A Comprehensive Gene-Environment Interaction Analysis in Ovarian Cancer using Genome-wide Significant Common Variants

Abstract

As a follow-up to genome-wide association analysis of common variants associated with ovarian carcinoma (cancer), this study considers seven well-known ovarian cancer risk factors and their interactions with 28 genome-wide significant common genetic variants. The interaction analyses were based on data from 9,971 ovarian cancer cases and 15,566 controls from 17 case-control studies. Likelihood ratio and Wald tests for multiplicative interaction and for relative excess risk due to additive interaction were used. The top multiplicative interaction was noted between oral contraceptive pill (OCP) use (ever vs never) and rs13255292 (P-value = 3.48 × 10⁻⁴). Among women with the TT genotype for this variant, the odds ratio for OCP use was 0.53 (95% CI=0.46–0.60) compared to 0.71 (95%CI=0.66–0.77) for women with the CC genotype. When stratified by duration of OCP use, women with 1–5 years of OCP use exhibited differential protective benefit across genotypes. However, no interaction on either the multiplicative or additive scale was found to be statistically significant after multiple testing correction. The results suggest that OCP use may offer increased benefit for women who are carriers of the T allele in rs13255292. On the other hand, for women carrying the C allele in this variant, longer (5+ years) use of OCP may reduce the impact of carrying the risk allele of this SNP. Replication of this finding is needed. The study presents a comprehensive analytic framework for conducting gene-environment analysis in ovarian cancer.

INTRODUCTION

Ovarian carcinoma (cancer) is a disease with high mortality; most women are diagnosed with advanced stage disease where five-year survival is less than 50% ¹. Effective screening modalities have been elusive ², and therefore primary prevention strategies remain the most promising avenue to minimize the incidence and mortality of ovarian cancer.

Several factors consistently associated with reduced or increased risk have been identified for ovarian cancer, including some that represent opportunities for chemoprevention or surgical intervention. Factors associated with reduced risk include oral contraceptive pill (OCP) ³ use aspirin use ⁴, tubal ligation ⁵, parity ³, salpingectomy ^6–9 and bilateral salpingo-oophorectomy (BSO). Common germline genetic variation ^10–20, first-degree family history of ovarian cancer ^{21, 22}, menopausal hormone therapy use ^23–25, greater body mass index (BMI) ²⁶ and endometriosis ²⁷ are risk factors for the disease. OCPs and aspirin use represent feasible chemoprevention strategies whereas salpingectomy is now recommended by many gynecologic societies as an ovarian cancer prevention approach for women seeking tubal sterilization, having a hysterectomy, or having other pelvic surgery.

Average lifetime risk of ovarian cancer diagnosis for women in the U.S. is 1.3% ²⁸, but this number varies greatly depending on the composite exposure history of risk factors ²⁹. Pearce et al. estimated the lifetime risk for women in the general population ranges from 0.35% (95%CI = 0.29% to 0.42%) to 8.8% (95%CI = 7.1% to 10.9%) depending on exposure history for six factors: OCP use, parity, tubal ligation, endometriosis, first degree family history of ovarian cancer and genetic risk score quintile ²⁹.

However, these lifetime risk estimates were limited to six risk factors and did not consider their interaction with individual genetic variants identified through genome-wide association studies (GWAS) ²⁸. The multiplicative scale is commonly used for gene-environment interaction (G × E) analysis. Additive interaction analysis has been suggested for case-control studies in many recent papers for a more mechanistic interpretation ^30–34. Validity of a truly multiplicative model implies existence of additive interaction when the two factors under consideration have non-null main effects ³⁵. Thus, failure to detect G × E interaction on multiplicative scale may imply there exists interaction on additive scale, but the ability to detect it depends on the sample size and the main and interaction effect sizes ³⁵. We present here our efforts to evaluate both multiplicative and additive gene-environment interactions in ovarian cancer using data from the international Ovarian Cancer Association Consortium (OCAC) comprising 17 case-control studies.

We have included 28 common genetic variants previously associated with risk of ovarian cancer in genome-wide association analyses for our G × E analyses ³⁶. Environmental factors included in our analysis are OCP use, parity, tubal ligation, breastfeeding, menopausal hormone therapy, usual adult BMI, and endometriosis. A small number of studies in OCAC had data available on aspirin use and thus we have not included this risk factor in our analysis here. Among our list of environmental factors, BMI, OCP use, tubal ligation, breastfeeding, and menopausal hormone therapy are of special interest because they are modifiable targets for prevention.

METHODS

Study Population

The OCAC is an international multidisciplinary consortium formed in 2005 (http://apps.ccge.medschl.cam.ac.uk/consortia/ocac/) with a goal of sharing data from worldwide ovarian cancer studies to establish reliable estimation of association between environmental and genetic factors related to risk of ovarian cancer ^{23, 37}. Cases were defined as women with ovarian carcinoma (i.e., invasive epithelial ovarian cancers), fallopian tube cancer and primary peritoneal cancer. Controls were women without ovarian cancer and who had at least one ovary. For both cases and controls, individuals with prior cancers except non-melanoma skin cancers were excluded.

