Background:

The study of gene–environment (GxE) interactions is a research priority for the NCI. Previously, our group analyzed NCI's extramural grant portfolio from fiscal years (FY) 2007 to 2009 to determine the state of the science in GxE research. This study builds upon our previous effort and examines changes in the landscape of GxE cancer research funded by NCI.

Methods:

The NCI grant portfolio was examined from FY 2010 to 2018 using the iSearch application. A time–trend analysis was conducted to explore changes over the study interval.

Results:

A total of 107 grants met the search criteria and were abstracted. The most common cancer types studied were breast (19.6%) and colorectal (18.7%). Most grants focused on GxE using specific candidate genes (69.2%) compared with agnostic approaches using genome-wide (26.2%) or whole-exome/whole-genome next-generation sequencing (NGS) approaches (19.6%); some grants used more than one approach to assess genetic variation. More funded grants incorporated NGS technologies in FY 2016–2018 compared with prior FYs. Environmental exposures most commonly examined were energy balance (46.7%) and drugs/treatment (40.2%). Over the time interval, we observed a decrease in energy balance applications with a concurrent increase in drug/treatment applications.

Conclusions:

Research in GxE interactions has continued to concentrate on common cancers, while there have been some shifts in focus of genetic and environmental exposures. Opportunities exist to study less common cancers, apply new technologies, and increase racial/ethnic diversity.

Impact:

This analysis of NCI's extramural grant portfolio updates previous efforts and provides a review of NCI grant support for GxE research.

Both genetic and environmental factors are known to contribute to cancer etiology (1–5), and risk of cancer is likely due to the interplay between genes and the environment. The study of gene–environment (GxE) interactions, or investigating how genetic variants modify the effect of lifestyle and the environment, is important as it can provide insight into biological processes and mechanisms of cancer etiology, identify individuals who may be more susceptible to cancer, and inform treatment decisions (6, 7). Although considerable work in GxE has been conducted to date, which has been summarized in recent review articles (8–10), the landscape of GxE research continues to be transformed by increasingly complex data types and larger datasets (11–13). Investigators are leveraging advances in next-generation sequencing (NGS; ref. 14) and other -omics data technologies (e.g., metabolomics, transcriptomics, and epigenetics; refs. 15–17) as sources of information to inform studies of genetic susceptibility to cancer. Similarly, researchers are incorporating innovative approaches to assess environmental exposures. Some examples include data from personal monitoring sensors, geographic information systems, or biomarker measurements using omics-based technologies to comprehensively assess an individual's exposure to environmental factors (18). In addition, new statistical methods are being developed to address the analytic challenges of integrating the diverse data types (19). The emergence of these approaches provides new opportunities (11, 20) for researchers to explore more fully the role of GxE in cancer etiology.

The NCI has made the study of GxE research a priority by committing resources to support multiple initiatives. The Institute has issued several funding opportunities announcements (FOA) and sponsored multiple workshops (11, 21, 22) over the past decade underscoring its commitment to this area of research. Because NCI recognizes the importance of GxE research in cancer, we previously conducted an analysis of its extramural grant portfolio from fiscal years (FY) 2007 to 2009 (23). In that portfolio analysis, we noted a number of possible opportunities for further investments in GxE research; they included: developing alternative approaches to exposure assessment, broadening the spectrum of cancer types investigated, developing new analytic and computational methods, and conducting GxE research using an agnostic approach, such as a gene–environment-wide interaction study (GEWIS).

To examine the current state of GxE research at NCI and explore potential changes in the portfolio over time, we conducted an analysis of the NCI extramural grant portfolio from FYs 2010 to 2018. This report characterizes the funded GxE applications and identifies the research gaps that remain.

To identify NCI-funded GxE research grants for inclusion in the portfolio analysis, the NIH's iSearch application was queried for newly awarded competing grants that supported research aims during FYs 2010 through 2018. Grants that were identified for possible inclusion were those that were primary research projects only (i.e., excluding center grants, supplements, and R25, T32, R13, U24, and U10 mechanisms; n = 10,485). Grants from the NCI division of cancer biology were excluded (n = 4,270). From these applications (n = 6,215), grants that contained both a genetic and an environmental search term and had a cancer outcome in the specific aims were selected for inclusion. A complete list of the genetic and environmental search terms can be found in Supplementary Table S1. A total of 429 research grants were initially identified using these search criteria. Four individuals (A.A. Ghazarian, G.Y. Lai, L.E. Mechanic, and N.I. Simonds) each evaluated a quarter of the grant's specific aims to determine whether it should be considered a GxE interaction application and was examining cancer outcomes. Only studies done among human populations were considered. Ten percent of the grants that were identified for possible inclusion (n = 43) were reviewed by all four reviewers together to ensure consistency on how the inclusion and exclusion criteria were being applied with a concordance between individual reviewers of 84%. Discordant results were discussed, and consensus results were recorded.

