+
MODEL
Kaohsiung Journal of Medical Sciences (2017) xx, 1e5
Available online at www.sciencedirect.com
ScienceDirect journal homepage: http://www.kjms-online.com
Original Article
Independent and additive interaction between polymorphisms of tumor necrosis factor aL308 and lymphotoxin aD252 on risk of hepatocellular carcinoma related to hepatitis B Jung-Fa Tsai a,b,*, Shinn-Chern Chen a,b, Zu-Yau Lin a,b, Chia-Yen Dai a,b, Jee-Fu Huang a,b, Min-Lung Yu a,b, Wan-Long Chuang a,b a
Hepatobiliary Division, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan b Faculty of Internal Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan Received 5 January 2017; accepted 25 April 2017
KEYWORDS Hepatitis B related hepatocellular carcinoma; Lymphotoxin a; Single nucleotide polymorphism; Synergy index; Tumor necrosis factor a
Abstract This caseecontrol study was aimed to assess the effect of genetic variants of tumor necrosis factor (TNF) a308 and lymphotoxin (LT) aþ252 on development of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC). Their geneegene interaction was also investigated. We enrolled 200 pairs of age- and sex-matched patients with cirrhotic HBV-HCC and unrelated patients with HBV-cirrhosis alone. Polymorphisms of TNFa308 and LTaþ252 were genotyped. Synergy index was used to calculate interaction between the variant genotypes. The results indicated that the frequency distribution of the variant genotypes (TNFa308 G/A and LTaþ252 G/G) in patients with HCC were significantly higher than those in patients with cirrhosis alone. Multivariate analysis indicated that TNFa308 G/A (odds ratio [OR], 2.34) and LTaþ252 G/G (OR, 2.04) were independent risk factors for HCC. By the clinical characteristics of study population, multivariate analysis demonstrated that independent factors associated with harboring the variant genotypes included cirrhosis with Child-Pugh C (OR Z 6.47 in cases and OR Z 11.56 in controls) and thrombocytopenia (OR Z 8.86 in cases and OR Z 7.74 in controls). Calculation of synergy index (SI) indicated that there are additive interaction between TNFa308 G/A and LTaþ252 G/G on risk of HCC (SI Z 1.29). In conclusion: There are independent and additive interactions between TNFa308 G/A and LTaþ252
Conflicts of interest: All authors declare no conflicts of Interest. * Corresponding author. Hepatobiliary Division, Department of Internal Medicine, Kaohsiung Medical University Hospital, 100 Tzyou 1 Road, Kaohsiung, 807, Taiwan. E-mail address:
[email protected] (J.-F. Tsai). http://dx.doi.org/10.1016/j.kjms.2017.04.009 1607-551X/Copyright ª 2017, Kaohsiung Medical University. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Please cite this article in press as: Tsai J-F, et al., Independent and additive interaction between polymorphisms of tumor necrosis factor a308 and lymphotoxin aþ252 on risk of hepatocellular carcinoma related to hepatitis B, Kaohsiung Journal of Medical Sciences (2017), http://dx.doi.org/10.1016/j.kjms.2017.04.009
+
MODEL
2
J.-F. Tsai et al. G/G on risk for HBV-HCC. They correlated with advanced hepatic fibrosis and severe liver damage, which might contribute to a higher risk for HCC. Copyright ª 2017, Kaohsiung Medical University. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/ by-nc-nd/4.0/).
Introduction A causal relationship among chronic hepatitis B virus (HBV) infection, cirrhosis and hepatocellular carcinoma (HCC) is well established [1,2], which in part are mediated by cytokines [2e6]. HCC is inherently associated with upregulation of cytokines [2,4]. Among them, aberrant expression of cytokines is thought to be critically involved [3e5]. As cytokines mediate inflammation and liver injury after HBV infection [3], there should be intimate association between cytokines and HBV-HCC. However, the results reported are inconsistent [7e11]. The proinflammatory cytokines tumor necrosis factor (TNF) a and lymphotoxin (LT) a are members of the TNF superfamily [5]. LTa has a close structural homology and about 30% amino acid sequence identify to TNFa. It carries out most TNFa activities. Both cytokines initiate similar (if not identical) biologic responses. The presence of single nucleotide polymorphism (SNP) within the coding and/or regulatory regions affect cytokine expression level [5,12]. They are important mediators of hepatic fibrogenesis [4e6] and tumorigenic [4,13]. Their SNPs might increase susceptibility to HCC through enhanced hepatic fibrosis and liver injury. However, such an association and their geneegene interaction remain to be elucidated. We conducted this study to clarify these issues.
