Interleukin-16 Gene Polymorphisms rs4778889,rs4072111,rs11556218,and Cancer Risk in Asian Populations: A Meta-Analysis

Abstract

Objectives: Some polymorphisms of the interleukin-16 (IL-16) gene are associated with various cancers. To resolve inconsistencies in published data, we performed a meta-analysis of studies of IL-16 polymorphisms and cancer risk. Materials and Methods: Seven eligible studies pooling 1678 cases and 1937 controls were quantitatively analyzed to evaluate three IL-16 polymorphisms (rs4778889, rs4072111, rs11556218) and cancer risk. Hardy-Weinberg equilibrium (HWE) for controls was evaluated by goodness-of-fit chi-squared tests. Odds ratios (ORs) and their 95% confidence intervals (CIs) were calculated for each genetic model and allelic comparison. Data were pooled using fixed- or random-effects models depending on heterogeneity across studies. Results: Our meta-analysis demonstrated that the IL-16 polymorphism rs11556218 was significantly associated with increased susceptibility to cancer in several models, including allelic contrast (OR=1.307; 95% CI, 1.108-1.541), heterozygote contrast (OR=1.650; 95% CI, 1.424-1.911), and dominant model (OR=1.605; 95% CI, 1.391-1.845). The result remained consistent after adjustment for age and gender. No significant association was found between IL-16 polymorphisms rs4778889 rs4072111 and cancer risk. Conclusions: The rs11556218 T/G polymorphism of the IL-16 gene was significantly associated with elevated cancer risk in Asian populations. Our results warrant larger, better-designed studies, including a greater ethnic variety.

Introduction

Cancer is associated with significant morbidity and mortality all over the world. A recent epidemiologic study in the United States revealed that one in three female Americans and one in two male Americans will suffer from cancer during their lifetime (Siegel et al., 2012). However, its etiology is far from being fully elucidated. Carcinogenesis might be related to multiple genes and environmental factors. Inflammation has been demonstrated to be one of the major hallmarks of multiple cancers (Guven Maiorov et al., 2013). It has been estimated that 25% of cancer cases worldwide involve a history of previous chronic infection and inflammation (Hussain and Harris, 2007).

Interleukin (IL)-16 is a proinflammatory cytokine that has a wide array of functions in multiple immunopathobiological processes (Yellapa et al., 2012). The gene encoding for IL-16 is located on chromosome 15q 26.1-3 (Cruikshank et al., 2000). IL-16 is initially expressed as a precursor protein and then activated by caspase 3 through cleavage of the 631 amino acid proprotein to form a C-terminal 121-amino acid peptide (Center et al., 1996; Chupp et al., 1998). The cellular source for IL-16 contains a group of inflammatory cells, including cytotoxic T lymphocytes (CD8+ T cells), monocytes/macrophages, dendritic cells, mast cells, epithelial cells, and fibroblasts (Glass et al., 2006; Gao et al., 2009b; Goihl et al., 2012). IL-16 stimulates the production of a variety of proinflammatory cytokines by monocytes with related biological sequelae (Mathy et al., 2000). All of these cytokines, including IL-6, IL-1β, IL-15, and tumor necrosis factor-α, have been demonstrated to be positively related to development and progression of cancer (Gao et al., 2009b). Evidence of increased IL-16 levels in serum or plasma has also been recognized in multiple malignant tumors in both preclinical and clinical trials (Kovacs, 2001; Bellomo et al., 2007; Gao et al., 2009b).

Polymorphisms (single-nucleotide polymorphisms; SNPs) of the IL-16 gene will alter the transcription level and affect the function of IL-16, eventually exhibiting differentiated expression in cancers. Recently, a variety of studies have highlighted some polymorphisms of the IL-16 gene with regard to susceptibility to a wide range of cancers (Gao et al., 2009a, 2009b; Zhu et al., 2010; Azimzadeh et al., 2011, 2012; Li et al., 2011). Three polymorphisms (rs4778889, rs4072111, rs11556218) were substantially explored. However, published data on the association of these polymorphisms and potential cancer risk remain inconsistent, partially because of the comparatively small sample size, the heterogeneity of cancer type, and differences in anthropometric data between cases and controls. To clarify whether the association of IL-16 polymorphisms with cancer risk may contribute to cancer prevention and interventions, we performed a comprehensive literature review and meta-analysis based on eligible published data.

