| Background/Aim:Hepatitis B virus (HBV) infection is a global health problem, and more than 350 million people in the world are chronic carriers of this virus, China is one of the highest endemic regions for hepatocellular carcinoma (HCC) in the world. Distinct clinical and virological characteristics of the HBV infection have been reported in different geographical parts of the world in many reports. The goal of the present study was to assess the risk of demographic, biochemical, and specific HBV gene mutations in the development of HCC.Methods:This nested case-control study was conducted within a large cohort of male HBV carriers in Qidong, Jiangsu Province, China. From 1996, a total of 2387 males who were seropositive for HBsAg and free of HCC at recruitment were followed up until October 2006. A serum specimen was collected from each participant at interview. We confirmed 196 incident HCC patients as cases. A total of 323 controls who were alive and had not been diagnosed with HCC throughout the follow-up period were recruited. The serum HBV DNA levels were determined using the Fluorescein quantitative polymerase chain reaction (FQ-PCR) detection system, according to the manufacturer’s instructions. HBV genes of the pre-S/enhancerⅡ/basal core promoter /precore regions were amplified by nested PCR and directly sequenced in both the forward and reverse directions.Results:1. After adjustment for age at recruitment, histories of cigarette smoking and alcohol consumption, compared with participants having serum HBV DNA levels of less than 2.69 log10 copies/mL, the adjusted OR was 4.53 (95% CI,2.69-7.63) for participants with serum HBV DNA levels of 4.00-5.99 log10 copies/mL; 8.71 (95%CI,5.29-14.34),6.00 log10 copies/mL or greater. Subjects with persistent high HBV DNA levels during follow-up had a higher risk of HCC. When we examined HBV DNA sequences in the pre-S and EnhⅡ/BCP/PC regions, pre-S deletion, T1653, T1762/A1764, and T1766 and/or A1768 mutations were significantly associated with HCC, showing adjusted ORs from 2.25 to 3.02.2. On further calculation using stepwise logistic regression analysis, the followings were found to be independent risk factors:AFP levels>20 ng/mL, high HBV DNA levels, pre-S deletion, and T1762/A1764 double mutations.3. Compared to patients with wild-type HBV or a single mutation, patients with combined mutations of T1762/A1764, pre-S deletion, and T1766 and/or A1768 had a higher risk of HCC.4. Among 134 HCC patients with success sequence data,16 HCC patients with sequential serum samples were selected for the longitudinal investigation of specific mutations (pre-S deletion, T1762/A1764 double mutations, and T1766 and/or A1768 mutations). Of these 16 patients,7 showed a gradual combination of pre-S deletion, T1762/A1764 double mutations, and T1766 and/or A1768 mutations.Conclusions:1. AFP levels>20 ng/mL, high HBV DNA levels, pre-S deletion, and T1762/A1764 double mutations at recruitment were independent risk factors of IICC.2. Combination of pre-S deletion and core promoter mutations increased the risk of HCC. Background/Aim:Hepatocellular carcinoma (HCC) is a major health problem worldwide. More than 500 000 new HCC cases are currently diagnosed throughout the world each year and approximately half of the patients with HCC come from China. Major risk factors for the development of HCC are chronic infection with hepatitis B virus (HBV). However, only a small fraction of chronic HBV carriers develop HCC, indicating that genetic and environmental factors may play an important role in the development of HCC. HCC is most likely a complex disease caused by a complex interplay between multiple genes, each with a small contribution to overall risk. The purpose of this study was to develop a practical approach for building multi-single nucleotide polymorphisms (SNPs) models to predict the susceptibility of HCC in chronic HBV carriers, and to validate the utility of this model in an independent cohort.Methods:All chronic HBV carriers involved in this study came from Qidong, China. In the first stage, we used the Affymetrix Genome-Wide Human SNP Array 5.0 to genotype 500447 SNPs in 50 HCC male cases with chronic HBV infection and 50 chronic HBV male carriers without HCC. After the features were ranked, using cross-validation in conjunction with machine learning classifiers including naive Bayes, alternating decision tree, J48 pruned decision tree, support vector machine (SVM), we identified a subset of SNPs as key discriminators between HCC and controls. In the second stage, using Illumina Golden Gate platform for SNP genotyping, we validate the accuracy of the models in an independent cohort.Results:After data filter, a total of 282 SNPs associated with HCC was individually significant at the P＜0.001 level. Through 10-fold cross validation of the 9 SNPs model, the naive Bayes classifier performed maximally among four predictive models, achieving 95.2% of total classifying accuracy in distinguishing between the two groups. The use of alternating decision tree resulted in 73.0% accuracy, the use of J48 pruned decision tree resulted in 81.8% accuracy, the use of SVM resulted in 90.9% accuracy. Second step, when validating the utility of this 9 SNPs model in an independent cohort with 157 HCC cases and 278 controls, we obtained a highest of 68.28% predictive power when using J48 pruned decision tree classifier with one SNP missing value. The sensitivity and specificity were 30.57% and 89.57%, respectively.Conclusions:In this study, we assessed several machine learning classifiers in genomic studies of HCC. The purpose of this study was to develop a practical approach for building multi-SNP models to predict HCC risk, and to validate this as being superior to traditional single SNP. Our findings suggested that our experiment may provide a plausible way to identify models in HCC. Although genomic profiling is not likely to be ready for clinical use for some time, as high-throughput technology for SNPs improves and as more SNPs are identified, it is likely that much higher predictive accuracy will be achieved and a useful clinical tool developed to screen the individuals at high risk of HCC in chronic HBV carriers. Background/Aim:Curative resection is considered the most effective treatment and the prognosis of hepatocellular carcinoma (HCC) was greatly improved. High possibility of intrahepatic recurrence remains one major obstacle for further improving the survival and prognosis. Therefore, adjuvant therapy to reduce tumor recurrence after resection and improve long-term survival is definitely needed. A retrospective cohort study was conducted to investigate the effect of interferon-a (IFN-a) therapy after curative resection on survival and recurrence in patients with hepatitis B virus (HBV)-related HCC.Methods:Of 568 HBV-related HCC patients who underwent curative resection,101 patients received postoperative IFN-a therapy (5 million units three times every week for 18 months). Clinical and pathological factors were compared between patients with postoperative IFN-a therapy or not. Risk factors for survival, early and late recurrence (2 years as cut-off) were studied. Risk factors that influenced prognosis were identified by the Cox proportional hazards model.Results:During a median follow-up time of 53.3 months (from 3.5 to 74 months),242 patients (42.6%) died. Recurrence was found in 333 patients (58.6%). There was no significant difference in clinicopathological factors between the two groups. Patients with postoperative IFN-a therapy had higher overall survival rates compared to those without IFN-a therapy (hazards ratio (HR):0.612,95% CI:0.422-0.889, P=0.010). No significant difference in disease-free survival rates was detected between the two groups (HR:0.786,95% CI:0.597-1.035, P=0.086). Multivariate analysis revealed that postoperative IFN-a therapy was an independent factor for overall survival (HR: 0.611,95% CI:0.421-0.887, P=0.010) and significantly reduced early recurrence (HR:0.562,95% CI:0.375-0.840, P=0.005). However, patients with postoperative IFN-αtherapy or not had similar cumulative late recurrence rates (P=0.396).Conclusions:IFN-αtherapy after curative resection prevented early recurrence and improved overall survival of patients with HBV-related HCC. |
PDF Full Text Request