Materials and methods Study population We enrolled 200 consecutive newly diagnosed patients with HCC as cases. Another 200, gender- and age-matched (5 year), unrelated consecutive patients with cirrhosis alone were enrolled as controls. All patients were hepatitis B surface antigen (HBsAg)-positive and negative for anti-HCV. Instead of HBV DNA, we used hepatitis B e antigen (HBeAg) to indicate viral replication. As HBV nucleos(t)ide therapy may decrease hepatic inflammation or regression of fibrosis, resulting in decrease risk for HCC, patients received this therapy were excluded. These patients were hospitalized or had visited outpatient clinics at Kaohsiung Medical University Hospital from 2005 to 2008. Patients with HCC were eligible for the study if they were diagnosed by aspiration cytology or biopsy and were free from any known disease with a genetic predisposition. Tumor staging was according to tumor-nodemetastasis (TNM) classification. Cirrhosis was diagnosed as described previously [14]. Patients with cirrhosis were classified into the 3 Child-Pugh grades based on their
clinical status. There was no space-occupying lesion in the liver in patients with cirrhosis alone, as evidenced by abdominal sonography. All participants were Chinese Han populations in southern Taiwan. This study was approved by the Investigation and Ethics Committee of the Kaohsiung Medical University Hospital. All procedures performed were in accordance with the ethical standards of the Investigation and Ethics Committee of the hospital. All participants gave their written informed consent prior to inclusion in the study.
DNA extraction Genomic DNA was isolated from EDTA preserved whole blood by a standard proteinase K digestion and phenolechloroform methods.
Polymorphism analysis The genotyping for the TNFa308 G > A and LTaþ252 A > G was performed by the polymerase chain reaction (PCR) according to the methods described previously, respectively [15,16].
Serologic examination HBsAg, HBeAg, anti-HBe, and aefetoprotein (AFP) were detected by radioimmunoassays (Abbott Laboratories, North Chicago, IL), Anti-HCV were detected by third generation Abbott HCV EIA (Abbott Laboratories, North Chicago, IL), respectively. For anti-HCV, reactive specimens were retested. Only repeatedly reactive specimens were interpreted as anti-HCV positive. Conventional liver function tests were measured by an autoanalyzer (Hitachi, Model 736, Tokyo, Japan).
Statistical analysis Genotypic deviation from HardyeWeinberg equilibrium was assessed by a goodness-of-fit c2 test. The following statistical methods were performed using the SPSS statistical package (version 19.0; IBM Co., Armonk, NY, USA): Univariate analyses included the ManneWhitney U test, c2 test with Yates’ correction or Fisher’s exact test when appropriate, and Mantel extension test for trend; Conditional logistic regression was performed as multivariate analysis. Two-tailed p values and 95% CI were given where appropriate. An alpha of 0.05 was used as the indicator of statistical significance. Calculation of the synergy Index (SI) was used for interaction between genes as previously described [17].
Please cite this article in press as: Tsai J-F, et al., Independent and additive interaction between polymorphisms of tumor necrosis factor a308 and lymphotoxin aþ252 on risk of hepatocellular carcinoma related to hepatitis B, Kaohsiung Journal of Medical Sciences (2017), http://dx.doi.org/10.1016/j.kjms.2017.04.009
+
MODEL
TNFa & LTa as risk factors of HBV-HCC
3
Results
Table 2 Frequencies of genotypes in TNFa308 and lymphotoxin aþ252 polymorphisms and risk of hepatocellular carcinoma.
Demographic profile of cases and controls There was no statistical difference in distribution of gender, median age, status of HBeAg and antiHBe, status of Child-Pugh classification, and peripheral blood platelets between Cases and Control group (Table 1). The TNM stages of I, II, III and IV in patients with HCC were 27, 78, 75 and 20, respectively.
Genotypic polymorphisms in cases and controls The results of each SNP in controls were in accordance with the HardyeWeinberg equilibrium (data not shown). As shown in Table 2, frequency distribution of the variant genotypes (LTaþ252 G/G and TNFa308 G/A) in cases were significantly higher than those in controls.
