Note of clarification of data in the paper entitled “Interferon gamma +874 T/A polymorphism increases the risk of cervical cancer: evidence from a meta-analysis”

Abstract

We read with great interest the paper entitled “Interferon gamma +874 T/A polymorphism increases the risk of cervical cancer: evidence from a meta-analysis” published online by Sun et al. Their results suggest that interferon gamma (IFNG) gene +874 T/A polymorphism might contribute to women’s susceptibility to cervical cancer. They also found that IFNG +874 T/A polymorphism is associated with increased cervical cancer risk in Asian female population. The result is encouraging. Nevertheless, several key issues are worth noticing. We re-evaluate the association between IFNG +874 T/A polymorphism and cervical cancer risk by performing an updated meta-analysis based on 2777 cases and 2542 controls of 11 studies. We found that IFNG +874 T/A polymorphism was not significantly associated with cervical cancer risk in overall population. We also observed that the polymorphism was associated with enhanced cervical cancer risk in Asian population and was relevant to increased squamous cell cervical cancer risk.

Keywords

Cervical cancer cervical intraepithelial neoplasia interferon gamma meta-analysis polymorphism

Dear Editor

With great interest, we recently read the article entitled “Interferon gamma +874 T/A polymorphism increases the risk of cervical cancer: evidence from a meta-analysis” published online by Sun et al.¹ In the paper, the authors conducted a meta-analysis to assess the association between interferon gamma (IFNG) gene +874 T/A polymorphism (rs2430561) and cervical cancer risk on the basis of a total of nine studies published before January 10, 2014, including 1116 cases and 1290 controls. In overall population, they observed that IFNG +874 T/A polymorphism was associated with elevated cervical cancer risk in the dominant genetic model (TA + AA vs TT) and in the TA versus TT model, with odds ratio (OR) (95% confidence interval (CI)) being equal to 1.399 (1.097–1.784) and 1.471 (1.137–1.903), respectively. In their subsequent stratified analyses by ethnicity, Sun et al. found significant association of IFNG +874 T/A variant with enhanced cervical cancer risk in Asian population (OR (95% CI) = 1.455 (1.062–1.933) in the dominant model and 1.494 (1.069–2.087) for TA vs TT model). The findings provided by Sun et al. are encouraging.

Nevertheless, after careful examination of the data supplied by Sun et al. (shown in Table 2 in their original text),¹ we found that several key issues should be noticed. First, two eligible case-control studies concerning IFNG +874 T/A polymorphism and cervical cancer risk,^2,3 published before 2014, were not included in Sun et al.’s analysis.¹ Second, three studies included in Sun et al.’s paper examined the association between IFNG +874 T/A polymorphism and cervical intraepithelial neoplasia (CIN) risk. It is widely recognized that CIN is a precancerous lesion of cervical cancer; however, not all patients with CIN will eventually develop into cervical cancer, and thus, it is appropriate to explore the relationship of IFNG +874 T/A polymorphism with cervical cancer and/or CIN risk, respectively. Third, two eligible studies about IFNG +874 T/A polymorphism and cervical cancer risk have been published after 2014.^4,5 Therefore, the conclusions induced from Sun et al.’s data are not entirely reliable. It is necessitated to summarize the association between IFNG +874 T/A polymorphism and cervical cancer/CIN risk more comprehensively and objectively.

A comprehensive search was carried out in the databases of MEDLINE/PubMed, Science Direct, Embase, China National Knowledge Infrastructure (CNKI), and Wanfang Medical Online with a combination of the following terms: “cervical cancer,” “cervical neoplasm” or “cervical carcinoma” and “Interferon-gamma” or “IFN-γ” or “IFNG” or “rs2430561” and “polymorphism” or “variant.” The references cited in the publications and review articles were also manually searched. Last search was updated on September 30, 2016.

Data inclusion criteria were as follows: (1) the papers reporting cervical cancer risk and IFNG +874 T/A polymorphism, (2) case-control studies or cohort studies, and (3) sufficient data to estimate the OR and 95% CI. For overlapping studies, the results including more information were included. Accordingly, papers lacking essential information and review articles were excluded.