Genetic Association Analysis

In total, 28 single nucleotide polymorphisms (SNPs) previously identified through GWAS were included from 75 OCAC sites (Table 1). The first 26 SNPs were found to be significantly associated with either ovarian cancer overall or one or more histotypes ³⁶. In addition, rs13255292 and rs10962643 were included because they were in the same region as two other significant SNPs but showed a strong independent association with ovarian cancer risk. The SNP at locus 15q26 (rs8037137), which was found to be genome-wide significant ¹³, was not included because not enough non-carriers were present in our analytic dataset for examining interactions. The genetic data included both genotyped and imputed variants (imputation being carried out using phase 2 Hapmap reference panel). More details regarding genotyping and imputation of the genetic data have been previously described ^{12, 17, 18, 20}. The methods for analyzing the SNP data in the OCAC have also been described previously ^{12, 17, 18, 20}. Briefly, logistic regression models were fit to examine the association between ovarian cancer and each genetic variant under an additive model (using risk allele dosage). The models were adjusted for ethnicity, genotyping panel and the leading principal components for each ethnicity. The summary results are shown in Table 1 and are also available through the OCAC website (http:/apps.ccge.medschl.cam.ac.uk/consortia/ocac/).

Table 1.

Odds ratios for marginal associations of 28 genetic susceptibility variants with ovarian cancer. Analysis used data with 26864 cases and 48034 controls from 75 study sites.

SNP	Previously published best hita	Chr	Position	Risk Allele	Baseline Allele	RAF	ORb	P-valueb
rs12023270	rs5872217015	1	38086578	T	C	0.264	1.08 (1.05,1.10)	2.65×10−8
chr2:111818658	rs216510918	2	111818658	C	A	0.277	1.06 (1.04,1.09)	2.03×10−6
rs874898	rs75259014	2	113974196	C	G	0.262	1.00 (0.98,1.03)	7.36×10−1*
rs1562314	rs71183014	2	177045560	T	A	0.638	1.10 (1.07,1.13)	2.84×10−14
rs11207182018		3	138849110	allele 1	G	0.270	1.03 (1.00,1.06)	5.17×10−2*
chr3:156397692	rs6227404117	3	156397692	T	C	0.048	1.47 (1.39,1.55)	7.73×10−47*
rs987020718		3	190525516	A	G	0.666	1.05 (1.03,1.08)	2.95×10−5
rs7705526	rs1006969010	5	1285974	A	C	0.343	1.10 (1.07,1.12)	5.52×10−14
chr5:66121089	rs55502517918	5	66121089	allele2	G	0.526	1.03 (1.00,1.05)	2.61×10−2*
chr8:82653644	8:8266881817	8	82653644	G	A	0.064	1.18 (1.12,1.23)	3.25×10−12*
rs988665118		8	128817883	G	A	0.435	1.06 (1.03,1.08)	2.89×10−6*
rs1325529218	NA	8	129076573	C	T	0.700	1.07 (1.05,1.10)	3.57×10−8*
rs10103314	rs1400482 12	8	129560744	A	C	0.883	1.15 (1.11,1.20)	5.76×10−15*
chr9:16915105	rs1096269220	9	16915105	C	G	0.834	1.24 (1.20,1.28)	4.54×10−41*
rs10962643	NA	9	16857403	C	A	0.699	1.17 (1.14,1.20)	1.13×10−35*
rs320203 18		9	104943226	C	A	0.842	1.03 (1.00,1.06)	5.21×10−2
chr9:136138765 15		9	136138765	G	allele 3	0.176	1.12 (1.08,1.15)	1.49×10−12*
rs7084454	rs14496237617	10	21821274	A	G	0.301	1.07 (1.05,1.10)	3.32×10−8*
rs7902587 18		10	105694301	T	C	0.091	1.08 (1.03,1.12)	4.54×10−4*
chr12:121403724	rs7953249 18	12	121403724	A	G	0.570	1.05 (1.03,1.07)	2.58×10−5*
chr15:91531995	rs8037137 13	15	91531995	C	T	0.829	1.08 (1.05,1.12)	1.18×10−6*
rs11658063	rs7405776 19	17	36103872	G	C	0.614	1.04 (1.02,1.07)	2.98×10−4*
chr17:43552537	rs1879586 17	17	43552537	A	G	0.164	1.12 (1.08,1.15)	2.22×10−12*
rs7217120	rs720782616	17	46484755	C	T	0.275	1.10 (1.07,1.13)	8.69×10−13*
rs809824418		18	21405553	G	A	0.741	1.04 (1.01,1.07)	4.23×10−3*
rs4808075 11		19	17390291	C	T	0.268	1.13 (1.10,1.16)	1.49×10−20*
rs74597329	rs68818714	19	39739155	G	T	0.301	1.02 (0.99,1.04)	2.63×10−1
rs6005807 18		22	28934313	T	C	0.095	1.09 (1.04,1.13)	3.35×10−5*

Abbreviations: SNP, single-nucleotide polymorphism; RAF, risk allele frequency; Chr, chromosome; OR, odds ratio; allele1, GCCAGATTCAGAAT; allele2, GACACACAC; allele3, GCGCCCACCACTA.

^aIf not specified, the previously published best hit is the same as the current best hit.

^bLogistic regression for ovarian cancer overall (regardless of histology), adjusted for ethnicity, study panel and leading principal components for each ethnicity (using a total of 47 principal components).

^*P-value > 0.01

Environmental Association Analysis

Environmental Variables (E):

A total of seven established environmental risk factors for ovarian cancer were of primary interest (Table 2), including four associated with decreased risk and three with increased risk for ovarian cancer or one specific histotype. These included: OCP use (measured as both ever/never and duration of OCP use (never users including <1 one year of use, 1-<5, 5+yr), tubal ligation (yes/no), breastfeeding (ever/never), parity (0, 1–2, 3+ full-term births (i.e., those lasting ≥6 months), type of menopausal hormone therapy use for more than 1 year after age 50 (never user, menopausal estrogen therapy only, any use of menopausal estrogen + progestin therapy), BMI (<25, 25-<30, 30+), and a history of endometriosis (yes/no).