A total of 107 grants (Fig. 1) were identified as relevant according to the inclusion criteria and specific genetic and environmental information was abstracted from these grants. More specifically, the following genetic terms were captured: candidate gene study, genome-wide association study (GWAS), epigenetic, targeted sequencing, and/or whole-exome/genome sequencing, as well as whether germline and/or somatic variation was being studied. Note that for a grant to be considered GWAS, analyses of GxE interactions were required to be agnostic (i.e., GEWIS). More specifically, even if a grant application used data from GWAS, it was not included in this category if assessment of GxE focused on specific candidate genes or regions that were identified previously. Grants were also characterized according to the following environmental term categories: infection and inflammation, drugs/treatment, exogenous hormones, reproductive factors, chemical environment, physical environment, lifestyle, energy balance, metabolomics, microbiome, social environment, and/or general (Supplementary Table S2). Even though metabolomics and microbiome measures may also be considered intermediate endpoints, they were included as environmental exposures because these measures can reflect possible exposures to environmental factors (24, 25). Environmental and genomic categories were not mutually exclusive as grant applications may examine multiple measures. Specific environmental terms, such as smoking, obesity, and physical activity, were also captured by reviewers. Grants were characterized as methods if they included statistical or analytic methods development. In addition, reviewers determined whether the grant application examined racial/ethnic differences. Finally, data on study design, cancer outcome(s) of interest (defined as cancer diagnosis or cancer survival), and cancer type were abstracted.

Figure 1.

Flow diagram of search strategy and review process of the NCI's extramural grant portfolio on GxE interaction research, FYs 2010–2018. Flow diagram of portfolio analysis search strategy and review process. QC, quality control.

Figure 1.

Flow diagram of search strategy and review process of the NCI's extramural grant portfolio on GxE interaction research, FYs 2010–2018. Flow diagram of portfolio analysis search strategy and review process. QC, quality control.

Close modal

Among the relevant grants, 10% were reviewed by all four reviewers, the remaining were reviewed in pairs to ensure that data abstraction was done consistently for quality control. Concordance between reviewer pairs ranged between 77% and 90% for each batch of grants reviewed. Any discordant results were discussed by all four reviewers and consensus results were recorded. In addition, other consistency checks were conducted on data, such as comparison of all specific environmental terms within environmental exposure categories and review of all specific genetic terms.

A time–trend analysis was conducted to explore whether any changes in research focus occurred over the 9-year study interval. The trend of overall number of NCI-awarded research grants was examined by calculating a Pearson correlation in Microsoft Excel. Relevant grants were grouped into 3-year groups by FY (2010–2012, 2013–2015, and 2016–2018) and temporal trends were explored by genomic categories (GWAS, candidate genes, and whole-exome/whole-genome sequencing) and environmental categories (drugs/treatment, lifestyle, and energy balance). Other categories had limited sample size for exploring these trends and thus, were not evaluated. Because candidate gene and whole-exome/whole-genome approaches may be used to determine both germline and somatic variation, the trend analysis was examined overall and restricted to germline studies. Finally, we evaluated whether changes in the number of GxE methods grant applications were observed over time.

A total of 107 grants were considered relevant according to the selection criteria. The different cancer types examined in the research grants that included studies of GxE interaction research are presented in Fig. 2. The most common cancer type examined was breast (19.6%), followed by colorectal (18.7%), blood cancers (14.0%), melanoma and other skin (9.3%), lung (7.5%), and prostate (7.5%). Less common cancer types in the portfolio (<5%) included: ovarian (4.7%), pancreatic (4.7%), brain/neurologic (3.7%), esophageal (3.7%), liver (3.7%), renal (3.7%), head/neck (2.8%), bladder (1.9%), gastric (1.9%), Kaposi sarcoma (1.9%), and cervical (0.9%). Approximately 10.3% of grants examined more than one cancer type and each cancer type was counted independently (i.e., such grants were included more than once). There were also grants (13.1%) that examined cancer in general, which we classified as “cancer, multiple types.” Eleven of the grants classified as “cancer, multiple types” were developing statistical or analytic methods that may be used for different cancer types. Several of the methods grants specified an individual cancer type (if evaluating a method using a specific dataset), but indicated methods were broadly applicable to other cancers. The other applications in this category were examining GxE in relation to risk of second cancers where any cancer type was included or were assessing prevention across cancer types.