Frequency distribution of TNF polymorphisms and risk for HCC As shown in Table 2, multivariate analysis indicated that SNPs of LTaþ252 G/G SNP (OR Z 2.04; 95% CI, 1.19e3.19) and TNFa308 (OR Z 2.34; 95% CI, 1.29e4.25) were independent risk factors for HCC.
Clinical characteristics in study population by status of the variant TNF genotypes (TNFaL308 G/ A and/or LTaD252 G/G) Univariate analysis indicated that there was a positive linear trend among cases and controls when cirrhosis was Table 1
Basic profiles in patients with HCC and controls.
Variables
Cases (n Z 200)
Controls (n Z 200)
p value*
Gender (male:female) Median age (range; yrs) HBsAg/HBeAg Positive/negative Positive/positive Cirrhosis Child-Pugh grade A B C Platelet (109/L) 130 >130 TNM stage of HCC I/II/III/IV
164:36
164:36
NS
56 (33e74)
56 (35e75)
NS NS
166 (83.0%) 34 (17.0%) 200 (100%)
172 (86.0%) 28 (14.0%) 200 (100%)
101 (50.5%) 50 (25.0%) 49 (24.5%) 100 (30e210) 155 (77.5%) 45 (22.5%)
NS 108 (54.0%) 51 (25.5%) 41 (20.5%) 83 (11e273) NS 153 (76.5%) 47 (23.5%)
27/78/75/20
HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; NS, nonsignificant; TNM, tumor-node-metastasis. *ManneWhitney U test for continuous variables and c2 test with Yates’ correction for category variables, respectively.
Genotype/ Cases Controls OR variant (n Z 200) (n Z 200) (95% CI) allele Ltaþ252 A > G A/A 0.30 A/G 0.45
0.45 0.43
G/G
0.12
0.25
TNF-a308 G > A G/G 0.77 G/A 0.23
0.91 0.10
A/A
0.0
0.0.
Adjusted OR* (95% CI)
1.0 1.57 (0.99e2.50) 2.48 2.04 (1.48e4.15) (1.19e3.49)
y
1.0 2.85 2.34 (1.60e5.6) (1.29e4.25)z
CI, confidence interval; OR, odds ratio; TNF, tumor necrosis factor; LT, lymphotoxin. *Adjusted for TNFa308, LTaþ252, TNFa308 LTaþ252, gender, age >50 years, and Child-Pugh grade of cirrhosis by logistic regression analysis. y p Z 0.037. z p Z 0.004.
categorized by Child-Pugh classification (pfor trend Z 0.001; Table 3). The higher the Child-Pugh grade, the higher the frequency of harboring the variant genotypes. HCC patients with higher serum AFP concentration (400 mg/L) and thrombocytopenia (platelet count 130 109/L) had significantly higher frequency of harboring the variant genotypes, respectively. However, all control patients had AFP < 400 mg/L. Multivariate analysis indicated that cirrhotic patients with Child-Pugh C (OR Z 6.47, 95% CI, 2.99e13.97 in cases; OR Z 11.56, 95% CI, 4.67e28.67 in controls, respectively) and thrombocytopenia (OR Z 8.86, 95% CI, 2.55e30.71 in cases; OR Z 7.74, 95% CI, 1.57e38.24 in controls, respectively) were independent factors for harboring the variant genotypes (Table 3).
Interaction between variant genotypes on risk for HCC As shown in Table 4, patients with either variant genotype had significantly association with risk for HCC. Calculation of synergy index indicated that there was an independent and additive interaction on risk for HCC (SI Z 1.29). However, there was no multiplicative interaction among them on multivariate analysis (Table 2).
Discussion This study demonstrated that increased frequencies of the variant genotypes (TNFa308 G/A and LTaþ252 G/G) related to risk of HBV-associated HCC (Table 2). Further analysis indicated that there was an independent and additive interaction between both variant genotypes (Tables 2 and 4). Moreover, all patients with Child-Pugh C and/or thrombocytopenia were independent factors harboring
Please cite this article in press as: Tsai J-F, et al., Independent and additive interaction between polymorphisms of tumor necrosis factor a308 and lymphotoxin aþ252 on risk of hepatocellular carcinoma related to hepatitis B, Kaohsiung Journal of Medical Sciences (2017), http://dx.doi.org/10.1016/j.kjms.2017.04.009
+
MODEL
4
J.-F. Tsai et al. Table 3
Clinical characteristics in study populations by status of the variant genotypes (TNFa308 G/A and/or LTaþ252 G/G).