Table 2.

Odds ratios for marginal associations of seven environmental risk factors with ovarian cancer risk with 13722 cases and 22975 controls from 19 study sites.

Environmental risk factor	Before Imputationa		After Imputationb
	Control	Case	Control	Case	ORc	P-valuec
OCP use
Never	0.347	0.444	0.351	0.452	Ref
Ever	0.645	0.536	0.649	0.548	0.62 (0.59,0.66)	5.24×10−73
(missing)	0.008	0.020
Duration of OCP use
Never users (including <1 year)	0.425	0.542	0.430	0.554	Ref
1−<5 year	0.229	0.208	0.232	0.215	0.70 (0.66,0.74)	8.23×10−32
5+ year	0.332	0.222	0.338	0.231	0.48 (0.45,0.51)	2.20×10−133
(missing)	0.014	0.028
Tubal ligation
No	0.693	0.777	0.762	0.824	Ref
Yes	0.208	0.160	0.238	0.176	0.73 (0.69,0.78)	1.81×10−23
(missing)	0.098	0.063
Breastfeeding
No	0.239	0.294	0.380	0.515	Ref
Yes	0.532	0.410	0.620	0.485	0.76 (0.71,0.80)	4.80×10−21
(missing)	0.229	0.296
Parity (number of full−term births)
0	0.148	0.241	0.149	0.243	Ref
1–2	0.487	0.434	0.489	0.438	0.59 (0.55,0.63)	1.94×10−65
3+	0.359	0.315	0.362	0.319	0.50 (0.46,0.53)	4.91×10−90
(missing)	0.006	0.011
Type of HT using more than 1 year after age 50
Never use	0.687	0.647	0.789	0.782	Ref
ET only	0.060	0.075	0.066	0.084	1.22 (1.12,1.34)	2.65×10−5
Any EPT	0.131	0.118	0.145	0.134	0.97 (0.90,1.04)	3.55×10−1
(missing)	0.121	0.160
BMI
< 25	0.392	0.370	0.516	0.485	Ref
25−<30	0.209	0.213	0.284	0.286	1.03 (0.98,1.09)	2.55×10−1
30+	0.144	0.174	0.200	0.229	1.15 (1.08,1.22)	6.11×10−6
(missing)	0.255	0.243
Endometriosis
No	0.703	0.695	0.937	0.902	Ref
Yes	0.047	0.076	0.063	0.098	1.60 (1.46,1.75)	3.41×10−23
(missing)	0.250	0.230

Abbreviations: OR, odds ratio; OCP, oral contraceptive pills; BMI, body mass index; HT, menopausal hormone therapy; ET, menopausal estrogen therapy; EPT, menopausal estrogen + progestin therapy; Ref, reference group.

^aHarmonized environmental data before imputation. Results of the complete cases analysis are provided in eTable 2.

^bBased on ten imputed E datasets.

^cLogistic regression model adjusted for reference age, race, education, family history, OCP use, tubal ligation, parity, BMI, endometriosis and study site.

Four other environmental variables were included in our analysis, as covariates: baseline age (<50, 50-<55, 55-<60, 60-<65, 65–70, 70+ years), race (non-Hispanic white, Hispanic White, Black, Other), education (less than high school, high school graduate, some college, college graduate) and first-degree family history of ovarian cancer (yes/no). In addition to these four covariates, study site, OCP use, tubal ligation, parity, BMI and endometriosis were also included in all models for the environmental association analysis and gene by environment interaction analysis.

Harmonization and Imputation of Environmental Data:

A brief description of environmental data harmonization across OCAC study sites is provided in eMethod 1 in the Supplementary Material. To optimize power and enhance the chance for discovery, we carried out multiple imputation of the environmental data. The maximal amount of data was used for imputation (see eMethod 1 and eFigure 1 in the Supplementary Material for details). A total of 19 studies comprising 13,722 cases and 22,975 controls with partially missing data were included for imputation. Of these 19 studies, 12 were from the US, 4 from Europe, 2 from Canada and 1 from Australia (see eTable 1 for a description of study sites). Further details for these 19 studies have been previously described (see Supplementary Material). The environmental variables included in our analysis were multiply imputed by chained equations (MICE) to produce ten imputed datasets. See details of imputation model in eMethod 2.1 in the Supplementary Material.

All analyses were performed on each of the ten imputed datasets, and coefficients/test statistics were properly combined to account for uncertainty due to imputation, following the recommended combination rule for multiply imputed datasets ³⁸ (see details in eMethod 2.3 in the Supplementary Material). Our marginal environmental association analysis was based on combined inference from the ten imputed versions of this harmonized E data. Logistic regression models were used for evaluating marginal associations between the environmental risk factors with ovarian cancer after adjusting for covariate. The estimated ORs, their 95% CIs, as well as two-sided Wald tests after accounting for imputation uncertainty are presented in Table 2 along with summary statistics of complete cases before imputation. Full results of the complete cases analysis using logistic regression models are presented in eTable 2.

Gene by Environment Interaction Analysis

After marginal analysis of the genetic and environmental risk factors, we considered gene by environment (G × E) interaction analysis both on the multiplicative (odds ratio/relative risk) and the additive (relative excess risk due to interaction/absolute risk) scale ³⁹. From the 19 studies with imputed environmental data, a subset of 17 case-control studies with 9,971 cases and 15,566 controls had available genetic data, thus G × E analyses were carried out on these 17 studies. Each imputed environmental dataset was merged with the genetic data for subsequent G × E analyses. Interaction analyses were then carried out separately on the ten imputed G × E datasets, and then all tests and coefficients reported were combined using appropriate multiple imputation combination rules ³⁸.