Figure 2.

Distribution of the cancer types examined among the 107 relevant GxE interaction grants funded by the NCI, FYs 2010–2018. Number of cancer types and categories captured in portfolio analysis. The bars are labeled with the total number of grants that investigated each cancer type. The total number of grants is 107; however, some grants investigated more than one cancer type and were counted more than once in the total for this specific figure.

Figure 2.

Distribution of the cancer types examined among the 107 relevant GxE interaction grants funded by the NCI, FYs 2010–2018. Number of cancer types and categories captured in portfolio analysis. The bars are labeled with the total number of grants that investigated each cancer type. The total number of grants is 107; however, some grants investigated more than one cancer type and were counted more than once in the total for this specific figure.

Close modal

Other key variables abstracted from the relevant grants are presented in Table 1. The majority of grants were studies of associations with cancer diagnosis (65.4%) compared with cancer survival (26.2%). The most common study designs used were case–control studies (53.3%) and cohort studies (31.8%). Only 6.5% of grants examined GxE in a randomized control trial. We found 15.9% of the relevant grants involved research exploring differences in racial/ethnic groups and 27.1% were methods-related applications.

Table 1.

Distribution of select grant characteristics among the 107 relevant GxE interaction grants funded by the NCI, FYs 2010–2018.

Grant characteristicsaNumber of grantsPercentage of grants
FY 
 2010 16 15.0% 
 2011 10 9.3% 
 2012 11 10.3% 
 2013 12 11.2% 
 2014 13 12.1% 
 2015 12 11.2% 
 2016 13 12.1% 
 2017 10 9.3% 
 2018 10 9.3% 
Outcome 
 Cancer diagnosis 70 65.4% 
 Cancer survival 28 26.2% 
 Cancer diagnosis and survival 8.4% 
Study design 
 Case–control 57 53.3% 
 Cohort 34 31.8% 
 Randomized control trial 6.5% 
 Multipleb 5.6% 
 Case-only 0.9% 
 Not specified 1.9% 
Health disparity 
 No 90 84.1% 
 Yes 17 15.9% 
Methods 
 No 78 72.9% 
 Yes 29 27.1% 
Genetics categoryc 
 Germline 80 74.8% 
 Somatic 16 14.9% 
 Germline and somatic 7.5% 
 Epigenetic 10 9.3% 
Genetics category (germline-only grants) 
 Candidate 62 77.5% 
 GWAS 24 30.0% 
 Whole exome/whole genome 11.3% 
 Targeted sequencing 3.8% 
 Epigenetics 3.8% 
Genetics category (somatic-only grants) 
 Candidate 37.5% 
 GWAS 0% 
 Whole exome/whole genome 50.0% 
 Targeted sequencing 37.5% 
 Epigenetics 12.5% 
Genetics category (germline and somatic grants) 
 Candidate 62.5% 
 GWAS 50.0% 
 Whole exome/whole genome 50.0% 
 Targeted sequencing 12.5% 
 Epigenetics 37.5% 
Grant characteristicsaNumber of grantsPercentage of grants
FY 
 2010 16 15.0% 
 2011 10 9.3% 
 2012 11 10.3% 
 2013 12 11.2% 
 2014 13 12.1% 
 2015 12 11.2% 
 2016 13 12.1% 
 2017 10 9.3% 
 2018 10 9.3% 
Outcome 
 Cancer diagnosis 70 65.4% 
 Cancer survival 28 26.2% 
 Cancer diagnosis and survival 8.4% 
Study design 
 Case–control 57 53.3% 
 Cohort 34 31.8% 
 Randomized control trial 6.5% 
 Multipleb 5.6% 
 Case-only 0.9% 
 Not specified 1.9% 
Health disparity 
 No 90 84.1% 
 Yes 17 15.9% 
Methods 
 No 78 72.9% 
 Yes 29 27.1% 
Genetics categoryc 
 Germline 80 74.8% 
 Somatic 16 14.9% 
 Germline and somatic 7.5% 
 Epigenetic 10 9.3% 
Genetics category (germline-only grants) 
 Candidate 62 77.5% 
 GWAS 24 30.0% 
 Whole exome/whole genome 11.3% 
 Targeted sequencing 3.8% 
 Epigenetics 3.8% 
Genetics category (somatic-only grants) 
 Candidate 37.5% 
 GWAS 0% 
 Whole exome/whole genome 50.0% 
 Targeted sequencing 37.5% 
 Epigenetics 12.5% 
Genetics category (germline and somatic grants) 
 Candidate 62.5% 
 GWAS 50.0% 
 Whole exome/whole genome 50.0% 
 Targeted sequencing 12.5% 
 Epigenetics 37.5% 

aGrant characteristics were coded as described in the “Materials and Methods.”

bGrants that included more than one study design were considered “multiple.”

cGenetics categories and definitions are provided in Supplementary Table S2.

