.12)39(45.88)..Cirrhosis.0.640.4305433(61.11)21(38.89)16535(53.85)30(46.15)ALT.2.680.10Negative9752(53.61)45(46.39)Positive2216(72.73)6(27.27)AST.0.000.95Negative9353(56.99)40(43.01)Positive2615(57.69)11(42.31)PLT.0.330.57Negative10561(58.10)44(41.90)Positive147(50.00)7(50.00)TBIL.0.940.33Negative9858(59.18)40(40.82)Positive2110(47.62)11(52.38)r-GT.0.010.92Negative6034(56.67)26(43.33)Positive5934(57.63)25(42.37)Lymphatic metastasis.0.760.38Negative9154(59.34)37(40.66)Positive2814(50.00)14(50.00)Maximum diameter.0.100.75≤ 5 cm6134(55.74)27(44.26)>5 cm5834(58.62)24(41.38)capsular.0.000.9606839(57.35)29(42.65)15129(56.86)22(43.14)Venous tumor thrombus.1.490.2208451(60.71)33(39.29)13517(48.57)18(51.43)Density.0.380.54Uniform4125(60.98)16(39.02)Uneven7843(55.13)35(44.87)Shape.8.26 0.04 round158(53.33)7(46.67)oval1812(66.67)6(33.33)lobulate3928(71.79)11(28.21)Irregular4720(42.55)27(57.45)Margin.0.220.64smooth1610(62.50)6(37.50)Non-smooth10358(56.31)45(43.69)Lipids.2.710.10011768(58.12)49(41.88)120(0.00)2(100.00)Hepatic atrophy.0.010.94010359(57.28)44(42.72)1169(56.25)7(43.75)Hepatic capsular retraction.0.290.5908849(55.68)39(44.32)13119(61.29)12(38.71)Intrahepatic duct dilatation.0.150.7009354(58.06)39(41.94)12614(53.85)12(46.15)schedule of reinforcement.0.120.94Ring strengthening4022(55.00)18(45.00)Uniform7443(58.11)31(41.89)Uneven53(60.00)2(40.00)TIC curve.1.030.31centripetal reinforcement6741(61.19)26(38.81)Wash in-wash out5227(51.92)25(48.08)Peritumoral enhancement.5.95 0.01 03526(74.29)9(25.71)18442(50.00)42(50.00) Comparison of patient characteristics according to MVI CT plain scan and dynamic enhanced scan characteristics See Table 1 for details. The MVI-positive group of 119 patients with cHCC-ICC was significantly different from the MVI-negative group in peripheral enhancement and tumor shape at the arterial phase (P < 0.05). MVI-positive group had a higher rate of peritumoral enhancement (Figs. 1 and 2) in the arterial phase whereas MVI-negative group had more oval and lobulated masses (P = 0.04). No statistical differences were observed between the two groups in terms of imaging signs such as capsule, intravenous tumor thrombus, tumor density, margin, internal lipid composition, hepatic affinity, hepatic capsular regression, intravascular duct differentiation, schedule of reinforcement, and dynamic enhancement curve. cHCC-ICC presenting with MVI in a 48-year-old man. (A) A plain CT scan showed an irregular low-density mass; (B-D) Early enhancement of the mass was observed during the arterial phase of CT enhancement, followed by rapid clearance of the contrast agent from the mass. Hepatic capsule retraction (green arrow) and peritumoral enhancement (red arrow) were observed cHCC-ICC presenting with MVI in a 53-year-old woman. (A) A plain CT scan revealed a large low-density mass in the right lobe of the liver, capsule retraction can be seen around the mass (green arrow). (B-D) The arterial phase of the enhanced CT scan showed significantly uneven enhancement and peritumoral enhancement (red arrow) Risk factor analysis: See Table 2 for details. According to the multivariate analysis, the increase in CEA (OR = 10.15, 95% CI: 1.11, 92.48, p = 0.04), hepatic capsular withdrawal (OR = 4.55, 95% CI: 1.44, 14.34, p = 0.01) and peritumoral enhancement (OR = 6.34, 95% CI: 2.18, 18.40, p < 0.01) are independent risk factors for predicting MVI. Table 2Univariate and multivariate analyses of preoperative CT imaging findings in predicting MVIUnivariate analysis(MVI)multivariate analyses(MVI)OR(95%CI) P OR(95%CI) P AgeSex0.71(0.22,2.27)0.57AFP1.89(0.90,3.97)0.09CEA5.25(1.12,24.6) 0.04 10.15(1.11,92.48) 0.04 CA-1990.52(0.23,1.15)0.11Hepatitis B1.55(0.68,3.54)0.29Cirrhosis1.35(0.65,2.80)0.43ALT0.43(0.16,1.20)0.11AST0.97(0.40,2.34)0.95PLT1.39(0.45,4.24)0.57TBIL1.59(0.62,4.11)0.33r-GT0.96(0.47,1.99)0.92Lymphatic metastasis1.46(0.62,3.42)0.38Maximum diameter0.89(0.43,1.84)0.75capsular1.02(0.49,2.12)0.96Venous tumor thrombus1.64(0.74,3.62)0.22Density1.27(0.59,2.75)0.54Shape(ref: round)oval0.57(0.14,2.34)0.44lobulated0.45(0.13,1.54)0.20irregular1.54(0.48,4.96)0.47Margin1.29(0.44,3.82)0.64Hepatic atrophy1.04(0.36,3.02)0.94Hepatic capsular retraction1.26(0.55,2.91)0.594.55(1.44,14.34) 