Genetic Risk of Hepatocellular Carcinoma in HCV Patients
Genetic Risk of Hepatocellular Carcinoma in HCV Patients
Patient characteristics are shown in Table 2. There was no significant difference between HCC cases and controls in terms of alcohol intake, total bilirubin, prothrombin time, alanine aminotransferase (ALT) and IFN therapy. Age, proportion of males, and proportion of Child–Pugh grade B were higher in patients with HCC than in patients without HCC, whereas serum albumin level and platelet count were lower in patients with HCC.
Among the 88 SNPs, we selected ones with minor allele frequency ≥ 10%, genotyping success rate ≥ 95%, and no evidence of deviation from Hardy–Weinberg equilibrium (P < 0.05) in control subjects. Seventy-five SNPs satisfied these criteria, of which 10 showed a positive association with HCC incidence (significant crude OR): CCND2 rs1049606, RAD23B rs1805329, miR-146a rs2910164, GRP78 rs430397, MDM2 rs2279744, ALDH2 rs671, TP53 rs1042522, WRN rs1801195, XPC rs2228000, and CEP164 rs573455. To evaluate the effects of these polymorphisms on the occurrence of HCC, unconditional logistic regression analysis was performed with adjustment for age, gender, drinking history, and Child–Pugh grade. As further validation, permutation tests for allele frequencies demonstrated significant differences with six SNPs: CCND2 rs1049606, RAD23B rs1805329, MDM2 rs2279744, ALDH2 rs671, CEP164 rs573455, and GRP78 rs430397 ( Table 3 ). After determining the best fitting genetic model for each variant ( Table 4 ), CEP164 rs573455 had the strongest association with HCC; adjusted OR for the CC genotype was 0.29 (95% CI: 0.15–0.56, P = 1.9 × 10) using TT+TC as the reference.
Multiplied odds ratio was calculated for the abovementioned six SNPs to assess the cumulative effects on HCC incidence after inverting protective ORs (OR < 1) (equivalently, using the OR for the reference genotype compared with risk genotype as reference). The risk genotype and risk OR for each SNP is shown in Table 4 . Calculated MOR ranged from 1.0 to 71.7 and was categorized into five groups using cutpoints 6, 13, 24, and 70. The difference in the relative OR of HCC incidence among the risk categories was highly significant. Compared with the very low risk group, relative OR of HCC was high in the very high risk group (OR = 19.1, 95% CI: 3.95–92.6, P = 1.08 × 10) ( Table 5 ). Table 5 also presents predicted cumulative incidence of HCC over 10 years according to stage of liver fibrosis in each MOR category. Ten-year predicted cumulative incidence ranged widely from 1.67 in the very low risk group of F0 or 1 to 96.2% in the very high risk group of F4. Within any particular fibrosis stage, the predicted incidence differed by anywhere from 1.7- to 15-fold among MOR categories, with the widest range in the lowest fibrosis stages. Even with fibrosis stage 0 or 1, predicted 10-year cumulative HCC incidence is 24.5% in the very high risk MOR group. Interferon (IFN) therapy was conducted in 61 cases (23%) before development of HCC. Twelve (5%) were sustained virologic responders and 49 (18%) were non-responders. No difference in MOR was observed between sustained virological responders and non-responders.
We compared the recurrence-free survival of RFA-treated patients between high risk and low risk groups. Characteristics of the 111 patients are shown in Table 6 . The high risk group was younger, but there was no significant difference in sex, tumor size, tumor number, alanine aminotransferase (ALT), total bilirubin, albumin, platelet, prothrombin time, or α-fetoprotein. The median follow-up period was 1043 days (range, 176–2533 days). The 3-year cumulative recurrence rates were 77% in the high risk group and 59% in the low risk group. Patients in the high risk group had higher recurrence rates than those in the low risk group (P = 0.038) (Fig. 1).
(Enlarge Image)
Figure 1.
Recurrence free survival of high risk group and low risk group. Solid line, high risk group (n = 53); dotted line, low risk group (n = 58). High risk group patients had higher recurrence rate than those in low risk group (P = 0.038).