Parameters
Group
Patients with variant genotypes Cases
Gender Age (yr)
% (positive/tested)
p
Male Female 50 >50
38.4 38.9 44.6 36.1
NS
A* B* , C* a Positive Negative 400 >400 130 >130
, 17.8*(18/101)** *** , 44.0* (22/50)** **** , 75.5*,a (37/49)*** **** 38.2 (13/34) 38.6 (64/166) 31.5 (34/108) 46.7 (43/92) 47.7 (74/155)b 6.7 (3/45)
Child-Pugh grade
HBeAg AFP (mg/L) Plateletb (109/L)
Controls y
(63/164) (14/36) (25/56) (52/144)
NS
% (positive/tested)
py
15.9 30.6 26.3 15.4
NS
(26/164) (11/36) (15/57) (22/143)
NS
0.001
NS 0.039 0.001
0.001 5.6# (6/108)##,### 19.6# (10/51)##,#### 51.2#,c (21/41)###,#### 28.6 (8/28) 16.9 (29/172) 18.5 (37/200) 0 (0/0) 22.9 (35/153)d 4.3 (2/47)
NS e 0.001
AFP, alpha-fetoprotein; NS, nonsignificant; positive, patients with variant genotypes; TNF, tumor necrosis factor; ULN, upper limit of normal. y 2 c test with Yates’correction. *,#pfor trend Z 0.001. **,##p Z 0.001. ***,###p Z 0.001. ****,####p Z 0.003. a,b; c,d Multivariate analysis with conditional stepwise logistic regression analysis indicated that cirrhosis with Child-Pugh C (OR, 6.47, 95% CI, 2.99e13.97 in cases, and OR, 11.56, 95% CI, 4.67e28.67 in controls) and thrombocytopenia (OR, 8.86, 95% CI, 2.55e30.71 in cases, and OR, 7.74, 95% CI, 1.57e38.24 in controls) were independent risk factors for harboring the variant genotypes.
Table 4
Interaction between TNFa308 G/A and LTaþ252 A/G on risk of hepatocellular carcinoma related to hepatitis B.
TNFa-308 G/A
LTaþ252 A/G
Absent Present Absent Present
Absent Absent Present Present
b
SE
p
OR (95% CI)a
0.015 0.014 0.001
1.0 2.15 (1.15e3.98) 2.52 (1.21e5.27) 4.32 (1.79e10.41)
Synergy indexb 1.29
0.76 0.92 1.46
0.31 0.37 0.44
b, coefficient; CI, confidence interval; OR, odds ratio; SE, standard error. a Adjusted for sex, age >50 years, HBsAg, TNFa-308 G/A, and LTaþ252 A/G by conditional logistic regression analysis. b Synergy Index Z (OR11 1)/(OR01 þ OR10 2), where OR11 is odds ratio of the joint effect of 2 risk factors; OR01 and OR10 are OR of each risk factor in the absence of the other [17].
these variant SNPs (Table 3). They might contribute to a higher risk for HCC. It is known that cirrhosis with Child-Pugh C [1e3] and low platelet count [18] reflect severity of hepatic fibrosis, portal hypertension and severity of liver disease. Hence, they may cause or aggravate persistent hepatic fibrosis (or cirrhosis), accelerate progression of cirrhosis to HCC. Growing evidence indicates that SNPs of TNFa gene loci play pivotal roles in HBV immunopathogenesis and hepatocarcinogenesis [3,19,20]. Circulating levels of TNFa and LTa increased during chronic HBV infection [3,21,22]. There are over-expression of hepatic mRNA and protein of TNFa and LTa in HBV-induced HCC, and correlate with severity of liver inflammation and tissue injury [7,21]. HCC cells produce TNFa [23] and LTa [4,24]. The elevated circulating cytokine concentration decreased after anti-
cancer therapy [25]. Based on these observations, it is reasonable to speculate that the high circulating cytokine levels and/or TNFa activity may be attributed to the variant genotypes. Therefore, these variants could be a causal predisposing factor for fibrosis and/or subsequent HCC. HBV is an oncogenic virus. HBV itself stimulates TNFa production, whereas hepatitis B x antigen (HBxAg) transactivates TNFa expression [3,26]. In HBV-induced chronic inflammatory response, both virus-induced and hostinduced cytokine production might contribute to viral carcinogenesis [3,4]. As SNPs of both TNFa and LTa increased cytokine secretions and thus increased TNFa activities. This may interpretate the additive interactions between both variant genotypes (Table 4).