Materials and Methods

Search strategy

We conducted a systematic search of studies addressing the association between IL-16 polymorphisms and cancer risk. Four biomedical electronic databases were retrieved without language limitations (updated on July, 2013): PubMed (US National Library of Medicine), Embase, Web of Knowledge, and CNKI (China National Knowledge Infrastructure). The key words and subject terms involved in the search strategy were as follows: “IL-16,” “interleukin-16,” “polymorphism,” “variant,” “alleles,” “single-nucleotide polymorphisms,” “SNP,” “cancer,” “neoplasms,” “carcinoma,” “tumor.” In addition, potential additional eligible studies were sought through a manual search of citations of included studies. The full search strategy implemented is shown in Fig. 1.

FIG. 1.

Diagrams for inclusion of studies in this meta-analysis.

Inclusion and exclusion criteria

Inclusion criteria were set to literature selections as follows: first, original case-control or cohort design articles concerning the association between IL-16 polymorphisms and cancers; second, studies should contain sufficient data on genotypes or allelic frequencies; and third, adequate data were available from which to calculate pooled odds ratios (ORs) and their corresponding 95% confidence intervals (CIs). The exclusion criteria included lack of control subjects in studies; article type of abstracts, reviews and comments; and studies not containing adequate information.

Data extraction

Data from all publications complying with the criteria described above were extracted by two authors independently. Any conflicts between evaluations were resolved by discussion until the authors reached a final agreement. For each included study, the following information was extracted: the last name of the first author, year of publication, country and ethnicity of the study population, source of control groups, genotyping method, gender and age distributions, and allele frequency in cases and controls.

Statistical analyses

Statistical analyses were undertaken using STATA version 11.1 (Stata Corp., College Station, TX) with two-sided p-values. Hardy-Weinberg equilibrium (HWE) for the controls was evaluated in each study by a chi-squared test for goodness-of-fit, and p<0.05 was regarded as evidence of inequality of genetic distributions. The minor allele frequency for the controls was also assessed to detect the least common allele occurring in reported populations. Genetic comparisons of extracted data used allelic contrast, homozygote contrast, heterozygote contrast, and dominant and recessive models. ORs and their 95% CIs were measured for each study. The statistical significance for pooled ORs was determined using the Z-test, and p_Z<0.05 was considered significant. The Cochran Q statistic was applied to evaluate the heterogeneity across studies (Egger et al., 1997b). When a significant Q statistic was achieved (p<0.1), a random-effects model (the DerSimonian-Laird method) was selected to calculate the pooled ORs (DerSimonian and Laird, 1986); otherwise, a fixed-effects model (the Mantel-Haenszel method) was applied (Donald and Donner, 1987). In this study, we also introduced another measurement, I², for more quantitative analysis of potential heterogeneity. I², which stands for the proportion of interstudy variability attributed to heterogeneity, ranges from 0% to 100% (Higgins and Thompson, 2002). High risk of heterogeneity is assured if I² is larger than 75%. The Egger's linear regression test was conducted to verify the bias of publication (Egger et al., 1997a), and p<0.1 represents the existence of publication bias.

Results

Characteristics of included studies

A total of 49 relevant articles were obtained using the search strategy described above. Among them, 42 were excluded based on titles and abstracts. Of those, 17 studies were duplicated; 21 were not about IL-16 polymorphisms and cancer risk; 3 studies were abstracts of committee reports or comments; 1 was a review article. Then, seven studies were left for full-text review. Further, one study was removed from duplication, another one was treated as two separate studies, for it contained two different cancer types with different case samples (Gao et al., 2009b). Finally, seven studies meeting the inclusion and exclusion criteria highlighted the association of three variants in the IL-16 gene and cancer risk (Table 1). Six were written in English, while the other one was written in Chinese. Among all the studies, seven were pooled for rs4778889 variants; six for rs4072111; and six for rs11556218. All of the studies had a case-control design with a population of Asian origin. Two were for colorectal cancer; two for nasopharyngeal carcinoma; and three for other types of cancer. All of the genotyping used PCR restriction fragment length polymorphism analysis. As shown in Tables 2, 3, and 4, genetic distributions of IL-16 gene polymorphisms are presented. HWE and minor allele frequency for controls were also calculated. Few articles for rs11556218 conformed to HWE; however, most of the articles for rs4778889 and rs4072111 showed goodness-of-fit.

Table 1.

