Increased Overall Survival and Decreased Cancer-Specific Mortality in Patients with Hepatocellular Carcinoma Treated by Transarterial Chemoembolization and Human Adenovirus Type-5 Combination Therapy: a Competing Risk Analysis

Background In analyzing cancer patient survival data, the problem of competing risks is often ignored. This study used a competing risk approach to evaluate the efficacy of recombinant human type-5 adenovirus (H101) in patients with hepatocellular carcinoma (HCC) treated by transarterial chemoembolization (TACE). Methods In this retrospective study, 476 patients were included. The cumulative probabilities of cancer-specific mortalities were analyzed by the Kaplan-Meier (KM) method and a competing risk model. Competing risk regression was used to assess the predictive factors for cumulative cancer-specific mortalities. Results Two hundred thirty-eight HCC patients received combination TACE and H101 therapy, and another 238 HCC patients received TACE therapy alone. For patients in the TACE with H101 group, estimated 1-, 2-, and 3-year overall survival (OS) rates were 61.0, 40.0, and 31.5%, respectively, while for patients in the TACE group, the estimated 1-, 2-, and 3-year OS rates were 55.0, 33.4, and 22.3%, respectively. The 1-, 2-, and 3-year cancer-specific mortality rates for patients in the TACE with H101 group vs. the TACE group were 37.3 vs. 42.0%, 55.7 vs. 63.5%, and 61.9 vs. 74.7%, respectively. Multivariate competing risk analysis established that a combination of TACE and H101 therapy was an independent factor in decreasing cancer-specific mortality. Conclusions Compared with TACE therapy, patients who were diagnosed with unresectable HCC treated with combined TACE and H101 therapy had increased OS and decreased cancer-specific mortality. The survival benefit was more obvious in patients with elevated AFP, absence of metastasis, single tumor, enlarged tumor, and HBsAg-positivity.


Introduction
Hepatocellular carcinoma (HCC) is the fifth most common cancer and the third leading cause of cancer-related mortality worldwide. 1 Several treatment procedures, including hepatectomy, liver transplantation, and radio-frequency ablation, are recommended for early-stage HCC. 2,3 However, due to the lack of symptoms during the early stage, 4 most of HCC cases are diagnosed at an advanced stage and unsuitable for curative therapy. 5 Transarterial chemoembolization (TACE), which focuses on delivering chemotherapeutic drugs to the tumor while blocking tumor-feeding arteries, has shown a survival Chaobin He and Yu Zhang contributed equally to this work. benefit for unresectable HCC. 6,7 However, diminished liver function, tumor enlargement, and portal vein involvement may lead to reduced efficacy of TACE. 8 Moreover, repeated TACE may also create a hypoxic microenvironment, which promotes tumor progression. 9,10 The poor prognosis 11 of unresectable HCC treated by TACE suggests that more improvements are needed to better benefit patients.
Genetic abnormalities are commonly observed during HCC formation, such as activated oncogenes 12 and inactivated tumor suppressor genes. 13 Deletion or mutation of wild-type p53 frequently occurs in HCC, indicating a poorer patient prognosis. 14 H101, which is generated by both E1B and E3 gene deletions, is a recombinant human type-5 adenovirus. 15 H101 infects tumor cells, ultimately killing them through viral oncolysis. 16 The active p53 gene in normal cells prevents the adenovirus from replicating and lysing cells, leading to selective H101 replication in cancer cells, rather than normal cells. In addition, this selectivity leads to tumor cell cytolysis without adverse side effects. 17 Tumor cell sensitivity to H101 in vitro is negatively reflected by the p53 gene sequence, which is thought to be due to p53 inactivation by several mechanisms. 18 Furthermore, H101 enhances the cellmediated immune and host immune systems, improving the efficacy of TACE treatment. 19 In addition, adenovirus safety was improved by deleting a 78.3-85.8-nm gene segment in the E3 region that encodes the adenovirus death protein. 20 Combined TACE and H101 therapy will likely benefit HCC patient survival 21 ; however, no phase III clinical trials have shown a survival benefit for this combination therapy to date. Therefore, it is necessary to explore the value of combining H101 with TACE treatment in HCC patients.
Competing risk refers to an event that precludes another event under investigation or fundamentally alters the probability of the outcome of interest. 22,23 In the survival analysis of HCC patients, a patient may experience cancer-specific death, non-cancerspecific death, survival, or lost to follow-up. Non-cancerspecific death is a competing risk event that prevents the event of interest, cancer-specific death. Failure to recognize the presence of a competing risk may result in misleading conclusions in clinical practice. In this case, it is unsuitable to use the Kaplan-Meier (KM) method to analyze survival data because this method treats competing events independently and overestimates the proportion of cancer-specific mortality. The cumulative incidence function (CIF) 24 accounts for the informative nature of this censorship and corresponds to the probability of a particular event occurring without assuming of independence between event types and can be used to analyze survival data.
Competing risk analysis has been adopted to analyze several cancers, including nasopharyngeal, 25 ovarian, 26 kidney, 27 and breast cancers. 28 However, to our knowledge, no relative reports focus on this analysis in HCC patients. In the current work, a competing risk analysis was conducted to explore the therapeutic effects of combined TACE and H101 therapy in HCC patients.

