Value of Perfusion Parameters and Histogram Analysis of Triphasic Computed Tomography in Pre-Operative Prediction of Histological Grade of Hepatocellular Carcinoma
Hepatocellular carcinoma (HCC) is the fifth most common cancer globally and the third leading cause of cancer-related deaths. Accurate pre-operative assessment of the histological grade of HCC is crucial for clinical decision-making and prognosis. However, non-invasive pre-operative histological evaluation remains a significant challenge. Tumor perfusion, which reflects the hemodynamic changes in HCC, has been shown to correlate with tumor aggressiveness and histological differentiation. This study explores the clinical value of quantitative liver perfusion parameters and histogram analysis derived from traditional triphasic enhanced computed tomography (CT) scans in predicting the histological grade of HCC.
Introduction
HCC is a highly aggressive malignancy with a poor prognosis, particularly in advanced stages. Treatment options include liver transplantation, surgical resection, radiofrequency ablation, transcatheter hepatic arterial chemoembolization, and targeted therapies. Surgical resection is one of the most effective treatments, but high recurrence rates (50%-60% at 3 years and 70%-100% at 5 years) pose significant challenges. The histological grade of HCC is a critical factor in predicting long-term survival and post-operative recurrence. HCCs with poorly differentiated components (P-HCCs) have higher recurrence rates, poorer prognosis, and lower survival rates compared to moderately and well-differentiated HCCs. Therefore, accurate pre-operative prediction of histological grade is essential for optimizing treatment strategies and improving patient outcomes.
Traditionally, histological evaluation of HCC is performed post-operatively through pathological examination of surgical specimens. However, non-invasive pre-operative assessment of tumor grade is highly desirable. Recent studies have explored the use of imaging techniques, including magnetic resonance imaging (MRI) and contrast-enhanced ultrasound, for grading HCC. However, these methods have limitations, such as variability in imaging protocols and limited accuracy in differentiating histological grades.
Perfusion CT (PCT) has emerged as a promising tool for evaluating tumor hemodynamics and perfusion status. PCT provides quantitative parameters that reflect the microcirculation of tumors, which can be used to assess tumor aggressiveness and histological differentiation. However, traditional PCT is associated with high radiation exposure and poor imaging quality, limiting its clinical application. The dual maximum slope model, first proposed by Blomley et al., offers a simplified approach to estimating liver perfusion parameters using standard triphasic CT scans. This model reduces radiation exposure while providing functional information about tumor perfusion.
This study aims to evaluate the clinical utility of liver perfusion parameters and histogram analysis derived from traditional triphasic CT scans in predicting the histological grade of HCC. The study also identifies the optimal perfusion parameters for differentiating between HCCs with poorly differentiated components (P-HCCs) and HCCs without poorly differentiated components (NP-HCCs).
Methods
Patient Selection
A total of 119 consecutive patients suspected of having malignant hepatic lesions were initially enrolled in this retrospective study. All patients underwent triphasic enhanced CT scans and subsequent hepatic resection within 30 days. After applying exclusion criteria, 52 patients with pathologically confirmed HCC were included in the analysis. Exclusion criteria included small arterially enhancing tumor portions (less than 5 mm), previous anti-tumor treatments, a long interval between CT scans and surgery (more than one month), portal vein thrombosis, and more than three concurrent lesions.
CT Imaging Protocol
All patients underwent non-contrast and triphasic contrast-enhanced CT scans using a Discovery 750HD CT scanner. The scanning protocol included arterial phase (30-35 seconds), portal venous phase (60-70 seconds), and delayed phase (180 seconds) after intravenous injection of iodinated contrast. The volumetric CT dose index (CTDIvol) was 24.8 ± 3.2 mGy, significantly lower than traditional perfusion imaging.