Characteristics of Studies Enrolled in This Meta-Analysis

								Number of population		Age		Gender (M/F)		Smoking%(past+current)
ID	Study	Year of publications	Study design	Country	Ethnicity	Source of controls	Type of cancer	Case	Control	Case	Control	Case	Control	Case	Control
1	Huang et al.	2012	Case-control	China	Asian	HB	NPC	75	75	NA	NA	NA	NA	NA	NA
2	Azimzadeh et al.	2011	Case-control	Iran	Asian	HB	CRC	260	405	45.3±16.5	56.3±12	139/121	196/209	13.8	11.6
3	Li et al.	2011	Case-control	China	Asian	PB	HCC	206	264	NA	NA	NA	NA	NA	NA
4	Zhu et al.	2010	Case-control	China	Asian	HB	RCC	335	340	55.1±11.4	54.6±12.7	222/113	215/125	37.9	35.6
5	Gao et al.	2009	Case-control	China	Asian	HB	NPC	206	373	48±13	46±11	154/52	266/107	NS	NS
6	Gao et al.	2009	Case-control	China	Asian	HB	CRC	376	480	60.3±13.2	58.8±11.2	226/150	306/174	61.2	58.8
7	Gao et al.	2009	Case-control	China	Asian	HB	GC	220	480	57.9±11.9	58.8±11.2	146/74	306/174	62.7	58.8

HB, hospital based; PB, population based; NPC, nasopharyngeal cancer; CRC, colorectal cancer; HCC, hepatocellular carcinoma; RCC, renal cell carcinoma; GC, gastric cancer; PCR-RFLP, polymerase chain reaction-restriction fragment length polymorphism; NA, not available.

Table 2.

Genotype Distributions of Interleukin-16 Polymorphism (rs4778889) of Enrolled Studies

		Case			Control			Case		Control
ID	Study	TT	CT	CC	TT	CT	CC	T	C	T	C	p_HWE	χ²	MAF(%)
1	Huang et al.	39	36	0	49	26	0	114	36	124	26	0.07	3.3	17.33
2	Azimzadeh et al.	178	73	9	274	112	19	429	91	660	150	0.09	2.83	18.52
3	Li et al.	158	42	6	182	76	6	358	54	440	88	0.55	0.35	16.67
4	Zhu et al.	199	122	14	171	135	34	520	150	477	203	0.34	0.92	29.85
5	Gao et al.	131	65	10	228	128	17	327	85	584	162	0.86	0.03	21.72
6	Gao et al.	246	119	11	294	164	22	611	141	752	208	0.89	0.02	21.67
7	Gao et al.	117	90	13	294	164	22	324	116	752	208	0.89	0.02	21.67

p<0.05 was regarded as significant.

MAF(%)<5% was regarded as inequality of genetic distributions.

HWE, Hardy-Weinberg equilibrium; χ², chi-squared; MAF, minor allele frequency.

Table 3.

Genotype Distributions of Interleukin-16 Polymorphism (rs4072111) of Enrolled Studies

		Case			Control			Case		Control
ID	Study	CC	CT	TT	CC	CT	TT	C	T	C	T	p_HWE	χ²	MAF(%)
1	Huang et al.	41	34	0	44	31	0	116	34	119	31	0.02	5.09	20.67
2	Azimzadeh et al.	196	56	8	324	77	4	448	72	725	85	0.81	0.06	10.49
3	Li et al.	110	80	16	136	104	24	300	112	376	152	0.52	0.4	28.79
5	Gao et al.	111	87	8	221	139	13	309	103	581	165	0.11	2.49	22.12
6	Gao et al.	235	123	18	283	179	18	593	159	745	215	0.11	2.55	22.40
7	Gao et al.	144	72	4	283	179	18	360	80	745	215	0.11	2.55	22.40

p<0.05 was regarded as significant.

MAF(%)<5% was regarded as inequality of genetic distributions.

Table 4.

Genotype Distributions of Interleukin-16 Polymorphism (rs11556218) of Enrolled Studies

		Case			Control			Case		Control
ID	Study	TT	TG	GG	TT	TG	GG	T	G	T	G	p_HWE	χ²	MAF(%)
1	Huang et al.	32	37	6	46	26	3	101	49	118	32	0.78	0.08	21.33
2	Azimzadeh et al.	62	178	20	124	226	55	302	218	474	336	0.00	9.04	41.48
3	Li et al.	122	62	22	160	78	26	306	106	398	130	0.00	10.99	24.62
5	Gao et al.	91	109	6	210	151	12	291	121	571	175	0.01	6.05	23.46
6	Gao et al.	143	219	14	265	197	18	505	247	727	233	0.01	6.51	24.27
7	Gao et al.	94	112	14	265	197	18	300	140	727	233	0.01	6.51	24.27

p<0.05 was regarded as significant.

MAF(%)<5% was regarded as inequality of genetic distributions.

Main results of meta-analysis

The detailed results of the association between the IL-16 polymorphism rs4778889 and cancer risk are presented in Table 5. In all genetic models, the overall results of pooled data did not show any statistically significant associations. The pooled results of studies in HWE were consistent with the overall one, for no deviation of HWE was detected.

Table 5.