Patients and Methods
Patients Clinical data were collected using a cohort of consecutive patients who received TACE therapy as their initial treatment at the Department of Hepatobiliary and Pancreatic Surgery of Sun Yat-sen University Cancer Center between January 2007 and July 2015. HCC was diagnosed based on the typical features of HCC identified by two radiological images or one radiological image combined with elevated alpha-fetoprotein (AFP) levels (≥ 400 ng/mL) or histopathological evidence, which is consistent with the diagnostic criteria for HCC used by American Association for the Study of the Liver guidelines. 29 The inclusion criteria for this study were as follows: (1) no previous treatment before TACE, (2) liver function Child-Pugh A or B, and (3) a follow-up period ≥ 1 year. The exclusion criteria were as follows: (1) liver function Child-Pugh C, (2) common diagnosis of secondary cancers, (3) any therapies other than TACE after initial TACE treatment, or (4) lost to follow-up.

Data Collection
We reviewed the patient files for the clinical and radiological data that were retrieved at the time of diagnosis before the initial TACE was performed. Clinical and radiological parameters including age, gender, white blood cell count (WBC), platelet (PLT) count, alanine transaminase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL), indirect bilirubin (IBIL), alkaline phosphatase (ALP), albumin (ALB), Creactive protein (CRP), AFP, hepatitis B surface antigen (HbsAg), splenomegaly, metastasis, vascular invasion, tumor number, tumor size, antiviral therapy, and tumor-nodemetastasis (TNM) stage were collected and analyzed.

Treatment Procedure
Each patient in this study received three cycles of uniform treatment protocols. The Seldinger technique was performed as previously reported. 30 Carboplatin at a dose of 300 mg (Bristol-Myers Squibb, NY, New York, USA) was used for hepatic artery infusion chemotherapy. Subsequently, 50 mg epirubicin (Pharmorubicin, Pfizer, Wuxi, Jiangsu, China) and 6 mg mitomycin (Zhejiang Hisun Pharmaceutical Co. Ltd., Taizhou, Zhejiang, China) mixed with Lipiodol (Lipiodol Ultra-Fluide; Andre Guerbet Laboratories, France) were used for chemolipiodolization. The Lipiodol dose was determined based on tumor location, size, and number and ranged from 5 to 30 mL. Sterile-purified H101 viruses were produced for human clinical use by Shanghai Sunway Biotech (Shanghai, China) and safety tested by the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Before the infusion chemotherapy, H101 was injected via catheter into the hepatic artery supplying the tumor(s). A total of 1.0 × 10 12 virus particles in 10 mL of 0.9% sodium chloride solution were administered. 31

Follow-Up
All patients in this study were followed regularly once every 2 months during the first year and once every 3 months thereafter. Radiological examinations, such as liver ultrasonography, computed tomogram (CT) scans, and magnetic resonance imaging (MRI), were performed as needed. Hematological tests, including AFP and liver function, were performed each time. Overall survival (OS) was defined as the duration from the date of the first TACE until death or the last follow-up. The last follow-up date was September 30, 2017. The median follow-up period was 13 months.