Perfusion Parameters Measurement
Tumor regions of interest (ROIs) were drawn on three or four representative slices to calculate perfusion parameters. CT hemodynamic kinetics software was used to measure hepatic arterial supply perfusion (HAP), portal vein blood supply perfusion (PVP), and arterial enhancement fraction (AEF). Perfusion parameters included total hepatic blood flow (HFtumor and HFliver), difference in flow between tumor and liver (DHF = HFtumor – HFliver), relative flow (rHF = DHF/HFliver), difference in HAP (DHAP = HAPtumor – HAPliver), relative HAP (rHAP = DHAP/HAPliver), difference in PVP (DPVP = PVPtumor – PVPliver), relative PVP (rPVP = DPVP/PVPliver), difference in AEF (DAEF = AEFtumor – AEFliver), and relative AEF (rAEF = DAEF/AEFliver). Histogram analysis was performed to assess tumor heterogeneity, including median, mean, standard deviation, variance, skewness, and kurtosis.
Statistical Analysis
Statistical analyses were performed using R software. Inter-observer agreement was assessed using weighted k statistics. The Mann-Whitney U test was used for non-normally distributed data. Receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cutoff values for predicting histological grade. A P-value less than 0.05 was considered statistically significant.
Results
Clinical and Pathological Characteristics
The study included 52 patients with HCC, comprising 36 males and 16 females, with a mean age of 52.8 years. Based on Child-Pugh classification, 38 patients were in class A, and 14 were in class B/C. Hepatitis virus markers were positive in 45 patients, and seven patients had a significant alcohol history. The mean tumor diameter was 11.4 ± 7.8 mm. Pathological evaluation classified the tumors as poorly differentiated (n=16), moderately differentiated (n=25), and well-differentiated (n=11). No significant differences were observed in clinical characteristics between P-HCC and NP-HCC groups.
Perfusion Parameters and Histogram Analysis
The study found significant differences in perfusion parameters between P-HCC and NP-HCC groups. The difference in total flow between tumor and liver (DHF) and relative flow (rHF) were significantly higher in NP-HCCs than in P-HCCs. Similarly, the difference in PVP between tumor and liver (DPVP) and the relative PVP (rPVP) were significantly higher in NP-HCCs. The variance of AEF was also higher in NP-HCCs compared to P-HCCs.
Predictive Ability of Perfusion Parameters
ROC analysis demonstrated that DPVP and rPVP had the highest area under the curve (AUC) of 0.697, with a sensitivity of 84.2% and specificity of 56.2%. DHF and rHF had higher specificity (87.5%) with AUC values of 0.681 and 0.673, respectively. The combined parameter of rHF and rPVP showed the highest AUC of 0.732, with a sensitivity of 57.9% and specificity of 93.8%. The combined parameter of DHF and rPVP had the highest positive predictive value (PPV) of 0.903, while the combined parameter of rPVP and DPVP had the highest negative predictive value (NPV) of 0.781.
Discussion
This study highlights the clinical utility of liver perfusion parameters and histogram analysis derived from traditional triphasic CT scans in predicting the histological grade of HCC. The findings demonstrate that quantitative perfusion parameters, including DHF, rHF, DPVP, rPVP, and AEF variance, can effectively differentiate between P-HCC and NP-HCC. The combined parameter of rHF and rPVP showed the highest predictive power, with an AUC of 0.732, suggesting its potential as a non-invasive biomarker for pre-operative histological grading.
The higher values of DHF and rHF in NP-HCCs may reflect increased arterial blood supply in well and moderately differentiated HCCs, while P-HCCs exhibit decreased arterial supply. Similarly, the higher values of DPVP and rPVP in NP-HCCs indicate a more prominent portal venous blood supply compared to P-HCCs. The variance of AEF, which reflects tumor heterogeneity, was also higher in NP-HCCs, further supporting its role in histological grading.
The study’s limitations include a relatively small sample size and retrospective design, which may introduce selection bias. Additionally, the use of ROIs in a single plane rather than volumetric analysis may limit the accuracy of perfusion parameter measurements. Future studies with larger cohorts and advanced imaging techniques are needed to validate these findings and improve the predictive value of perfusion parameters.
Conclusion
Liver perfusion parameters and histogram analysis derived from traditional triphasic CT scans provide a non-invasive method for predicting the histological grade of HCC. The combined parameter of rHF and rPVP demonstrated the highest predictive power, offering a potential tool for pre-operative assessment of tumor differentiation. These findings underscore the importance of functional imaging in optimizing treatment strategies and improving outcomes for patients with HCC.
doi.org/10.1097/CM9.0000000000001446
Was this helpful?
0 / 0