For both multiplicative and additive interaction analysis, we started with global likelihood ratio tests (LRTs) for each G × E pair as several environmental factors had multiple categories resulting in tests for interactions with multiple degrees of freedom (df). These global joint tests, serving as a screening step for G × E interactions, were carried out for a total of 196 (7×28=196) G × E pairs. After the global tests, we then followed up on the suggestive interactions (with global test P-value < 0.2) and carried out a two-sided Wald test for interactions involving each separate category of an environmental risk factor.

For the k-th SNP $G_k\left(\mathrm{k}=1,\dots ,28\right)$, coded as a continuous allelic dosage, the j-th environmental risk factor $E_j\left(j=1,\dots ,7\right)$, and a set of confounders/covariates $\left\{C_q\right\}(q=1,\dots ,Q)$, the basic fitted model for the probability of ovarian cancer of the i-th subject, namely, ${\pi }_i$, is of the following form:

$$\begin{array}{l}logit\left({\pi }_i|G_{ki},E_{ji},C_{1i},\dots ,C_{Qi}\right)\\ ={\beta }_0+{\beta }_G{G}_{ki}+{\sum }_{l=1}^L{\beta }_{El}I\left(E_{ji}=l\right)+{\sum }_{l=1}^L{\beta }_{GEl}I\left(E_{ji}=l\right)G_{ki}+{\sum }_{q=1}^Q{\sum }_{m=1}^{M_q}{\beta }_{C_{q}m}I\left(C_{qi}=m\right),\end{array}$$ [M1]

where L = (levels of $E_j$) – 1, $M_q$ = (levels of $C_q$) – 1, and Q is the number of adjusted covariates.

Multiplicative Interaction Tests:

For testing the multiplicative interaction between $G_k$ and $E_j$, we first used the global LRT with L degrees of freedom to test for the joint null hypothesis $H_0:{\beta }_{GE1}={\beta }_{GE2}=\dots ={\beta }_{GEL}=0$. If the global test P-value < 0.2, we further assessed the multiplicative interaction at each level of $E_j$ by using a Wald test with one degree of freedom for the null hypothesis $H_0:{\beta }_{GEl}=0$ for the l-th level.

Additive Interaction Tests:

Due to limitations of existing software (CGEN) ⁴⁰ for testing additive interactions with continuous dosage data, we used the maximal probable genotype for imputed SNPs. We further conducted the LRTs with binary collapsing of SNPs assuming a dominant genetic susceptibility model (given the constraints in software) ³¹. For a given SNP $G_k$ and an environmental risk factor $E_j$ with L categories, a global LRT with L df was used for the following joint null hypothesis

$$H_0:\frac{\left\{\exp \left({\beta }_{E1}\right)+\exp \left({\beta }_G\right)-1\right\}}{\text{exp}\left({\beta }_{E1}+{\beta }_G\right)}=\exp \left({\beta }_{GE1}\right),\dots ,\frac{\left\{\exp \left({\beta }_{EL}\right)+\exp \left({\beta }_G\right)-1\right\}}{\text{exp}\left({\beta }_{EL}+{\beta }_G\right)}=\exp \left({\beta }_{GEL}\right),$$

where the regression coefficients $\left(\beta \right)$ are log odds ratio parameters described in model [M1]. This null hypothesis is based on a rare disease assumption ⁴¹, which is tenable for our study (lifetime risk of ovarian cancer in the US is approximately 1.3%) ⁴². If the global LRT P-value < 0.2, we further assessed the additive interaction at each level of $E_j$ through the relative excess risk due to interaction (RERI) ⁴¹. At the l-h level of $E_j$_, a Wald test with one degree of freedom (35) was used to test for the null hypothesis:

$$H_0:RER{I}_{GEl}=0,\text{ where }RER{I}_{GEl}=\exp \left({\beta }_{El}+{\beta }_{GEl}+{\beta }_G\right)-\exp \left({\beta }_{El}\right)-\exp \left({\beta }_G\right)+1.$$

After the screening step, we further explored the structure of the most promising interactions (defined as global test P-value < 0.01). This was accomplished by exploring odds ratios corresponding to E in sub-groups defined by G (for the multiplicative interaction) or absolute risks for ovarian cancer in each configuration of the values of (G, E) (for the additive interaction). To better understand these two different scales of interaction, we also compared the observed joint ORs with the corresponding expected ORs under the multiplicative and the additive nulls.

To estimate sub-group specific absolute risk (AR) for each stratum defined by a given SNP $G_k$ and environmental risk factor, we need the relative risk and the joint distribution of $G_k$ and $E_j$. The former was estimated from the fitted model [M1], and the latter was empirically estimated from the observed joint frequency of $E_j$ and $G_k$ in the control population (details in eMethod3 from the Supplementary Material). Table 4 presents the bootstrap confidence intervals for the estimated ARs and the risk differences (RDs) (see details in eMethod4 in the Supplementary Material). The results for G × E analysis are presented in Table 3 (multiplicative interaction), Table 4 (additive interaction) and eTable 5 (observed and expected joint OR under the two different nulls). All calculations were performed in the statistical software R ^{30, 40}.

Table 4.

Absolute risks and risk differences stratified by levels of environmental risk factor and levels of genotype (for G-E pairs with global likelihood ratio test p-value < 0.01 on additive scale. Analysis used the G×E data with 9971 cases and 15566 controls from 17 study sites).