Statistical Analysis
Continuous data are presented as means and ranges and compared using Student's t test. Categorical data are shown as frequencies and proportions and compared using a Chisquare test and Fisher's exact test. Univariate and multivariate analyses were performed using the Cox regression model and the associated 95% confidence interval (CI) was calculated. OS was analyzed using the KM method. The log-rank test was used to compare the differences between groups. All statistical analyses were performed using SPSS version 22 (SPSS Inc., Chicago, IL, USA). MedCalc software version 11.4.2.0 (http://www.medcalc.be) was used to compare the survival. A two-tailed P value < 0.05 was considered statistically significant. The cumulative incidence of overall mortality and cancer-specific mortality was determined by the competing risk analysis. Non-cancer-specific mortality was evaluated as competing mortality in this study. The combined effects of the variables on overall mortality and cancer-specific mortality were evaluated by the Cox proportional hazards analysis of the Fine and Grey model. 32

Patient Characteristics
This study included 476 HCC patients who received TACE therapy during the study period. Of all the included patients, 238 who received TACE with H101 were sorted into the TACE with H101 group, and the remaining 238 patients who received TACE alone were sorted into the TACE group.
Baseline characteristic comparisons between the two groups are shown in Table 1. The clinical data include 430 males (90.3%) and 46 females (9.7%) with a median age of 55 years (range, 15-94 years). Most patients (94.3%) in this cohort were HBsAg-positive. No patients were infected with the hepatitis C virus (HCV). Most patients had an enlarged tumor size (tumor size > 5 cm) and multiple tumors that were identified by radiography. Of the patients, 5.5 and 4.6% had metastases in the TACE with H101 and TACE groups, respectively. The proportion of HBsAg positivity was slightly higher in the TACE group than in the TACE with H101 group. Other than this, only TBIL significantly differed between the two groups.

OS Analysis
The patients in TACE with H101 group had significantly favorable prognoses compared with patients in the TACE group (P = 0.047, Fig. 1). For patients in the TACE with H101 group, the estimated 1-, 2-, and 3-year OS rates were 61.0, 40.0, and 31.5%, respectively, while for patients in the TACE group, the estimated 1-, 2-, and 3-year OS rates were 55.0, 33.4, and 22.3%, respectively. The median OS for patients in the TACE with H101 and TACE groups were 13.7 and 13.1 months, respectively. Patients with elevated AFP values, metastases, vascular invasion, multiple tumors, larger tumors, and elevated TNM stages had poorer OS based on the univariate analysis. Apart from these variables, age, antiviral therapy, and combined TACE and H101 therapy were all associated with OS (Table 2).

Multivariate Analysis
Variables that were significantly associated with OS were analyzed by multivariate Cox regression analysis.