0.01 Intrahepatic duct dilatation1.19(0.50,2.84)0.70schedule of reinforcement(ref: Ring strengthening)Uniform0.88(0.41,1.91)0.75Uneven0.81(0.12,5.42)0.83TIC curve1.46(0.70,3.04)0.31Peritumoral enhancement2.89(1.21,6.9)0.026.34(2.18,18.40) < 0.01 Univariate and multivariate analyses of preoperative CT imaging findings in predicting MVI schedule of reinforcement (ref: Ring strengthening) The three imaging characteristics that have the value of predicting MVI - CEA elevation, sensitivity, and specificity, i.e., PPV and NPV of the hepatic capsular withdrawal and peritumoral enhancement, as well as their combination to predict MVI are shown in Table 3. When these three imaging signs are combined, the specificity of MVI prediction was 70.59% (series connection), and the sensitivity was 100% (parallel connection). Table 3Diagnostic performance of clinical and CT Characteristics in prediction of MVICharacteristicsSensitivitySpecificityPPVNPVCEA96.08%17.65%46.67%85.71%Hepatic capsular retraction76.47%27.94%44.32%61.29%Peritumoral enhancement82.35%38.24%50.00%74.29%Combination of three findings(series connection)54.90%70.59%58.33%67.61%Combination of three findings(parallel connection)100.00%1.47%43.22%100.00% Diagnostic performance of clinical and CT Characteristics in prediction of MVI Discussion Predicting MVI in patients with primary liver cancer has been a hot research topic in the past decade because it is closely related to postoperative recurrence and metastasis and affects disease-free survival and overall survival time of patients. High-risk factors related to MVI in HCC include tumor diameter, number, AFP level, arterial phase annular enhancement, arterial phase peritumoral enhancement, etc. [14–16] Whether MVI plays a key role in the prognosis of patients with ICC is still unclear. Studies that are focused on MVI in ICC are rarely seen and only a few reports are comprehensive [17, 18]. Age, γ-glutamyl transpeptidase, preoperative tumor number, larger tumor diameter, and intrahepatic bile duct dilatation may be independent risk factors for predicting MVI in ICC. CHCC-ICC has the tumor tissue components of both HCC and ICC, so its imaging signs are complex and changeable. When HCC was the main component, the enhancement mode of enhanced scanning was similar to that of HCC, with obvious enhancement in the arterial phase and rapid clearance in the venous phase and delayed phase; With ICC as the main component, the arterial phase showed marginal enhancement, and the venous phase and delayed phase showed progressive centripetal enhancement [19, 20]. A few research reports on preoperative MVI prediction of cHCC-ICC [13, 14] found that high AFP level (> 400 ng/mL), arterial phase peritumoral enhancement, multiple nodules, γ-glutamyl transpeptidase could be independent risk factors for predicting MVI. However, the risk factors reported in those studies are not completely the same. In addition, when there is fat deposition in the tumor, it is a protective factor for MVI, which is similar to other studies that predict MVI in HCC [21, 22]. The prognosis of HCC is better when there is fat in the tumor on MRI, with longer disease-free survival and total survival. As we all know, CEA is a tumor marker for colorectal cancer and other adenocarcinomas, and has recently attracted much attention as a potential marker for ICC: Compared with moderately differentiated ICC, CEA increases are more often seen in poorly differentiated ICC, and the 1-year and 5-year survival of patients with CEA increases are shorter [23, 24]. This suggests that it is a tumor marker of poor prognosis in ICC. Our data show that the increase of CEA is positively correlated with the positive MVI in cHCC-ICC, but whether there is a causal relationship between these two indicators is unknown. Hepatic capsular retraction refers to the localized depression of the smooth envelope of the liver, which was once considered to be related to liver malignancy. Many benign lesions and treatments can cause hepatic capsular retraction [25]. It may indicate the potential process of liver fibrosis or focal atrophy of liver parenchyma caused by vascular injury/chronic biliary inflammation