SNPs	Environmental risk factor		N (cases/controls)a			Estimated ARs or RDs for E stratified by SNPs (95%CI)c			Global LRTd	Wald Teste
risk/baseline allele	variable	category	Genotype			Genotype			(df)	(df)
rs11658063 G/C	Type of HT		CC	CG	GG	CC	CG	GG
		Neither	589/1142	2609/4518	3310/4956	1.27% (1.23%,1.32%)	1.30% (1.28%,1.33%)	1.33% (1.26%,1.40%)	Ref	Ref
		ET only	66/98	281/409	416/454	1.36% (1.15%,1.57%)	1.63% (1.46%,1.79%)	1.96% (1.59%,2.33%)
		RDb				0.09% (−0.14%,0.31%)	0.33% (0.15%,0.50%)	0.63% (0.24%,1.02%)	3.29×10−3(2)	3.01×10−2(1)
		Any EPT	105/207	498/952	606/1046	1.16% (1.04%,1.28%)	1.21% (1.12%,1.30%)	1.27% (1.09%,1.44%)
		RD				−0.12%(−0.26%,0.03%)	−0.09%(−0.20%,0.01%)	−0.06%(−0.26%,0.13%)		7.04×10−1(1)
		missing	122/202	582/762	787/820
rs9886651G/A	OCP use		AA	AG	GG	AA	AG	GG
		Never	1278/1718	2053/2502	900/1028	1.52% (1.42%,1.62%)	1.70% (1.64%,1.76%)	1.91% (1.77%,2.04%)	Ref	Ref
		Ever	1666/3105	2640/4978	1194/2072	1.07% (1.02%,1.12%)	1.10% (1.07%,1.13%)	1.14% (1.07%,1.21%)
		RD				−0.45%(−0.57%, −0.33%)	−0.60%(−0.69%, −0.51%)	−0.77%(−0.93%, −0.60%)	5.32×10−3(2)	9.90×10−3(1)
		missing	70/47	113/79	57/37

Abbreviation: SNP, single-nucleotide polymorphism; AR, absolute risk; RD, risk difference; OCP, oral contraceptive pills; HT, menopausal hormone therapy; ET, menopausal estrogen therapy; EPT, menopausal estrogen + progestin therapy; Ref, reference group; df, degree of freedom.

^a:Number of cases and controls were estimated from the original merged G×E data (before imputation) with 9971 cases and 15566 controls from 17 study sites, using maximal probable genotypes for imputed SNPs.

^b:The risk difference corresponds to given category compared to the reference group, stratified by SNP.

^c:ARs were estimated from logistic regression model by empirically estimated distribution of E and SNPs, while fixing all other covariates at their mode (determined from the original data).

^d:LRT was performed for jointly testing additive interactions, assuming dominant effect model of SNPs (due to limitation of software).

^e:1-df Wald test corresponds to the test individual RERI term (SNP = 2 vs SNP = 0, E = k vs E = reference group) is zero or not.

All models were estimated from logistic regression model with SNP, E variable, SNP E variable, assuming log-additive model (except for additive LRT which assumes dominant effect), using maximal probable genotypes for imputed SNPs, adjusted for reference age, race, education, family history, OCP use, tubal ligation, parity, BMI, endometriosis and study site and were performed on imputed datasets of G-E (9971 cases, 15566 controls) with proper pooling.

Table 3.

Odds ratios for marginal associations of seven environmental risk factors with ovarian cancer risk with 13722 cases and 22975 controls from 19 study sites.

SNP	Environmental risk factor		N (cases/controls)a			Estimated ORb for E stratified by G (95%CI)			Globalc LRT	Waldd Test
Risk/Baseline allele	Variable	Category	Genotype			Genotype			(df)	(df)
rs13255292 C/T	OCP use		TT	TC	CC	TT	TC	CC
		Never	396/503	1758/2175	2077/2570	Ref			Ref	Ref
		Ever	446/1069	2286/4336	2768/4750	0.5 (0.46,0.60)	0.61 (0.57,0.66)	0.71 (0.66,0.77)	3.48×10−4 (1)	3.47×10−4 (1)
		Missing	24/15	96/56	120/96
rs13255292 C/T	Duration of OCP use		TT	TC	CC	TT	TC	CC
		< 1 yr	451/636	2213/2670	2546/3145	Ref			Ref	Ref
		1−<5 yr	171/362	854/1522	1082/1662	0.58 (0.50,0.69)	0.68 (0.63,0.74)	0.79 (0.72,0.87)	7.26×10−3 (2)	4.74×10−3 (1)
		5+ yr	209/568	945/2269	1178/2470	0.43 (0.37,0.5)	0.48 (0.44,0.52)	0.53 (0.49,0.58)		2.43×10−2 (1)
		Missing	35/21	128/106	159/135
rs10962643 C/A	Parity (full term birth)		AA	AC	CC	AA	AC	CC
		0	230/220	940/940	1194/1080	Ref			Ref	Ref
		1–2	398/835	1741/3184	2202/3536	0.52 (0.44,0.61)	0.56 (0.51,0.6)	0.60 (0.54,0.66)	7.52×10−3 (2)	1.99×10−1 (1)
		3+	243/579	1242/2459	1664/2614	0.38 (0.32,0.46)	0.46 (0.42,0.5)	0.55 (0.49,0.61)		2.86×10−3 (1)
		Missing	11/15	47/58	59/46
chr9:16915105 C/G	Parity (full term birth)		GG	GC	CC	GG	GC	CC
		0	73/72	624/649	1667/1519	Ref			Ref	Ref
		1–2	111/300	1129/2285	3101/4970	0.46 (0.36,0.58)	0.52 (0.47,0.59)	0.60 (0.55,0.65)	5.25×10−3 (2)	5.10×10−2 (1)
		3+	70/220	749/1679	2330/3753	0.33 (0.26,0.43)	0.42 (0.37,0.48)	0.53 (0.48,0.58)		1.25×10−3 (1)
		missing	2/7	37/36	78/76

Abbreviation: SNP, single-nucleotide polymorphism; OR, odds ratio; OCP, oral contraceptive pills; yr, year; Ref, reference group; df, degree of freedom, LRT, likelihood ratio test.