We reassess the association between IFNG +874 T/A polymorphism and cervical cancer/CIN risk by performing an updated meta-analysis on the basis of a total of 14 studies containing 3068 cases and 2755 controls.^2–13 The Cochrane Q statistics test was used to examine the heterogeneity among studies. A fixed-effect model or a random-effect model was adopted to estimate the combined effects according to the result of a heterogeneity test.¹⁴ A fixed-effect model is chosen while the effects are assumed to be homogeneous; otherwise, a random-effect model is used. We first examine the relationship between IFNG +874 T/A polymorphism and cervical cancer/CIN risk, that is, pooling the cervical cancer and CIN cases together. Then, subgroup meta-analyses were conducted to investigate the association of IFNG +874 T/A polymorphism with cervical cancer risk or CIN risk, respectively. Further stratified analyses were also carried out by ethnicity, Hardy–Weinberg equilibrium (HWE) of the control group, and histological types of cervical cancer. Cumulative meta-analysis was conducted to investigate the tendency of results by accumulating single study year by year, which could be used to determine whether or not new relevant studies are needed. Sensitivity analysis was implemented via sequentially removing each study per time from the meta-analysis, which was employed to evaluate the stability of the results of current meta-analysis. Deviation from HWE in control subjects may bias the estimates of genetic effects in genetic association studies and meta-analysis¹⁵; thus, the Pearson chi-square (χ²) test was used to examine whether the genotype distribution in each control group is deviated from HWE. The funnel plot was drawn to assess publication bias visually. In addition, Begg’s test and Egger’s test were used to examine the publication bias.^16,17 All the statistical analyses were performed on the STATA/SE 10.1 software package (Stata Corporation, College Station, TX). Statistical significance was determined as a two-sided p value less than 0.05 for any test or model.

The general characteristics of the included studies in this meta-analysis were shown in Table 1. The summary ORs for the association of IFNG +874 T/A polymorphism with cervical cancer/CIN risk were listed in Table 2. Overall, we found a significant association between IFNG +874 T/A polymorphism and cervical cancer/CIN risk in the dominant genetic model, with OR (95% CI) being equal to 1.262 (1.007–1.581; as shown in Figure 1 and Table 2). However, no significant associations of IFNG +874 T/A polymorphism with cervical cancer risk or CIN risk were observed, when stratified meta-analysis by disease category (dividing into cervical cancer and CIN) was performed, with OR (95% CI) being 1.240 (0.952–1.615) for cervical cancer and being 1.382 (0.837–2.282) for CIN in the dominant model (Table 2). Considering that the majority of the included studies explored the association between IFNG +874 T/A polymorphism and cervical cancer risk, we concentrated our analysis on this association based on 11 studies with 2777 cases and 2542 controls (Table 1). In the subsequent subgroup analysis by ethnicity of cervical cancer cases, we observed that IFNG +874 T/A polymorphism was significantly associated with elevated cervical cancer risk in Asian population but not in Caucasian and mixed population in the dominant model, with OR (95% CI) being equal to 1.390 (1.032–1.873), 1.221 (0.567–2.630), and 1.333 (0.879–2.021), respectively (Table 2). The finding in Asian population is similar to the result of Sun et al.¹ In the stratified analysis by histological types of cervical cancer, we found that IFNG +874 AA+TA genotypes were relevant to enhanced squamous cervical cancer risk, with OR (95% CI) being 1.566 (1.178–2.082) (Table 2). No significant associations between IFNG +874 T/A polymorphism and cervical cancer risk were found in stratified analysis based on the studies whose control groups were in HWE (Table 2).

Table 1.

Characteristics of the included studies of the meta-analysis of INFG +874 T/A polymorphism and cervical cancer risk.

