A Novel Scoring System for Patients with Recurrence of Hepatocellular Carcinoma After Undergoing Minimal Invasive Therapies

Introduction

Hepatocellular carcinoma (HCC), as the sixth most common malignancy and the fourth tumor-related death worldwide, is a global health problem, especially in developing countries.¹ In today’s clinical practice, the higher recurrence rate is one of the leading causes of mortality, negatively influencing the long-term prognosis of patients with HCC.^2,3 Thus, monitoring for recurrence and providing timely treatment might improve the prognosis of patients with HCC.⁴

Minimal invasive therapies such as local ablation coupled with transcatheter arterial chemoembolization (TACE) have been an alternative treatment for patients with poor liver function, fewer liver donors, and bad economic conditions.^5,6 This could be explained as follows: (1) TACE can label a tumor and cut off its blood supply by blocking the hepatic artery, leaving the tumor starved; (2) TACE can enhance the therapeutic potential of locoregional ablation.^7–9

Over the last few years, numerous scholars have devoted themselves to studying some markers that are associated with the early recurrence of patients with HCC after TACE or curative ablation therapy, including tumor number, fibrinogen (Fib), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and so on.^10,11 Nevertheless, relevant studies on combined therapies are lacking.

Therefore, the objective of this study was to establish a novel scoring system based on clinicopathological data of patients who underwent combined therapies as the initial treatment. It was assumed that the system might effectively assess the risk of recurrence and help clinicians adjust the follow-up strategy, administer timely anti-tumor treatment to patients, and eventually improve their prognosis.

Materials and Methods

Subjects of Study

From January 1, 2015, to December 31, 2016, a total of 305 subjects who were newly diagnosed with HCC were initially treated with combined therapies in the Beijing You’an Hospital. The diagnosis of HCC was completed based on the radiological/histological criteria proposed by the guidelines of the American Association for the Study of Liver Disease.¹² This study’s inclusion criteria were as follows: (1) Age from 18 to 75 years, (2) TACE combined with locoregional ablation as the initial anticancer treatment, and (3) no other malignancies that may affect the prognosis. The exclusion criteria were as follows: (1) Radiological evidence of invasion into the major portal/hepatic vein branches, (2) the presence of extrahepatic metastases, (3) severe coagulation disturbance, and (4) secondary liver cancer. The study protocol, conformed to the 1990 Declaration of Helsinki, was approved by the Ethics Committee of the Beijing You’an Hospital.

Data collected were as follows: demographic data (age, sex, history of hypertension, and diabetes mellitus), etiologies of HCC [hepatitis B virus (HBV), hepatitis C virus (HCV), and alcoholic liver disease], tumor-related indices [number of tumors, size of tumors, and alpha-fetoprotein (AFP) level], liver function indices [cirrhosis, Child-Pugh class, and alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin, and gamma-glutamyl transferase (γ-GT) levels], routine blood examinations [(neutrophil count, lymphocyte count, platelet count, and Fib and albumin-to-prealbumin ratio (APR) levels)], ablation modalities (radiofrequency ablation, microwave ablation, and argon-helium cryoablation).

TACE Procedure

The procedure was performed by three qualified hepatologists. A microcatheter was placed approximately in the tumor-feeding artery by the selective/superselective technique. A mixture of doxorubicin (Pfizer Inc., NY, USA) and lipiodol (Guerbet, Villepinte, France) was injected, and Gelfoam was used for embolization. Occlusions of the feeding artery and disappearance of the vessel stain were identified as the endpoint of embolization.

Ablation Procedure

Local ablation was performed by qualified hepatologists within 2 weeks after TACE. With the guidance of computed tomography (CT) or magnetic resonance imaging (MRI), the ablative position and modality were determined. Then anesthesia was injected at the puncture site, and the ablation needle was inserted into the skin. Multiple sites, overlapping ablation, and repeated ablation were considered according to the tumor number and size to achieve complete ablation. A safety margin of 0.5–1.0 cm of the adjacent non-neoplastic tissue was ablated to ensure complete coverage.

Follow-Up

All patients were followed-up at the outpatient clinic. Abdominal CT or MRI was carried out 4–6 weeks after the treatment to determine whether the ablation was complete. A salvage treatment was required until the ablation was complete. The follow-up involved the standard physical and blood examination every 3 months, ultrasonography every 3–6 months, and CT/MRI every 6 months.

If the CT/MRI displayed an enhancing area within, adjacent to, or out of the original neoplasms, recurrence would be recognized. The recurrence-free survival (RFS) was delineated as the time from the termination of treatments to the first record of a detectable relapse or the date of mortality without evidence of HCC-related recurrence. The overall survival (OS) was defined as the time from the termination of treatments to the date of HCC-related mortality or the data of the last follow-up in which the expiration date in the present study was April 1, 2019. Individuals were treated again with TACE or ablation when recurrence was confirmed. The data of patients were kept confidential. In the present study with minimum risk, the requirement for informed consent was waived because it was difficult to reconnect with the patient.