Meta-Analysis of the Association Between Interleukin-16 Polymorphism (rs4778889) and Cancer Risk

Genetic model	Variables	No. of study	Statistic model	OR	95% CI		Z	p_Z	I² (%)	Het χ²	p_Het
C vs. T	Total	7	R	0.925	0.763	1.121	0.80	0.426	65.0	17.15	0.009
	HWE	7	R	0.925	0.763	1.121	0.80	0.426	65.0	17.15	0.009
	CRC	2	F	0.873	0.727	1.050	1.44	0.149	0.0	0.35	0.556
	Others	5	R	0.957	0.714	1.282	0.30	0.768	75.9	16.58	0.002
TC vs. TT	Total	7	R	0.952	0.772	1.175	0.46	0.647	55.4	13.44	0.037
	HWE	7	R	0.952	0.772	1.175	0.46	0.647	55.4	13.44	0.037
	CRC	2	F	0.92	0.736	1.151	0.73	0.466	0.0	0.39	0.53
	Others	5	R	0.974	0.706	1.344	0.16	0.873	69.2	12.99	0.011
CC vs. TT	Total	6	R	0.771	0.488	1.220	1.11	0.267	50.0	10.01	0.075
	HWE	6	R	0.771	0.488	1.220	1.11	0.267	50.0	10.01	0.075
	CRC	2	F	0.654	0.377	1.132	1.52	0.129	0.0	0.13	0.724
	Others	4	R	0.856	0.416	1.761	0.42	0.673	68.7	9.58	0.023
CC+TC vs. TT	Total	7	R	0.933	0.749	1.164	0.61	0.541	62.5	16.01	0.014
	HWE	7	R	0.933	0.749	1.164	0.61	0.541	62.5	16.01	0.014
	CRC	2	F	0.886	0.715	1.099	1.10	0.270	0.0	0.41	0.521
	Others	5	R	0.925	0.688	1.355	0.2	0.838	74.1	15.47	0.000
CC vs. TC+TT	Total	6	F	0.745	0.546	1.016	1.86	0.063	38.6	8.15	0.148
	HWE	6	F	0.745	0.546	1.016	1.86	0.063	38.6	8.15	0.148
	CRC	2	F	0.671	0.389	1.157	1.43	0.151	0.0	0.07	0.789
	Others	4	R	0.871	0.459	1.653	0.42	0.673	61.5	7.78	0.051

p<0.05 was regarded as significant.

R, random-effect model; F, fixed-effect model.

As listed in Table 6, no evidence was found in the association of IL-16 polymorphism rs4072111 and cancer risk. Likewise, no statistical associations were revealed in studies restricted to HWE.

Table 6.

Meta-Analysis of the Association Between Interleukin-16 Polymorphism (rs4072111) and Cancer Risk

Genetic model	Variables	No. of study	Statistic model	OR	95%CI		Z	p_Z	I² (%)	Het χ²	p_Het
T vs. C	Total	6	F	0.993	0.879	1.120	0.12	0.905	42.5	8.70	0.122
	HWE	5	F	1.000	0.832	1.202	0.00	0.999	52.9	8.50	0.075
	Colorectal	2	F	1.107	0.757	1.617	0.52	0.601	71.4	3.5	0.061
	Others	4	F	0.915	0.811	1.033	1.43	0.153	25.5	6.71	0.243
TC vs. CC	Total	6	F	0.971	0.836	1.127	0.39	0.696	17.4	6.05	0.301
	HWE	5	F	0.960	0.824	1.119	0.52	0.604	29.8	5.70	0.223
	Colorectal	2	R	0.944	0.749	1.189	0.49	0.624	56.6	2.30	0.129
	Others	4	F	0.990	0.814	1.205	0.10	0.924	17.8	3.65	0.302
TT vs. CC	Total	5	F	1.036	0.719	1.492	0.19	0.851	39.7	6.63	0.157
	HWE	5	F	1.036	0.719	1.492	0.19	0.851	39.7	6.63	0.157
	Colorectal	2	F	1.544	0.864	2.762	1.47	0.143	50.8	2.03	0.154
	Others	3	F	0.793	0.490	1.284	0.94	0.345	0.8	2.02	0.365
TT+TC vs. CC	Total	6	F	0.981	0.849	1.134	0.26	0.796	34.7	7.66	0.176
	HWE	5	F	0.972	0.838	1.127	0.38	0.705	45.5	7.34	0.119
	Colorectal	2	R	1.041	0.694	1.561	0.19	0.848	67.5	3.08	0.079
	Others	4	F	0.954	0.816	1.117	0.58	0.561	35.1	4.63	0.201
TT vs. TC+CC	Total	5	F	1.055	0.736	1.512	0.29	0.771	33.8	6.04	0.196
	HWE	5	F	1.055	0.736	1.512	0.29	0.771	33.8	6.04	0.196
	Colorectal	2	F	1.608	0.903	2.862	1.61	0.107	39.0	1.64	0.201
	Others	3	F	0.616	0.34	1.116	1.6	0.11	35.5	3.1	0.212

p<0.05 was regarded as significant.