First author (Ref. No.)	Year	Ethnicity	Category	Quality score	Cases genotype count		Control genotype count		HWE (p)
First author (Ref. No.)	Year	Ethnicity	Category	Quality score	n	TT/TA/AA	n	TT/TA/AA	HWE (p)
Govan⁶	2003	African	CC	8	96	5/30/61	140	7/31/102	0.033
Govan⁶	2003	Mixed	CC	8	165	14/66/85	265	26/81/158	0.002
Fernandes⁷	2008	Caucasian	CIN	7	42	5/19/18	87	16/28/43	0.007
Guzman²	2008	Mixed	SCC	5	229	27/103/99	193	33/93/67	0.940
Kordi⁸	2008	Asian	CC	6	200	24/125/51	200	25/75/100	0.072
Gangwar⁹	2009	Asian	SCC	6	200	23/91/86	230	45/115/70	0.856
von Linsingen¹⁰	2009	Mixed	CIN	8	96	13/45/38	50	11/24/15	0.811
Ivansson³	2010	Caucasian	CC	9	1290	286/650/354	802	157/371/274	0.121
Wang¹¹	2011	Asian	SCC	6	186	21/84/81	200	38/96/66	0.769
do Carmo¹²	2012	Mixed	CC	6	9	2/3/4	76	13/46/17	0.062
do Carmo¹²	2012	Mixed	CIN	7	153	25/81/47	76	13/46/17	0.062
Maskoen¹³	2013	Asian	CC	7	64	9/32/23	42	3/22/17	0.320
Zidi⁴	2014	Caucasian	CC	6	122	32/65/25	160	64/68/28	0.186
Yuan⁵	2015	Asian	CC	9	216	9/50/157	234	9/63/162	0.363
Sum (CIN)	–	–	–	–	291	43/145/103	213	40/98/75	–
Sum (SCC)	–	–	–	–	790	92/387/311	823	141/379/303	–
Sum (CC)	–	–	–	–	2777	452/1299/1026	2542	420/1061/1061	–

HWE: Hardy–Weinberg equilibrium; CC: cervical cancer; CIN: cervical intraepithelial neoplasia; SCC: squamous cell carcinoma.

Table 2.

Summary ORs for the association between INFG +874 T/A variant and cervical cancer or cervical intraepithelial neoplasia risk.

Comparisons	Sample size	No. of studies	Hypothesis tests			Heterogeneity tests			Publication bias test (p)
Comparisons	Case/control	No. of studies	OR (95% CI)	Z	p	χ² (df)	p	I² (%)	Begg’s test	Egger’s test
Cervical cancer and cervical intraepithelial neoplasia
A vs T	6136/5510	14	1.067 (0.891–1.278)	0.71	0.479	52.90 (13)	<0.001	75.4	0.622	0.230
AA vs TA + TT	3068/2755	14	1.018 (0.776–1.334)	0.13	0.899	59.93 (13)	<0.001	78.3	0.702	0.228
TA + AA vs TT	3068/2755	14	1.262 (1.007–1.581)	2.02	0.044	21.14 (13)	0.070	38.5	0.025	0.244
Cervical intraepithelial neoplasia
A vs T	582/426	3	1.216 (0.931–1.588)	1.44	0.150	1.00 (2)	0.605	<0.1	0.602	0.846
AA vs TA + TT	291/213	3	1.250 (0.835–1.871)	1.08	0.279	2.36 (2)	0.307	15.2	0.117	0.580
TA + AA vs TT	291/213	3	1.382 (0.837–2.282)	1.26	0.206	0.97 (2)	0.616	<0.1	0.602	0.412
Cervical cancer
A vs T	5554/5084	11	1.037 (0.842–1.278)	0.34	0.733	49.47 (10)	<0.001	79.8	0.586	0.379
AA vs TA + TT	2777/2542	11	0.977 (0.718–1.330)	0.15	0.882	55.13 (10)	<0.001	81.9	0.815	0.363
TA + AA vs TT	2777/2542	11	1.240 (0.952–1.615)	1.59	0.111	19.49 (10)	0.034	48.7	0.036	0.417
Asian cervical cancer
A vs T	1732/1812	5	1.050 (0.706–1.561)	0.24	0.809	28.14 (4)	<0.001	85.8	0.327	0.695
AA vs TA + TT	866/906	5	0.988 (0.530–1.843)	0.04	0.971	37.37 (4)	<0.001	89.3	0.142	0.789
TA + AA vs TT	866/906	5	1.390 (1.032–1.873)	2.17	0.030	6.08 (4)	0.193	34.2	0.050	0.088
Mixed population cervical cancer
A vs T	806/1068	3	1.105 (0.756–1.614)	0.51	0.607	5.09 (2)	0.079	60.7	0.602	0.921
AA vs TA + TT	403/534	3	1.174 (0.620–2.224)	0.49	0.623	7.74 (2)	0.021	74.2	0.602	0.668
TA + AA vs TT	403/534	3	1.333 (0.879–2.021)	1.35	0.176	0.92 (2)	0.632	<0.1	0.117	0.292
Caucasian cervical cancer
A vs T	2824/1924	2	1.055 (0.628–1.771)	0.20	0.840	8.37 (1)	<0.001	88.1	0.317	–
AA vs TA + TT	1412/962	2	0.867 (0.539–1.392)	0.59	0.554	2.54 (1)	0.111	60.6	0.317	–
TA + AA vs TT	1412/962	2	1.221 (0.567–2.630)	0.51	0.609	7.64 (1)	0.006	86.9	0.317	–
Cervical cancer risk studies (in HWE)
A vs T	5032/4274	9	1.101 (0.862–1.405)	0.77	0.442	46.90 (8)	<0.001	82.9	0.677	0.267
AA vs TA + TT	2516/2137	9	1.066 (0.737–1.543)	0.34	0.733	52.33 (8)	<0.001	84.7	0.677	0.306
TA + AA vs TT	2516/2137	9	1.262 (0.925–1.720)	1.47	0.142	19.41 (8)	0.013	58.8	0.095	0.408
Squamous cell carcinoma risk
A vs T	1580/1646	4	1.165 (0.762–1.783)	0.71	0.480	26.39 (3)	<0.001	88.6	0.174	0.234
AA vs TA + TT	790/823	4	1.080 (0.530–2.202)	0.21	0.832	36.18 (3)	<0.001	91.7	0.497	0.105
TA + AA vs TT	790/823	4	1.566 (1.178–2.082)	3.09	0.002	2.34 (3)	0.504	<0.1	0.174	0.188