Statistical Analysis

Continuous and categorical variables, expressed as the mean ± standard deviation and frequency, were compared by the Student t-test and Chi-square test, respectively. The baseline γ-GT was compared by the Mann–Whitney U-test. The univariate and multivariate Cox proportional hazard model was used to identify independent risk factors of recurrence in patients with HCC receiving combined therapies. The optimal cutoff values were calculated in accordance with the time-dependent receiver operating characteristic curves and Youden’s index. The recurrence and survival curves among the three groups with different risks were generated by the Kaplan–Meier estimator and were compared by the log rank test. The prediction value of the scoring system was determined by the area under the curve (AUC). A P value less than 0.05 was regarded as statistically significant. All the data were processed by SPSS 25.0 (IBM, NY, USA).

Results

Patient Characteristics and Follow-Up results

A total of 71 subjects were excluded from the present study in line with the inclusion and exclusion criteria. The selection method is shown in Figure 1 in detail. Eventually, the clinicopathological data of 234 patients with HCC treated with combined therapies as the initial treatment were retrospectively analyzed. The present study included 182 male (77.8%) and 52 female (22.2%) patients. In addition, 64 patients (27.4%) had high blood pressure and 50 (21.4%) had type 2 diabetes mellitus. Further, 97 patients (41.5%) had a history of smoking and 75 (32.1%) had a history of drinking. For etiologies, 176 (75.2%) patients had HBV-related HCC, 43 (18.4%) had HCV-related HCC, and 15 (6.4%) had alcohol-related HCC (Table 1). The median follow-up was 38.2 months (25~75th percentiles, 36.0~43.3 months). A total of 143 (61.1%) patients were diagnosed with tumor recurrence, and 31 (13.2%) died of HCC. The cumulative recurrence rate of 1, 2, and 3 years was 24.4% (57/234), 48.7% (114/234), and 59.0% (138/234), and the corresponding OS rate was 98.3% (230/234), 94.0% (220/234), and 88.5% (207/234), respectively.

Table 1 Demographic and Clinicopathological Data of HCC Patients

Figure 1 Patients selection process.

Prognostic Factors Associated with HCC Recurrence

Based on whether relapses occurred within 2 years or not, all patients were divided into recurrence and non-recurrence groups. Subsequently, demographic and clinicopathological data of these two groups were compared. The results showed a statistical difference in some variables including the history of smoking, Child-Pugh class, size and number of tumors, AFP, γ-GT, Fib, BUN, and APR (Table 2). The variables with the P value less than 0.1 in Table 2 were analyzed by univariate and multivariate Cox regression analysis for identifying variables associated significantly with a poor RFS rate. Finally, the findings of this study suggested that gender, number of tumors, AFP, Fib, and APR were independent risk factors of HCC recurrence (Table 3).

Table 2 The Comparison of Clinical Data Between Recurrence/Non-Recurrence Group

Table 3 Prognostic Factors for RFS by Cox Proportional Hazards Regression Model

Establishment of the Scoring System

Each variable selected by multivariate analysis in Table 3 was assigned different scores according to their HR value. Men were assigned a score of 2; patients with multiple tumors were assigned a score of 2; patients with different AFP levels (<7 ng/mL, 7–400 ng/mL, and ≧400 ng/mL) were assigned scores of 0, 1, and 2, respectively; patients with different Fib levels (<3.105 mg/dL and ≧3.105 mg/dL) were assigned scores of 0 and 1, respectively; and patients with different APR levels (<0.250 and ≧0.250) were assigned scores of 0 and 4, respectively (Table 4). In conclusion, a scoring system was obtained, which ranged from 0 to 11 by calculating each patient’s score. Individuals with scores of 0–3 were defined to be at low recurrence risk, 4–7 at immediate recurrence risk, and 8–11 at high recurrence risk. The time-dependent AUC of 1 year, 2 years, and 3 years was 0.680 (95% CI: 0.604–0.756), 0.728 (95% CI: 0.664–0.793), and 0.709 (95% CI: 0.641–0.776), respectively (P < 0.0001).

Table 4 The Assignment of Variables According HR

Comparison of Clinicopathological Data Among the Three Groups with Different Recurrence Risks

The comparison of demographic and clinicopathological data revealed that some variables, such as history of smoking, Child-Pugh class B, high ALT, high AST, large size of tumors, and so on, representing the poor functional status of liver were significantly associated with a higher recurrence risk (Table 5). Especially, a higher γ-GT level and lower BUN and TC levels indicated a higher recurrence risk (Figure 2). And then, grouped by cut-off values of these three variables, the differences in the recurrence rates between the groups were statistically significant (Figure 3).