As shown in Table 7, evidence from six studies has demonstrated that the IL-16 polymorphism rs11556218 was significantly associated with an increased susceptibility to cancer in several genetic models, including allelic contrast (G vs. T: OR=1.307; 95% CI, 1.108-1.541; p_Z=0.001; p_Het=0.052), heterozygote contrast (TG vs. TT: OR=1.650; 95% CI, 1.424-1.911; p_Z=0.000; p_Het=0.17), and dominant model (TG+GG/TT: OR=1.605; 95% CI, 1.391-1.845; p_Z=0.000; p_Het=0.125). No potential associations were observed in either the homozygote contrast (GG vs. TT: OR=1.209; 95% CI, 0.893-1.638; p_Z=0.219; p_Het=0.223) or recessive model (GG vs. TT+TG: OR=0.930; 95% CI, 0.698-1.214; p_Z=0.622; p_Het=0.122). However, among all the studies concerning the rs11556218 polymorphism, only one conformed to HWE. After excluding the other studies, the results of association remained consistent. Statistically significant associations were found in allelic contrast, heterozygote contrast, and the dominant model (p<0.05).

Table 7.

Meta-Analysis of the Association Between Interleukin-16 Polymorphism (rs11556218) and Cancer Risk

Genetic model	Variables	No. of study	Statistic model	OR	95% CI		Z	p_Z	I² (%)	Het χ²	p_Het
G vs. T	Total	6	R	1.307	1.108	1.541	3.18	0.001	54.3	10.95	0.052
	HWE	1	F	1.789	1.065	3.005	2.20	0.028	NA	NA	NA
	Colorectal	2	R	1.018	0.815	1.273	1.10	0.272	84.9	6.63	0.010
	Others	4	R	1.341	1.122	1.601	3.24	0.001	25.3	4.02	0.260
TG vs. TT	Total	6	F	1.650	1.424	1.911	6.67	0.000	35.6	7.76	0.17
	HWE	1	F	2.046	1.042	4.016	2.08	0.038	NA	NA	NA
	Colorectal	2	F	1.861	1.491	2.323	5.49	0.000	24.1	1.32	0.251
	Others	4	F	1.498	1.230	1.824	4.02	0.000	31.3	4.37	0.224
GG vs. TT	Total	6	F	1.209	0.893	1.638	1.23	0.219	28.2	6.97	0.223
	HWE	1	F	2.875	0.669	12.349	1.42	0.156	NA	NA	NA
	Colorectal	2	F	0.948	0.6	1.499	0.23	0.82	50.9	2.03	0.154
	Others	4	F	1.477	0.983	2.220	1.88	0.061	0.0	2.97	0.397
TG+GG/TT	Total	6	R	1.605	1.391	1.845	6.49	0.000	42.1	8.63	0.125
	HWE	1	F	2.131	1.110	4.092	2.27	0.023	NA	NA	NA
	Colorectal	2	F	1.754	1.412	2.183	5.07	0.000	58.2	2.39	0.122
	Others	4	R	1.497	1.239	1.808	4.18	0.000	40.5	5.04	0.169
GG/TT+TG	Total	6	F	0.930	0.698	1.241	0.49	0.622	42.5	8.70	0.122
	HWE	1	F	2.087	0.502	8.675	1.01	0.311	NA	NA	NA
	Colorectal	2	F	0.658	0.431	1.007	1.93	0.054	47.4	1,90	0.168
	Others	4	F	1.283	0.864	1.912	1.24	0.216	0.0	1.90	0.593

p<0.05 was regarded as significant.

In the subgroup analysis stratified by cancer type, significant risk of colorectal cancer was found for the rs11556218 polymorphism in the heterozygote contrast (TG vs. TT: OR=1.861; 95% CI, 1.491-2.323; p_Z=0.000; p_Het=0.251) and the dominant model (TG+GG/TT: OR=1.754; 95% CI, 1.412-2.183; p_Z=0.000; p_Het=0.122). No other significant association was detected for either the rs4778889 or rs4072111 polymorphisms (p>0.05). In the subgroup analysis stratified with regard to gender (Table 8), significant results were identified in both male and female populations for the rs11556218 polymorphism in the allelic contrast or the dominant model. However, for rs4778889, allele C was identified to have a slight association compared with allele T in the female population.