OR: odds ratio; CI: confidence interval; df: degrees of freedom; HWE: Hardy–Weinberg equilibrium.

Figure 1.

Forest plots for the association between IFNG +874 T/A polymorphism and cervical cancer or cervical intraepithelial neoplasia risk.

The cumulative meta-analysis accumulated the studies in the sequence of year of publications. As illustrated in Figure 2, significant association between IFNG +874 T/A polymorphism and cervical cancer risk was observed for the first time in 2009, when Gangwar et al.’s study was published and a meta-analysis was performed then. We also examined that the summary ORs became statistically insignificant when Ivansson et al.’s study was published in 2010; however, there seemed to be a tendency of significance of association between IFNG +874 T/A polymorphism and cervical cancer risk in the dominant model (Figure 2). Sensitivity analysis was performed to evaluate the stability of the result of the present meta-analysis. As shown in Figure 3, all studies except one did not alter the result of being no significant association between IFNG +874 T/A polymorphism and cervical cancer risk. The exception was Ivansson et al.’s study. The result displayed that the association between IFNG +874 T/A polymorphism and cervical cancer risk became statistically significant when Ivansson et al.’s study was excluded from the current meta-analysis (Figure 3). This seemed to explain why Sun et al.’s study provided a positive association¹; as mentioned above, Ivansson et al.’s study was erroneously not included in Sun et al.’s meta-analysis. Meanwhile, the results of sensitivity analysis also revealed that Ivansson et al.’s study seemed to be a main source of heterogeneity of our current meta-analysis, because the statistical significance of heterogeneity (χ² = 19.49, df = 10, p = 0.034, I² = 48.7%; Table 2 and Figure 1) altered to be insignificant (χ² = 8.66, df = 9, p = 0.470, I² < 0.1%) and the OR (95% CI) was 1.432 (1.155–1.776) in the fixed-effect model for IFNG +874 AA + TA versus TT when Ivansson et al.’s study was excluded from our present meta-analysis.

Figure 2.

Cumulative meta-analysis for the association between IFNG +874 T/A polymorphism and cervical cancer risk.

Figure 3.

Sensitivity analysis for the association between IFNG +874 T/A polymorphism and cervical cancer risk.

The shape of the funnel plot seemed to be approximately symmetrical among total population (Figure 4). Both Begg’s and Egger’s tests suggested that there was no significant publication bias in current meta-analysis with the exception of the ORs for TA + AA versus TT in overall and cervical cancer comparison group (p = 0.025 and 0.036 in Begg’s test); however, the OR for TA + AA versus TT in cervical cancer risk comparison was not statistically significant, as shown in Table 2. These results indicated that publication biases might have few effects on our current meta-analysis.