Table 5 The Comparison of Clinical Data Among Groups with Different Recurrence Risk

Figure 2 The association between γ-GT, BUN, and TC and different recurrence risk. Abbreviations: γ-GT, gamma-glutamyltransferase; BUN, blood urea nitrogen; TC, total cholesterol.

Figure 3 The cumulative recurrence risk in different levels of γ-GT, BUN, and TC. Abbreviations: γ-GT, gamma-glutamyltransferase; BUN, blood urea nitrogen; TC, total cholesterol.

The 1-, 2-, and 3-year cumulative RFS rates of the low-risk group were 4%, 12.2%, and 30.6%, while those of the intermediate-risk group and high-risk group were 23.4%, 51.6%, 60.0% and 47.3%, 78.2%, 83.6%, respectively. Differences between groups were statistically significant. The 1-, 2-, and 3-year cumulative OS rates of the low-risk group were 99.0%, 99.0%, and 99.0%, while those of the intermediate-risk group and high-risk group were 100%, 95.2%, 91.1%, and 98.2%, 89.1%, 74.5%, respectively. A statistically significant difference was also observed between the groups (Figure 4).

Figure 4 Kaplan-Meier survival curve of RFS and OS for patients with different recurrence risk. Abbreviations: RFS, recurrence-free survival; OS, overall survival.

Discussion

The present study indicated that the 3-year OS rate of patients with HCC after receiving percutaneous ablation coupled with TACE reached 88.5%, suggesting that TACE is one of the effective alternative treatments for HCC. However, the therapeutic effect and quality of life of patients are influenced by the higher recurrence rate.^2,3 The absence of accurate markers might deprive individuals with a high risk of relapse of access to timely interventions. Thus, valuable markers or models should be developed to improve prognosis of patients with HCC.

Many scholars are gradually focusing on demographic and laboratory data, which are secure, repeatable, and easily available for predicting the recurrence of patients with HCC receiving only TACE or ablation therapy. Various studies showed that several predictors might indicate poor RFS, some of which are as follows: (1) noninvasive fibrosis markers such as aspartate APR index and age-to-platelet index;^13,14 (2) liver function markers such as albumin and albumin-to-bilirubin index;^15,16 (3) tumor-related markers such as tumor number, AFP, lens culinaris agglutinin-reactive fraction of AFP-L3, and des-gamma-carboxy prothrombin;^17–19 (4) inflammatory markers such as NLR, PLR, and prognostic nutritional index; and^20,21 (5) pathological markers such as microvascular invasion and lowly-differentiated neoplasm.^22,23

But little attention has been paid to the relationship between these indicators or combined markers and prognosis of patients with HCC treated with TACE plus percutaneous ablation. However, the present study illustrated the development of a novel scoring system based on gender, tumor number, AFP, Fib, and APR selected by univariate and multivariate analysis, which might evaluate the recurrence risk of patients with HCC after undergoing combined therapies.

First, the present study showed that men had a higher risk of relapse compared with women, which could be explained by two points: androgen receptor activation might contribute to HCC cell progression and invasion, and the endogenous metabolite of estrogen could suppress tumor growth because of its antiproliferative, proapoptotic, and antiangiogenic activities.^24,25 Second, multiple tumors are the characteristic of multicentric development, which could not be usually detected by an imaging test, with a higher possibility to relapse and metastasize Third, AFP, an independent risk factor for recurrence in the present study, might be an indication of vascular invasion and HCC progression, contributing to an increased risk for early recurrence and is in agreement with Yang’s conclusion.^26,27 Then, deposited Fib in tumor tissue, as an extracellular matrix, could promote adhesion and migration of tumor cells by contributing to the angiogenesis of tumors.²⁸ Finally, APR, a new combination of albumin and prealbumin, perhaps was explained by two viewpoints: Supplemental infusion of albumin or blood transfusion might make exogenous albumin persist in the body for several days as a consequence of a long half-life of 20 days;²⁹ prealbumin, as a sensitive marker to reflect protein synthesis and secretion owing to a short half-life of 0.5 day, might predict the recurrence risk of HCC.³⁰

In this study, discrepancies were found in the BUN and TC levels among the groups with different risks of relapse. Therefore, it was considered that abnormality in protein and lipid metabolism among the three groups might be associated with the recurrence of HCC. Further research would be undertaken to confirm these ideas.

Few studies have explored the prognostic markers of patients with HCC after undergoing TACE plus ablation therapies. This study presented a risk scoring system based on APR, a novel marker. However, confounders related to the retrospective study could not be avoided, and parameters such as NLR, PLR, and other newer predictors were not investigated. Thus, further prospective studies including more markers are needed to be carried out.

Conclusion

In conclusion, the novel scoring system could help stratify patients with HCC into groups with different recurrence risks, following which a personalized surveillance strategy and timely antineoplastic therapies could be implemented.