^aNumber of cases and controls were estimated from the original merged GxE data (before imputation) with 9971 cases and 15566 controls from 17 study sites, using maximal probable genotypes for imputed SNPs.

^bORs were estimated from the logistic regression model with SNP, E variable, SNP E variable.

^cLRT was performed for jointly testing multiplicative interactions.

^dWald test for individual multiplicative interaction.

All models were estimated from the logistic regression model with SNP, E variable, SNP E variable, assuming log-additive model, using dosage data for imputed SNPs, adjusted for reference age, race, education, family history, OCP use, tubal ligation, parity, BMI, endometriosis and study site and were performed on imputed datasets of G-E (9971 cases, 15566 controls) with proper pooling.

RESULTS

The marginal G analysis was carried out on 26,864 cases and 48,034 controls and the results are shown in Table 1. These results are available through the OCAC website (http://apps.ccge.medschl.cam.ac.uk/consortia/ocac/). A total of 36,697 women with 13,722 ovarian cancer cases from 19 sites were included in the marginal E analysis using the imputed datasets. All seven environmental risk factors were associated with ovarian cancer in the expected direction (Table 2). OCP use for five or more years was associated with a 52% decrease in risk of ovarian cancer compared to never users (OR=0.48, 95%CI = 0.45 to 0.51). Tubal ligation (OR=0.73, 95%CI = 0.69 to 0.78) and breastfeeding (OR=0.76, 95%CI = 0.71 to 0.80) showed similar magnitudes of decreased risk. Also, having more than 3 children (versus none) was associated with a 50% (OR=0.5, 95%CI = 0.46 to 0.53) reduction in risk of ovarian cancer. Using menopausal estrogen therapy only for more than one year (OR=1.22, 95%CI = 1.12 to 1.34), being obese (OR=1.15, 95%CI = 1.08 to 1.22), and history of endometriosis (OR=1.60, 95%CI = 1.46 to 1.75) were all associated with increased risk of ovarian cancer. The inference remained robust before and after imputation (eTable 2.).

Gene by Environment Interaction Results

Global Likelihood Ratio Tests:

The global LRT essentially serves as a screening approach to identify a list of potentially interesting interactions. All interactions with global LRT P-value < 0.2 (40 on multiplicative scale and 41 on additive scale) are listed in eTable 3, while more detailed analysis of the top interactions, which showed the strongest significance (P-value < 0.01; 4 on multiplicative and 2 on additive scale), are shown in Table 3 and Table 4, respectively.

According to Global LRT results, the top interaction on the multiplicative scale was identified with the SNP rs13255292 and OCP use (ever and never use: P-value = 3.48 × 10⁻⁴; duration of use [<1 yr, 1–5 yr, 5+ yr]: P-value = 7.26 × 10⁻³) (Table 3). None of the observed interactions were significant based on a Bonferroni threshold of 0.05/(28 × 7)= 2.55 × 10⁻⁴.

Wald Tests for Multiplicative interactions:

For the most promising multiplicative interactions reported in Table 3 we carried out an in-depth analysis to better understand the structure of interactions by estimating the ORs (with accompanying Wald CIs and tests) corresponding to E in strata defined by G. For example, the OR for OCP use among women with the TT genotype for rs13255292 is estimated to be 0.53 (95%CI = 0.46 to 0.60), whereas for the CC genotype the estimated OR is 0.71 (95%CI = 0.66 to 0.77) suggesting a stronger protective effect of OCP use among TT genotypes (Table 3, Figure 1A).

Figure 1A–1C.

ORs of oral contraceptive (OCP) use, marginally, or stratified by number of risk allele of rs13255292. The ORs were calculated from a logistic regression model assuming log-additive effect of SNPs. (A) OR of OCP (ever vs never) (B) OR of 1 to 5 years of OCP use (vs < 1 year) (B) OR of more than 5 years of OCP use (vs < 1 year).

When OCP use was further stratified by duration, we observed an interesting pattern in its interaction with rs13255292. The estimated OR corresponding to 1–5 year of OCP use vs < 1 year use in the TT genotype group was 0.58 (95%CI = 0.50 to 0.69) compared to an OR of 0.79 (95%CI = 0.72 to 0.87) among women with CC genotype, showing effect modification by the risk allele (C) of rs13255292 (Table 3, Figure 1B). This is akin to the result with ever/never user. However, the OR corresponding to 5+ years of OCP use vs < 1 year of use for the TT genotype group was 0.43 (95%CI = 0.37 to 0.50) and for the CC genotype was 0.53 (95%CI = 0.49 to 0.58) (Table 3, Figure 1C). With overlapping confidence intervals, there is no significant difference in the odds ratios for long-term OCP users across genotype sub-groups. Table 3 shows that the P-value of the Wald test for interaction of rs13255292 and 1–5 years of OCP use (vs < 1 yr) was lower (P-value = 4.74 × 10⁻³), when compared to the P-value for interaction of the same variant with 5+ years of OCP use (vs < 1 yr) (P-value = 2.43 × 10⁻²).