Figure 4.

Funnel plot of the association between IFNG +874 T/A polymorphism and cervical cancer risk.

In conclusion, the results supplied by Sun et al.¹ should be taken into account with caution. Further well-designed studies with large sample size are needed to reach a definitive conclusion about the relationship between IFNG +874 T/A polymorphism and cervical cancer risk. We hope our remark will help for more accurate elaboration and substantiation of the findings presented by Sun et al.¹

Footnotes

Acknowledgements

H.Y. and M.Y. contributed equally to this article.

Declaration of Conflicting Interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This work was financially supported by the scientific research project of Hubei Provincial Department of Education (No. Q20152008).

References

Sun

Pen

. Interferon gamma +874 T/A polymorphism increases the risk of cervical cancer: evidence from a meta-analysis. Tumour Biol 2015; 36(6): 4555–4564.

Guzman

Yambartsev

Goncalves-Primo

. New approach reveals CD28 and IFNG gene interaction in the susceptibility to cervical cancer. Hum Mol Genet 2008; 17(12): 1838–1844.

Ivansson

Juko-Pecirep

Gyllensten

UB.

Interaction of immunological genes on chromosome 2q33 and IFNG in susceptibility to cervical cancer. Gynecol Oncol 2010; 116(3): 544–548.

Zidi

Benothmen

Sghaier

. Association of IL10-1082 and IFN-γ+874 polymorphisms with cervical cancer among Tunisian women. ISRN Genetics 2014; 2014: 706516.

Yuan

Fan

Yao

. Association study of single-nucleotide polymorphisms of STAT2/STAT3/IFN-γ genes in cervical cancer in Southern Chinese Han women. Asian Pac J Cancer Prev 2015; 16(8): 3117–3120.

Govan

Carrara

Sachs

. Ethnic differences in allelic distribution of IFN-g in South African women but no link with cervical cancer. J Carcinog 2003; 2(1): 3.

Fernandes

Goncalves

Simoes

. A pilot case-control association study of cytokine polymorphisms in Brazilian women presenting with HPV-related cervical lesions. Eur J Obstet Gynecol Reprod Biol 2008; 140(2): 241–244.

Kordi Tamandani

Sobti

Shekari

. Expression and polymorphism of IFN-gamma gene in patients with cervical cancer. Exp Oncol 2008; 30(3): 224–229.

Gangwar

Pandey

Mittal

RD.

Association of interferon-gamma +874A polymorphism with the risk of developing cervical cancer in north-Indian population. BJOG 2009; 116(12): 1671–1677.

10.

von Linsingen

Bompeixe

Maestri

. IFNG (+874 T/A) polymorphism and cervical intraepithelial neoplasia in Brazilian women. J Interferon Cytokine Res 2009; 29(5): 285–288.

11.

Wang

Zhang

Walayat

. Association between cytokine gene polymorphisms and cervical cancer in a Chinese population. Eur J Obstet Gynecol Reprod Biol 2011; 158(2): 330–333.

12.

do Carmo Vasconcelos de Carvalho

de Macedo

de Lima

. IFN-gamma and IL-12B polymorphisms in women with cervical intraepithellial neoplasia caused by human papillomavirus. Mol Biol Rep 2012; 39(7): 7627–7634.

13.

Maskoen

Susanto

IFNG Polymorphism (+874 T>A) is not a risk factor for cervical cancer. Universa Medicina 2013; 32(1): 29–36.

14.

DerSimonian

Laird

Meta-analysis in clinical trials. Control Clin Trials 1986; 7(3): 177–188.

15.

Zintzaras

Impact of Hardy-Weinberg equilibrium deviation on allele-based risk effect of genetic association studies and meta-analysis. Eur J Epidemiol 2010; 25(8): 553–560.

16.

Begg

Mazumdar

Operating characteristics of a rank correlation test for publication bias. Biometrics 1994; 50(4): 1088–1101.

17.

Egger

Davey Smith

Schneider

. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997; 315(7109): 629–634.