and may explain the cHCC-ICC in the more aggressive MVI-positive group, which is more likely to lead to vascular and bile duct injury around the focus. Another reason may be that the proportion of our cases located in the periphery of the liver was too high. It has been supported by a large number of literature [10, 26–28] that MVI in HCC can be predicted by the arterial phase peritumoral enhancement. It has also been confirmed that it can be used to predict the MVI in cHCC-ICC, a rare subtype of primary liver cancer. Due to the rarity of this tumor, its clinicopathological features are still not well understood. The surgical resection plan, indication for liver transplantation, and the corresponding adjuvant treatment plan are still under exploration. More clinical and imaging studies on cHCC-ICC will be conducted in the future. We can do a better job in using CT, MRI, and other imaging techniques to predict the efficacy of various treatment regimens and forecast the disease-free survival and overall survival of patients with this disease. Limitations: Our cases were from three hospitals and the CT machines used for scanning were different from the scanning parameters, which may lead to inconsistent interpretation of some image signs. In addition, we did not extract the image omics features in this study but used manual film reading, which would have certain limitations and subjectivity. In the next step, we will add the image omics features to predict MVI before surgery and increase the sample size to make the prediction model more accurate. Besides, our research used CT as an imaging tool, whose effect of observing lipid components with small structures in the tumor is not as good as MRI. We will collect more cases of cHCC-ICC undergoing MRI examination for an accurate preoperative diagnosis and predict its therapeutic efficacy and survival. Conclusion Our multivariate analysis found that CEA elevation, liver capsule depression, and arterial phase peritumoral enhancement were independent risk factors for predicting MVI in cHCC-ICC. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Acknowledgements We would like to thank the whole study team at our Hospitals, for continuous support. Abbreviations Combined hepatocellular intrahepatic cholangiocarcinoma microvascular invasion CEA:Carcinoembryonic antigen alpha-fetoprotein DCE:dynamic contrast enhanced Author’s contribution ZJL, LL, and YPL participated in the study design, evaluated the results, and wrote the first and revised manuscript. CGF and YPL participated in the study design and provided the contrast medium. LZ, GGQ, and NYJ carried out the image analysis and revised manuscripts. LZ participated in the design of the study and redesigned the data analysis. All authors read and approved the final manuscript. Funding This work was supported by the Natural Science Foundation of Guangdong Province (2018A030313951), the science and technology plan project of Ganzhou (2022–RC1339, 2022–RC1349). Data Availability Data to replicate findings are in the Figures and Tables of the main paper. Due to patient privacy protection, any additional materials of the study are only available upon individual request directed to the corresponding author. Declarations Ethics approval and consent to participation The retrospective study was approved by the ethics committee of Ganzhou People’s Hospital, and the requirement for signed informed consent was waived. We confirm that all methods were performed in accordance with the relevant guidelines. Consent for publishing Not applicable. Competing interests The authors declare that there is no conflict of interest regarding the publication of this paper. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Zhong-Jian Liao, Lun Lu and Yi-Ping Liu contributed equally to this work. Contributor Information Ning-yang Jia, Email: jianingyang6@163.com. Ling Zhang, Email: 41733348@qq.com. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement Data to replicate findings are in the Figures and Tables of the main paper. Due to patient privacy protection, any additional materials of the study are only available upon individual request directed to the corresponding author.