Table 8.

Stratification Analysis of Three Studied Polymorphisms According to Gender Subgroups

Polymorphism	Genetic models	No. of studies	Statistic model	OR	95% CI		Z	p_Z	Het χ²	p_Het	I² (%)
Male
rs4778889	C vs. T	3	F	0.928	0.763	1.128	0.75	0.452	3.91	0.142	48.8
	TC+CC vs. TT	4	F	0.917	0.753	1.118	0.86	0.392	3.97	0.264	24.5
rs4072111	T vs. C	3	R	0.706	0.397	1.256	1.18	0.236	9.21	0.010	78.3
	CT+TT vs. CC	3	F	1.213	0.959	1.534	1.61	0.107	1.27	0.530	0.0
rs11556218	G vs. T	3	R	1.352	1.037	1.763	2.23	0.026	4.75	0.093	57.9
	TG+GG vs. TT	3	F	1.721	1.366	2.169	4.60	0.000	1.11	0.575	0.0
Female
rs4778889	C vs. T	3	F	1.332	1.044	1.699	2.31	0.021	4.11	0.128	51.4
	TC+CC vs. TT	4	R	0.935	0.608	1.438	0.30	0.761	8.97	0.030	66.6
rs4072111	T vs. C	3	R	0.608	0.311	1.187	1.46	0.145	9.50	0.009	78.9
	CT+TT vs. CC	3	R	0.608	0.311	1.187	1.46	0.145	9.50	0.009	78.9
rs11556218	G vs. T	3	R	1.140	0.748	1.736	0.61	0.543	8.58	0.014	76.7
	TG+GG vs. TT	3	F	1.725	1.291	2.305	3.69	0.000	1.34	0.511	0.0

In addition, among seven studies that were enrolled in this meta-analysis, six provided ORs adjusted by age and gender between cases and controls through linear regressions. To further eliminate any confounding factors that might have affected our conclusions, adjusted and unadjusted pooled ORs were calculated. Because the data contained in studies are limited, only three genetic models were compared. As shown in Table 9, the overall results of adjusted ORs are consistent with the unadjusted ones. A significant association was found for the rs11556218 polymorphism in the G versus T and TG versus TT models (p<0.05).

Table 9.

Meta-Analysis of the Association Between Three Studied Polymorphisms of Interleukin-16 and Cancer Risk Adjusted by Age and Gender

Polymorphism	Genetic models	Pooled	Statistic model	OR	95% CI		Z	p_Z	Het χ²	p_Het
rs4778889	C vs. T	Ad	R	0.929	0.737	1.171	−0.623	0.533	14.812	0.011
		Un	R	0.926	0.737	1.164	0.66	0.511	17.08	0.004
	TC vs. TT	Ad	R	0.976	0.753	1.265	−0.183	0.855	12.564	0.028
		Un	R	0.968	0.755	1.241	0.26	0.796	13.31	0.021
	CC vs. TT	Ad	R	0.879	0.452	1.712	−0.378	0.706	12.932	0.012
		Un	R	0.734	0.428	1.259	1.12	0.261	9.30	0.054
rs4072111	T vs. C	Ad	F	0.946	0.817	1.095	−0.744	0.457	5.298	0.258
		Un	F	0.956	0.836	1.093	0.65	0.513	7.09	0.131
	CT vs. CC	Ad	F	0.879	0.735	1.05	−1.423	0.155	3.221	0.522
		Un	F	0.919	0.779	1.084	1.00	0.316	3.71	0.447
	TT vs. CC	Ad	F	1.139	0.59	2.201	0.389	0.698	5.02	0.17
		Un	R	1.046	0.550	1.990	6.50	0.090	6.50	0.090
rs11556218	G vs. T	Ad	R	1.303	1.05	1.616	2.408	0.016	10.137	0.038
		Un	R	1.300	1.062	1.591	2.55	0.011	10.83	0.029
	TG vs. TT	Ad	F	1.773	1.457	2.158	5.721	0.000	7.249	0.123
		Un	F	1.646	1.400	1.936	6.02	0.000	7.76	0.101
	GG vs. TT	Ad	R	1.392	0.868	2.233	1.373	0.17	7.895	0.096
		Un	F	1.215	0.884	1.670	1.20	0.230	6.95	0.138

Ad, adjusted; Un, unadjusted.