Please cite this article in press as: Tsai J-F, et al., Independent and additive interaction between polymorphisms of tumor necrosis factor a308 and lymphotoxin aþ252 on risk of hepatocellular carcinoma related to hepatitis B, Kaohsiung Journal of Medical Sciences (2017), http://dx.doi.org/10.1016/j.kjms.2017.04.009
+
MODEL
TNFa & LTa as risk factors of HBV-HCC
5
Our findings are consistent with previous reports that TNFa308 G/A genotype are involved in hepatic carcinogenesis [7,8]. However, inconsistent results have been reported [11e15]. However, this study has several potential limitations. The first one is the rather small sample size. Second, our analysis may have a type I error due to multiple comparisons and/or a type II error due to small sample number size. Third, the ethic background may bias the estimate. This study was performed in the Han Chinese; therefore, the observed findings may not be generalizable to other populations. Therefore, the results should be confirmed in a larger series as well as in patients of different ethic origin. Fourth, this study did not compare “The duration of cirrhosis from diagnosis to study enrollment”. This parameter might reflect more accuracy the duration of HBV infection and liver disease activity. When using it to assess risk for HCC, it may decrease lead-time bias and length bias. Fifth, the selected cutoff values for thrombocytopenia (platelet count 130 109/L) used in this study was adapted from Lackner et al. [27]. The normal range in our hospital is between 16.0 109/L and 40.0 109/L. Based on these ranges, the percentage of thrombocytopenia should be The selected cutoff values for significant fibrosis (platelet count 130 109/L) was adapted from Lackner et al. [27]. In conclusion, there are independent and additive interactions between these SNPs on risk of HBV-related HCC. Chronic HBsAg carriers with these variants should receive more intensive surveillance for early detection of HCC.
[14]
References
[19]
[1] de Martel C, Maucort-Boulch D, Plummer M, Franceschi S. World-wide relative contribution of hepatitis B and C viruses in hepatocellular carcinoma. Hepatology 2015;62:1190e200. [2] Su CH, Lin Y, Cai L. Genetic factors, viral infection, other factors and liver cancer: an update on current progress. Asian Pac J Cancer Prev 2013;14:4953e60. [3] Li X, Liu X, Tian L, Chen Y. Cytokine-mediated immunopathogenesis of hepatitis B virus Infections. Clin Rev Allergy Immunol 2016;50:41e54. [4] Haybaeck J, Zeller N, Wolf MJ, Weber A, Wagner U, Kurrer MO, et al. A lymphotoxin-driven pathway to hepatocellular carcinoma. Cancer Cell 2009;16:295e308. [5] Aggarwal BB, Gupta SC, Kim JH. Historical perspectives on tumor necrosis factor and its superfamily: 25 years later, a golden journey. Blood 2012;119:651e65. [6] Luedde T, Schwabe RF. NF-kB in the liverdlinking injury, fibrosis and hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2011;8:108e18. [7] Sghaier I, Zidi S, Mouelhi L, Honold G, Blo ¨mer K, Pape GR, et al. The relationship between TNF alpha gene polymorphisms (-238/-308), TNF RII VNTR (p75) and outcomes of hepatitis B virus infection in Tunisian population. Gene 2015;568:140e5. [8] Wang B, Wang J, Zheng Y, Zhou S, Zheng J, Wang F, et al. A study of TNF-alpha-238 and -308 polymorphisms with different outcomes of persistent hepatitis B virus infection in China. Pathology 2010;42:674e80. [9] Saxena R, Chawla YK, Verma I, Kaur J. IFN-g (þ874) and not TNF-a (-308) is associated with HBV-HCC risk in India. Mol Cell Biochem 2014;385:297e307. [10] Panigrahi R, Sarkar N, Biswas A, Pal A, Saha D, Singh SP, et al. Association of TNF-a promoter polymorphism with HBV
[11]
[12]
[13]
[15]
[16]
[17] [18]
[20]
[21]
[22]
[23]
[24]
[25] [26]
[27]