Results
Patient characteristics are shown in Table 2. There was no significant difference between HCC cases and controls in terms of alcohol intake, total bilirubin, prothrombin time, alanine aminotransferase (ALT) and IFN therapy. Age, proportion of males, and proportion of Child–Pugh grade B were higher in patients with HCC than in patients without HCC, whereas serum albumin level and platelet count were lower in patients with HCC.
SNPs Related to HCC
Among the 88 SNPs, we selected ones with minor allele frequency ≥ 10%, genotyping success rate ≥ 95%, and no evidence of deviation from Hardy–Weinberg equilibrium (P < 0.05) in control subjects. Seventy-five SNPs satisfied these criteria, of which 10 showed a positive association with HCC incidence (significant crude OR): CCND2 rs1049606, RAD23B rs1805329, miR-146a rs2910164, GRP78 rs430397, MDM2 rs2279744, ALDH2 rs671, TP53 rs1042522, WRN rs1801195, XPC rs2228000, and CEP164 rs573455. To evaluate the effects of these polymorphisms on the occurrence of HCC, unconditional logistic regression analysis was performed with adjustment for age, gender, drinking history, and Child–Pugh grade. As further validation, permutation tests for allele frequencies demonstrated significant differences with six SNPs: CCND2 rs1049606, RAD23B rs1805329, MDM2 rs2279744, ALDH2 rs671, CEP164 rs573455, and GRP78 rs430397 ( Table 3 ). After determining the best fitting genetic model for each variant ( Table 4 ), CEP164 rs573455 had the strongest association with HCC; adjusted OR for the CC genotype was 0.29 (95% CI: 0.15–0.56, P = 1.9 × 10) using TT+TC as the reference.
Prediction of HCC Development by MOR
Multiplied odds ratio was calculated for the abovementioned six SNPs to assess the cumulative effects on HCC incidence after inverting protective ORs (OR < 1) (equivalently, using the OR for the reference genotype compared with risk genotype as reference). The risk genotype and risk OR for each SNP is shown in Table 4 . Calculated MOR ranged from 1.0 to 71.7 and was categorized into five groups using cutpoints 6, 13, 24, and 70. The difference in the relative OR of HCC incidence among the risk categories was highly significant. Compared with the very low risk group, relative OR of HCC was high in the very high risk group (OR = 19.1, 95% CI: 3.95–92.6, P = 1.08 × 10) ( Table 5 ). Table 5 also presents predicted cumulative incidence of HCC over 10 years according to stage of liver fibrosis in each MOR category. Ten-year predicted cumulative incidence ranged widely from 1.67 in the very low risk group of F0 or 1 to 96.2% in the very high risk group of F4. Within any particular fibrosis stage, the predicted incidence differed by anywhere from 1.7- to 15-fold among MOR categories, with the widest range in the lowest fibrosis stages. Even with fibrosis stage 0 or 1, predicted 10-year cumulative HCC incidence is 24.5% in the very high risk MOR group. Interferon (IFN) therapy was conducted in 61 cases (23%) before development of HCC. Twelve (5%) were sustained virologic responders and 49 (18%) were non-responders. No difference in MOR was observed between sustained virological responders and non-responders.
Impact of MOR on Recurrence-free Survival of RFA-treated HCC Patients
We compared the recurrence-free survival of RFA-treated patients between high risk and low risk groups. Characteristics of the 111 patients are shown in Table 6 . The high risk group was younger, but there was no significant difference in sex, tumor size, tumor number, alanine aminotransferase (ALT), total bilirubin, albumin, platelet, prothrombin time, or α-fetoprotein. The median follow-up period was 1043 days (range, 176–2533 days). The 3-year cumulative recurrence rates were 77% in the high risk group and 59% in the low risk group. Patients in the high risk group had higher recurrence rates than those in the low risk group (P = 0.038) (Fig. 1).
(Enlarge Image)
Figure 1.
Recurrence free survival of high risk group and low risk group. Solid line, high risk group (n = 53); dotted line, low risk group (n = 58). High risk group patients had higher recurrence rate than those in low risk group (P = 0.038).
Source...