Publication bias

For the comparatively small number of studies enrolled in this meta-analysis, it was difficult to verify the publication bias using the commonly used Begg's test and funnel plots, while Egger's regression was more sensitive in this aspect. Egger's linear regression in this meta-analysis exhibited no publication bias, and p-values of all polymorphisms were greater than 0.1, respectively (rs4778889, t=0.72, p=0.503 for TC vs. TT; rs4072111, t=1.16, p=0.312 for CT vs. CC; rs11556218, t=−0.53, p=0.626 for TG vs. TT).

Discussion

IL-16, a proinflammatory cytokine that has a wide range of biological functions, is also involved in tumor growth and progression (Yellapa et al., 2012). Certain single-nucleotide changes would result in the substitution of amino acid sequelae, thus affecting the expression of the IL-16 protein. Multiple loci of the IL-16 polymorphism have been discovered in relation to susceptibility to cancer. Among them, three polymorphisms located in the promoter or exon regions (nonsynonymous mutation: rs4072111 C/T, Ser/Pro substitution: rs11556218 T/G, Asn/Lys substitution in exon 6 area: rs4778889 T/C in the promoting area) (Nakayama et al., 2000) were the main targets of interest that were explored. The present meta-analysis was based on seven eligible studies quantitatively evaluating the three IL-16 polymorphisms with regard to the potential risk of cancer, including a pool of 1678 cases and 1937 controls, all of Asian origin. Our results from pooled data for the rs11556218 polymorphism indicated that the allele G had a 1.307-fold increased risk factor of systemic cancer when compared with those carrying the allele T. Significant results were also reached under a heterozygote model (OR=1.650; 95% CI, 1.424-1.911) and dominant model (OR=1.605; 95% CI, 1.391-1.845). This phenomenon implies that allele G may carry more pathogenic risks of the occurrence of systemic cancer.

Our results are consistent with previous studies addressing the rs11556218 polymorphism and cancer risk. Gao et al. (2009b), investigating a Chinese cohort stratified by gender with regard to the relationship between IL-16 gene polymorphisms and the risk of development of gastric and colorectal cancers, found that the rs11556218 T/G polymorphism of the IL-16 gene is significantly associated with cancer risk in both male and female populations in allelic contrast. Li et al. (2011), analyzing three polymorphisms of the IL-16 gene in 206 hepatocellular carcinoma cases and 264 healthy controls, found that both the TG genotype and G allele of the rs11556218 T/G polymorphism appear to have a higher risk of hepatocellular carcinoma than wild types. Recent publications also draw similar conclusions in nasopharyngeal carcinoma in a Chinese population and colorectal cancer in an Iranian population (Gao et al., 2009a; Azimzadeh et al., 2011).

Although we found the magnitude of association with cancer risk was relatively low, close attention should be paid to this polymorphism, for it affected a large portion (TG, 42.45%; G, 28.37%) of healthy controls in the pooled study populations. However, no association was observed under a dominant model (GG vs. TT). This conclusion was unexpected. It was assumed that the GG genotype would carry more pathogenic genes for carcinogenesis than would the TG genotype. For this part, we found low expression of the GG genotype (7.14%) in overall study populations, which may explain this discrepancy.

For the rs4778889 and rs4072111 polymorphisms, no associations were identified in the overall population by any of the genetic models. However, for rs4778889, although the association was low, allele C was identified to carry more cancer risk compared with allele T in the female subgroup.

Heterogeneity is observed across the studies of three polymorphisms in Table 4 through Table 6, and thus, we selected random-effect models to combine these pooled ORs. Studies conforming to HWE were also selected and analyzed to eliminate the bias of genetic distributions imposed on the synthesized results. The significance of the overall associations was not statistically altered after eliminating studies that deviated from HWE. To further eliminate selection bias among cases and controls, we compared a pool of adjusted and unadjusted ORs of studies in this meta-analysis. No substantial changes were discovered between the two ORs. This finding indicated that the anthropometric data (age and gender) of cases and controls were well matched in selected studies. Heterogeneity among studies was significantly changed by stratification of cancer types in this study, demonstrating that the type of cancer was a major source of heterogeneity.

Several limitations of our meta-analysis should be noticed when interpreting the results. (1) The number of studies and sample size were relatively small, which may result in reduced statistical power. (2) Genetic distributions may vary according to different ethnicities. The studies enrolled in this study were of Asian origin, mainly accumulating in China and Iran. Information with regard to other ethnicities and populations was insufficient. (3) Because of the lack of data in the original publications, we could not further stratify the results by other potential confounding factors to more fully explore the origin of heterogeneity. (4) The association of gene-gene and gene-environment interactions should be established and analyzed in future studies of the association between IL-16 polymorphisms and cancer risk.