Wald Test for Additive interaction/RERI:

For the most statistically significant additive interactions in Table 4, we estimated the sub-group specific absolute risks (ARs) and risk differences (RDs) in each E by G stratum. For example, for the strongest additive interaction based on the global likelihood ratio tests in Table 4, there was suggestive evidence that rs11658063 modified the effect of menopausal estrogen therapy use, compared to never use of menopausal hormone therapy (P-value = 3.01 × 10⁻²). Among women with the GG genotype, never users of menopausal hormone therapy had an estimated AR of 1.33% (95%CI =1.26% to 1.40%) while women who used menopausal estrogen therapy had an estimated AR of 1.96% (95%CI = 1.59% to 2.33%), leading to an absolute risk increase of 0.63% (95%CI = 0.24% to 1.02%) (Table 4, eFigure 2).

For women with the CC genotype, the estimated AR was 1.27% (95%CI = 1.23% to 1.32%) for never receiving menopausal hormone therapy and 1.36% (95%CI = 1.15% to 1.57%) for receiving menopausal estrogen only therapy. This implies virtually no increased risk from taking menopausal estrogen only therapy among women with the CC genotype (95%CI = −0.14% to 0.31%; Table 4, eFigure 2). The results on the additive interactions were in general weaker in terms of the strength of P-values.

DISCUSSION

We have conducted a comprehensive multiplicative and additive interaction analysis of previously identified common genetic variants and environmental factors unequivocally associated with ovarian cancer risk. We observed six suggestive interactions (with P-value < 0.01), four on the multiplicative scale and two on the additive scale. The lack of statistical significance of interactions after multiple testing correction from a large collection of data and well-curated studies enable us to conclude that it is unlikely that there are substantive interactions with single variants and environmental factors regardless of the choice of scale. This is consistent with what has been observed for other cancers. One may argue that the Bonferroni threshold for multiple comparisons is likely to be conservative for this set of correlated environmental factors, but the general pattern of findings remains consistent with smaller magnitude of interaction effect sizes. However, there are several interesting findings from this analysis that may be worthwhile to follow-up in future G × E studies of ovarian cancer.

Mechanistic Insight:

In addition to guiding targeted prevention strategies, G × E analysis has the potential to provide mechanistic insight into the complex multifactorial structure of the underlying biological pathway. One issue complicating observed gene-environment interactions of even confirmed susceptibility loci is that the true casual alleles and the biological impact of the variants are unknown. Our top interaction is between OCP use and rs13255292. This variant lies in the 8q24 region which harbors several risk loci for ovarian cancer ¹⁸ and other cancers ^{43, 44}. The SNP is in the PVT1 gene which interacts with the oncogene MYC ⁴⁵. MYC has long been reported to be at least in part under hormonal control ^{46, 47} thus an interaction with OCP use is plausible. Conversely, our top additive interaction is between menopausal estrogen use and rs11658063 which falls in HNF1B. To our knowledge there is no relationship between HNF1B and hormones thus underscoring the difficulty of understanding these gene-environment interactions given our limited understanding of the function of the variants and even more broadly the biological role of the genes.

Exposure Pathways and Potential for Targeted Prevention:

The strongest interactions are observed with OCP use or menopausal estrogen use which are modifiable exposures. Our most promising finding is the potential interaction between SNP rs13255292 and OCP use. This finding, if replicated could potentially lead to improved understanding of exposure pathways.

Analytic Architecture and the Choice of Scale for Measuring Interaction:

We present a comprehensive analytical framework to carry out post-GWAS G × E analysis on both multiplicative and additive scale. Our framework starting with data harmonization and imputation followed by Global likelihood ratio tests and single df Wald tests provides a principled analytic architecture for such analysis. Our analysis reiterates the well-known fact that testing the additive and multiplicative nulls are very similar when the marginal associations are weak but could depart when both marginal associations are large in magnitude and the sample size is finite. In eTable 5, we present observed joint odds ratios for strata defined by G and E along with the expected odds ratios under the multiplicative null and the additive null. We use our top hit rs13255292 and OCP use (ever versus never) and length of OCP use (<1yr, 1-<5 yrs, 5+ yrs) as an illustration. One can note that the expected ORs are fairly close under both models. However, their estimated departure from the observed joint OR is more pronounced for the 1-<5 yrs sub-group when compared to 5+ yrs, explaining the suggestive evidence for rejecting the null.

We discussed the multiplicative interaction results for rs13255292 and OCP use in the previous section. We now explore the structure of additive interaction for this G × E result (Figure 2A-2C). Marginally, without including any genetic information, from a pure environmental association analysis we observed a relationship between duration of OCP use and risk reduction for ovarian cancer. For 1–5 years of OCP use (vs <1 year) the estimated absolute risk difference was 0.47% (95%CI = 0.37% to 0.56%), while the estimated absolute risk difference for long-term use of OCPs (5+ year vs <1 year) was 0.84% (95%CI = 0.77% to 0.92%) (Figure 2B-2C, eTable 4), in agreement with previous findings that longer duration of OCP use is associated with larger risk reduction in ovarian cancer ³. However, when stratified by rs13255292 genotype, we observed an interesting pattern. Among individuals with TT genotype, the corresponding absolute risk difference estimate for 1–5 year of OCP use (vs <1 year) was 0.69% (95%CI = 0.49% to 0.88%), whereas among individuals with CC genotypes the corresponding risk reduction estimate was 0.36% (95%CI = 0.22% to 0.50%), implying potential effect modification by the C allele at locus rs13255292 (P-value = 1.12 × 10⁻²) (Figure 2B, eTable 4). In contrast, the absolute risk difference is estimated at 0.95% (95%CI = 0.78% to 1.12%) for women with TT genotype and at 0.79% (95%CI = 0.69% to 0.90%) in women with CC genotype. This indicates that longer OC use is associated with greater risk reduction overall and the risk reduction might be even greater for women with the TT genotype than those with the CC genotype. From Figure 2B-2C we observe the interplay between “nature vs nurture” with risk due to germline genetic mutations offset by long-term use of a modifiable protective factor. This analysis also highlights the benefit of measuring duration of exposure as opposed to a coarse indicator of ever/never use.