Discussion
To our knowledge, this study is the first to demonstrate the prognostic value of H101 combined with TACE in HCC   Abbreviations as in Table 2 our study. This benefit mainly originated from decreased cancer-specific mortality, consistent with previous reports. 31,34 Furthermore, subgroup analyses were adopted, and comparing cumulative mortality showed that cumulative mortalities differed significantly between the TACE with H101 and TACE groups. In addition, the survival benefit for H101 was more obvious when HCC patients had elevated AFP, no metastases, single tumors, enlarged tumors, or HBsAg positivity. Survival analysis is often used to assess the time to an event of interest in follow-up studies. Apart from the event of interest, other events may prevent target outcome from occurring. In the HCC patient survival analysis, noncancer-specific death prevented the appearance of cancerspecific death. In standard survival analysis, the risk of cancer-specific death is incorrect if non-cancer-specific deaths occur. 35 Therefore, the KM method may overestimate the cumulative mortality of HCC patients who received TACE with or without H101. In this study, noncancer-specific mortality was accounted for as competing mortality, and cancer-specific mortality was compared with the results from the standard survival analysis.
Competing risk methods are used to analyze risk factors in biomedical research, especially in cancer, either at the screening or treatment stage, which may influence decision making. 36,37 What is more, multivariate Cox regression may lead to confounding in exploring predicted factor values when competing risks are present. 38 Evaluating factor efficacy will be more realistic using competing risk analysis. Multivariate analysis showed significant differences in both cancer-specific mortalities and overall survival rates between the two groups in this study. The results of the current work may further consolidate the role of H101 in TACE therapy for HCC patients.
H101 is an E1B/E3B-deleted adenovirus that restricts p53-mutated neoplasm replication, sparing p53 wild-type tissues. 39 The decreased cancer-specific mortality for HCC patients treated with combined TACE and H101 therapy compared with TACE therapy alone may be explained by the following mechanism. Carboplatin, a chemotherapy drug commonly used for TACE at our institution, induces apoptosis and cell cycle arrest through p53 apoptosis. 40 Tumor cell inhibition will be enhanced when TACE and H101 are used together. In addition, over 80% of patients with HCC in Asia are hepatitis B virus (HBV)positive 41 and over 90% of patients in this study are HBV-positive. HBV produces HBV X protein (HBx), which inhibits p53 gene expression. 42 Therefore, H101 provides a survival benefit for HCC patients. In this study, it was revealed that patients had decreased cancer-specific mortality in TACE with H101 group compared with TACE group when patients were HBV-positive, while the differences were not significant between the two groups when patients were HBV-negative. In addition, some reports have shown that antivirus therapy for HBV-positive patients was associated with prolonged OS after TACE. 43,44 Similarly, antivirus therapy was an independent prognostic factor in both multivariate analyses and the Fine and Gray regression model for HCC patients in this study. It was also suggested that the combining H101 with antivirus therapy may increase curative effects for HCC patients after TACE therapy, which requires further investigation.
Similar to other report, 45 our study revealed that elevated AFP values, presence of vascular invasion, and enlarged tumors were also independent predictors for HCC patients in this study. Interestingly, subgroup analyses showed that patients who received combined TACE and H101 therapy had significantly decreased cancer-specific mortality in the elevated-AFP and enlarged tumor subgroups. Cancer-specific mortalities in patients after TACE therapy were higher than those of the combined TACE and H101 therapy in the presence of vascular invasion subgroup, although the differences were not significant in this study. Cancer-specific mortalities were comparable between the two groups in patients whose AFP values were lower than 400 mg/mL or whose tumor sizes were smaller than 5 cm in this study. One possible explanation is that the p53 gene mutations and loss of p53 gene heterozygosity are common in HCC, especially HCC with a heavy tumor burden, which is reflected by the elevated APF values and enlarged tumors. 46,47 H101 may be more effective for tumors in which p53 gene mutations or deletions are more frequent. 48 What is more, blood flow is abundant in HCC with a heavy tumor burden. 49 Higher H101 concentrations in vascular-rich areas where tumor cells grow faster may inhibit tumor cell growth in a timely and effective manner.
The major limitations of the present study were its retrospective nature and the single-center experiment. In addition, most included patients were predominantly HBV-infected in China. Whether this result can be applied to patients with HCV infection requires further confirmation. Additionally, longer follow-up times may be needed to observe more endpoints to more precisely estimate cancer-specific mortality. Large-scale, further prospective, randomized-controlled, long-term studies are needed to confirm our results.
In conclusion, based on the competing risk model, we demonstrated that combining TACE and H101 therapy decreased cancer-specific mortality in HCC patients compared with TACE therapy alone. The survival benefit was more obvious in patients with elevated AFP, absence of metastasis, single tumor, enlarged tumor, and HBsAg positivity. The results of this study may further consolidate the role of H101 in TACE therapy for patients with HCC.