In conclusion, despite these limitations, our meta-analysis provided a more accurate estimate of the association between IL-16 polymorphisms and cancer risk. Our results showed that the rs11556218 T/G polymorphism of the IL-16 gene was significantly associated with an elevated cancer risk in an Asian population. Our results warrant larger, better-designed studies in this field, including a greater variety of ethnicity and cancer type.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Azimzadeh

, Romani

, Mohebbi

, et al. (2011) Interleukin-16 (IL-16) gene polymorphisms in Iranian patients with colorectal cancer. J Gastrointestin Liver Dis, 20:371-376.

Azimzadeh

, Romani

, Mohebbi

, et al. (2012) Association of polymorphisms in microRNA-binding sites and colorectal cancer in an Iranian population. Cancer Genet, 205:501-507.

Bellomo

, Allegra

, Alonci

, et al. (2007) Serum levels of interleukin-16 in lymphoid malignancies. Leuk Lymphoma, 48:1225-1227.

Center

, Kornfeld

, Cruikshank

(1996) Interleukin 16 and its function as a CD4 ligand. Immunol Today, 17:476-481.

Chupp

, Wright

, Wu

, et al. (1998) Tissue and T cell distribution of precursor and mature IL-16. J Immunol, 161:3114-3119.

Cruikshank

, Kornfeld

, Center

(2000) Interleukin-16. J Leukoc Biol, 67:757-766.

DerSimonian

, Laird

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

Donald

, Donner

(1987) Adjustments to the Mantel-Haenszel chi-square statistic and odds ratio variance estimator when the data are clustered. Stat Med, 6:491-499.

Egger

, Davey

, Smith

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

10.

Egger

, Smith

, Phillips

(1997b) Meta-analysis: principles and procedures. BMJ, 315:1533-1537.

11.

Gao

, Liang

, Xue

, et al. (2009a) Genetic polymorphism of interleukin-16 and risk of nasopharyngeal carcinoma. Clin Chim Acta, 409:132-135.

12.

Gao

, Rao

, Wang

, et al. (2009b) The association of interleukin-16 polymorphisms with IL-16 serum levels and risk of colorectal and gastric cancer. Carcinogenesis, 30:295-299.

13.

Glass

, Sarisky

, Vecchio

(2006) Not-so-sweet sixteen: the role of IL-16 in infectious and immune-mediated inflammatory diseases. J Interferon Cytokine Res, 26:511-520.

14.

Goihl

, Rolle

, Kahne

, et al. (2012) Methodologic issues in the measurement of interleukin-16 in clinical blood samples using immunoassays. Cytokine, 58:1-5.

15.

Guven Maiorov

, Keskin

, Gursoy

, Nussinov

(2013) The structural network of inflammation and cancer: merits and challenges. Semin Cancer Biol, 23:243-251.

16.

Higgins

, Thompson

(2002) Quantifying heterogeneity in a meta-analysis. Stat Med, 21:1539-1558.

17.

Huang

, Lao

, Qin

, et al. (2012) Association between polymorphisms of interleukin-16 gene and susceptibility to nasopharyngeal carcinoma in Guangxi. Chongqing Medicine, 41:220-223,226. [In Chinese]

18.

Hussain

, Harris

(2007) Inflammation and cancer: an ancient link with novel potentials. Int J Cancer, 121:2373-2380.

19.

Kovacs

(2001) The serum levels of IL-12 and IL-16 in cancer patients. Relation to the tumour stage and previous therapy. Biomed Pharmacother, 55:111-116.

20.

, Deng

, Chen

, et al. (2011) Genetic polymorphism of interleukin-16 influences susceptibility to HBV-related hepatocellular carcinoma in a Chinese population. Infect Genet Evol, 11:2083-2088.

21.

Mathy

, Scheuer

, Lanzendorfer

, et al. (2000) Interleukin-16 stimulates the expression and production of pro-inflammatory cytokines by human monocytes. Immunology, 100:63-69.

22.

Nakayama

, Wasi

, Ajisawa

, et al. (2000) A new polymorphism in the promoter region of the human interleukin-16 (IL-16) gene. Genes Immun, 1:293-294.

23.

Siegel

, DeSantis

, Virgo

, et al. (2012) Cancer treatment and survivorship statistics, 2012. CA Cancer J Clin, 62:220-241.

24.

Yellapa

, Bahr

, Bitterman

, et al. (2012) Association of interleukin 16 with the development of ovarian tumor and tumor-associated neoangiogenesis in laying hen model of spontaneous ovarian cancer. Int J Gynecol Cancer, 22:199-207.

25.

Zhu

, Qin

, Yan

, et al. (2010) IL-16 polymorphism and risk of renal cell carcinoma: association in a Chinese population. Int J Urol, 17:700-707.