Figure 2A–2C.

Estimated absolute risk (AR) of ovarian cancer given OCP use and number of copies of C allele, among non-Hispanic white college graduates aged below 50 with no family history of ovarian cancer, BMI below 25, no tubal ligation, no endometriosis, with one child. The ARs were calculated from a logistic regression model assuming log-additive effect of SNPs while all covariates fixed at their most frequent level as described above. (A) ARs stratified by OCP (ever vs never) and genotype (B) ARs stratified by 1 to 5 years of OCP use (vs < 1 year) and genotype (F) ARs stratified by more than 5 years of OCP use (vs < 1 year) and genotype. Risk differences were also reported as the solid black bar.

Prior work in G × E for ovarian cancer has focused solely on multiplicative interactions. We previously reported no departures from a multiplicative model with the first six risk loci identified through GWAS with a reduced set of exposures ³. Follow-up work identified an interaction with menopausal estrogen therapy use and rs10069690 in the TERT gene ⁴⁸, but that finding was not replicated in the present analysis which included a larger set of studies. Fridley and colleagues have reported on G × E taking a candidate gene approach with several promising findings ⁴⁹. There are several studies in other cancers examining G × E on the multiplicative scale with limited success in identifying interactions, but to our knowledge, only prostate cancer and bladder cancer have been studied on the additive scale. In prostate cancer, suggestive additive interactions between vitamin D, confirmed genetic variants and risk have been identified ⁵⁰. In bladder cancer, additive interaction has been explored between confirmed genetic loci and smoking with risk of disease ³¹. In this work the authors were able to demonstrate that the absolute risk of bladder cancer for current smokers varied from 2.9% to 9.9% based on the polygenetic risk score quartile. These results are similar to our findings on the additive scale with absolute risk differing based on genetics and hormone therapy use; an interesting next step for our work is to consider the polygenetic risk score for all of these confirmed ovarian cancer susceptibility alleles.

There are several limitations of the current analysis. Though we considered both multiplicative and additive interactions, the logistic model in (M1) is linear in covariates and exposures. We ignored potential non-linearity and exposure x exposure as well as exposure x covariate interactions. Similarly, we ignored any higher order interactions. A completely non-parametric machine learning approach, based on a recursive partition of the predictor space may avoid misspecification of the model, but would lack interpretability from an epidemiologic and public health perspective. We also acknowledge that this exploration of interaction is purely statistical, a more causal interpretation in a biological sense will require functional validation. One may also want to explore G × E interaction with loci that are not significant at genome-wide threshold but are significant at a less stringent threshold or even conduct genome-wide G × E scans.

The associations between ovarian cancer risk and some of the variants included here were limited to specific histotypes of ovarian cancer, however we have only presented results for all epithelial ovarian cancers combined. Developing histotype-specific risk stratification approaches is not feasible because for any given histotype the absolute risk is unlikely to ever reach an actionable threshold on a population level. In addition, risk reducing strategies are the same across histotypes and thus there is little benefit to considering histotype specific results from a precision prevention perspective. Heterogeneous associations between environmental risk factors and ovarian cancer risk by histology has previously been well characterized ^{3, 23, 27}. There is value in understanding histotype associations for disease etiology and mechanisms and this will be the focus of future work.

The analyses presented here offer insight into potential biological mechanisms, opportunities for ovarian cancer risk stratification, and approaches to studying gene-environment interactions. Ideally, replication for the six promising findings would be undertaken, but this is challenging with ovarian cancer given that most studies with the relevant data are included here. Functional studies for the regions harboring our most promising findings are underway and it is possible that the association described here may help inform those investigations ⁵¹. Also, gene-environment interaction analyses can also be used to identify novel genetic associations ⁵¹ and thus a deeper evaluation of variants that are still borderline significant, but do not exactly achieve a genome-wide threshold is warranted for subsequent G × E analysis. Of particular interest will be to conduct risk stratification and risk prediction analysis using a summative polygenic risk score and to conduct an agnostic genome-wide search for G × E interaction. Despite the limitations the comprehensive framework of data harmonization, imputation, screening test followed by characterization of effect and risk estimates that has been used in this analysis can serve as a robust model for future gene-environment interaction analyses.

Novelty and Impact:

Our paper conducts gene x environment interaction analysis on both additive and multiplicative scales using data from 9,971 ovarian cancer (OC) cases and 15,566 controls. Seven OC risk factors are considered with 28 variants identified from previous GWAS. The top interaction was between oral contraceptive pill (OCP) use (ever vs never) and rs13255292 (P-value=3.48×10⁻⁴). The protective benefit of OCP use differs by genotype suggesting that prevention strategies need tailoring to an individual’s genotypic profile.

Abbreviations:

AR

absolute risk

BMI

body mass index

BSO

bilateral salpingo-oophorectomy

CI

confidence interval

df

degrees of freedom

G × E

gene-environment interaction

GWAS

genome-wide association study

LRT

likelihood ratio test

OCAC

Ovarian Cancer Association Consortium

OCP

oral contraceptive pill

OR

odds ratio

RD

risk difference

SNP

single nucleotide polymorphism