associated disease outcome among HBV infected patients from Orissa, southern part of east India. Clin Exp Hepatol 2014;4:202e8. Jin YJ, Lee D, Chung YH, Kim JA, Kim SE, Lee YS, et al. Tumor necrosis factor-alpha gene polymorphism associated with development of hepatitis B virus-associated hepatocellular carcinoma. J Clin Gastroenterol 2015;49:e76e81. Messer G, Spengler U, Jung MC, Honold G, Blo ¨mer K, Pape GR, et al. Polymorphic structure of the tumor necrosis factor (TNF) locus: an NcoI polymorphism in the first intron of the human TNF-beta gene correlates with a variant amino acid in position 26 and a reduced level of TNF-beta production. J Exp Med 1991;173:209e19. Szlosarek P, Charles KA, Frances R, Balkwill FR. Tumour necrosis factor-ɑ as a tumor promoter. Eur J Cancer 2006;42:5e50. Tsai JF, Chang WY, Jeng JE, Ho MS, Wang LY, Hsieh MY, et al. Hepatitis C virus infection as a risk factor for nonalcoholic liver cirrhosis in Taiwan. J Med Virol 1993;41:296e300. Jeng JE, Tsai JF, Chuang LY, Ho MS, Lin ZY, Hsieh MY, et al. Tumor necrosis factor-a 308.2 polymorphism is associated with advanced hepatic fibrosis and higher risk for hepatocellular carcinoma. Neoplasia 2007;9:987e92. Tsuchiya N, Tokushige K, Yamaguchi N, Hasegawa K, Hashimoto E, Yamauchi K, et al. Influence of TNF gene olymorphism in patients with acute and fulminant hepatitis. J Gastroenterol 2004;39:859e66. Rothman KJ. Interactions between causes. In: Modern epidemiology. Boston: Little brown and company; 1986. p. 311e26. Lu SN, Wang JH, Liu SL, Hung CH, Chen CH, Tung HD, et al. Thrombocytopenia as a surrogate for cirrhosis and a marker for the identification of patients at high-risk for hepatocellular carcinoma. Cancer 2006;107:2212e22. Tarhuni A, Guyot E, Rufat P, Sutton A, Bourcier V, Grando V, et al. Impact of cytokine gene variants on the prediction and prognosis of hepatocellular carcinoma in patients with cirrhosis. J Hepatol 2014;61:342e50. Elahi MM, Asotra K, Matata BM, Mastana SS. Tumor necrosis factor alpha-308 gene locus promoter polymorphism: an analysis of association with health and disease. Biochim Biophys Acta 2009;1792:163e72. Naslednikova IO, Konenkov VI, Ryazantseva NV, Novit ski VV, Tkachenko SB, Zima AP, et al. Role of genetically determined production of immunoregulatory cytokines in immunopathogenesis of chronic viral hepatitides. Bull Exp Biol Med 2007; 143:706e12. Jung MC, Spengler U, Schraut W, Hoffman R, Zachova R, Eisenburg G, et al. Hepatitis B virus antigen-specific T-cell activation in patients with acute and chronic hepatitis B. J Hepatol 1991;13:310e7. Bortolami M, Venturi C, Giacomelli L, Scalerta R, Bacchetti S, Marino F, et al. Cytokine, infiltrating macrophage and T cellmediated response to development of primary and secondary human liver cancer. Dig Liver Dis 2002;34:794e801. Lee SH, Park SG, Lim SO, Jung G. The hepatitis B virus X protein up-regulates lymphotoxin alpha expression in hepatocytes. Biochem Biophys Acta 2005;1741:75e84. Nakazaki H. Preoperative and postoperative cytokines in patients with cancer. Cancer 1992;70:709e13. Feitelson MA, Reis HM, Tufan NL, Sun B, Pan J, Lian Z. Putative roles of hepatitis B x antigen in the pathogenesis of chronic liver disease. Cancer Lett 2009;286:69e79. Lackner C, Struber G, Liegl B, Leibl S, Ofner P, Bankuti C, et al. Comparison and validation of simple noninvasive tests for prediction of fibrosis in chronic hepatitis C. Hepatology 2005;41:1376e82.
Please cite this article in press as: Tsai J-F, et al., Independent and additive interaction between polymorphisms of tumor necrosis factor a308 and lymphotoxin aþ252 on risk of hepatocellular carcinoma related to hepatitis B, Kaohsiung Journal of Medical Sciences (2017), http://dx.doi.org/10.1016/j.kjms.2017.04.009