The majority of the 107 relevant grants (74.8%) examined germline variation, 14.9% examined only somatic alterations, and 7.5% of grants explored both types of variation (Table 1). Most of the germline-only grants (n = 80) examined GxE interactions with specific candidate genes (77.5%) compared with agnostic genome-wide GxE analysis approaches using GWAS (30%) or whole-exome/whole-genome sequencing data (11.3%). In contrast, 37.5% of the somatic-only grants assessed GxE interactions with specific candidate genes. In addition, somatic-only grants and those applications that evaluated both germline and somatic variations frequently explored GxE agnostically using whole-exome/whole-genome sequencing (50% each). Combined (germline, somatic, or germline and somatic grants), most grants focused on GxE using specific candidate genes (69.2%) compared with agnostic approaches using genome-wide (26.2%) or whole-exome/whole-genome NGS approaches (19.6%). The majority (60%) of the targeted sequencing grants were measuring somatic variation and one application was exploring both germline and somatic changes. Finally, approximately 9.3% of grants investigated an epigenetic marker. The majority of grants exploring analytic methods focused on germline variation (89.7% of 29 methods grants).

Environmental exposure data were also abstracted for all relevant grants using the terms described in the Materials and Methods section. The environmental terms that were evaluated in our portfolio analysis of GxE interaction research are presented in Fig. 3. The most common environmental exposures that were examined in the GxE research grants included energy balance (46.7%), followed by drugs/treatment (40.2%), and lifestyle factors (37.4%). Other environmental exposures that were studied included exogenous hormones (11.2%), infection/inflammation (9.3%), reproductive factors (7.5%), chemical environment (5.6%), social environment (5.6%), metabolomics (3.7%), physical environment (2.8%), and microbiome (0.9%). The majority of the methods applications (48.3% of 29 methods applications) also examined lifestyle and/or variables associated with energy balance (41.3% of 29). Ten of the methods applications examined exposures in general or did not list a specific exposure as the focus of methods development. The two most common specific exposures for all relevant applications observed within each environmental exposure category are shown in Table 2. Specific environmental exposures that were most frequently studied included diet (n = 33), smoking (n = 29), alcohol use (n = 18), chemotherapeutic drugs (n = 22), and physical activity (n = 14).

Figure 3.

Distribution among the most common environmental exposure categories examined among the 107 relevant GxE interaction grants funded by the NCI, FYs 2010–2018. Number of environmental exposures captured in portfolio analysis. Environmental exposure categories used in this portfolio analysis are listed, and the bars are labeled with the total number of grants that investigated each environmental exposure category. The total number of grants is 107; however, some grants investigated more than one environmental exposure and were counted more than once in the total for this specific figure.

Figure 3.

Distribution among the most common environmental exposure categories examined among the 107 relevant GxE interaction grants funded by the NCI, FYs 2010–2018. Number of environmental exposures captured in portfolio analysis. Environmental exposure categories used in this portfolio analysis are listed, and the bars are labeled with the total number of grants that investigated each environmental exposure category. The total number of grants is 107; however, some grants investigated more than one environmental exposure and were counted more than once in the total for this specific figure.

Close modal
Table 2.

Most commonly observed environmental exposures in the NCI's extramural grant portfolio on GxE interaction research, FYs 2010–2018.

Environmental exposure categorySpecific exposure
Energy balance Diet (n = 33), physical activity (n = 14) 
Drugs/treatment Chemotherapeutic drugs (n = 22), NSAIDs (n = 10) 
Lifestyle Smoking (n = 29), alcohol use (n = 18) 
Exogenous hormones Hormone replacement therapy (n = 8), oral contraceptive use (n = 2) 
Infection/inflammation Human papillomavirus (n = 3), hepatitis infection (n = 2) 
Reproductive factors Parity (n = 4), age at menarche (n = 4) 
Chemical environment Heavy metals (n = 3), occupational chemical exposures (n = 2) 
Social environment Socioeconomic status (n = 2), neighborhood characteristics (n = 2) 
Metabolomics Metabolites not otherwise specified (n = 3) 
Physical environment Sun exposure (n = 2), ionizing radiation (n = 1) 
Microbiome Microbiome not otherwise specified (n = 1) 
Environmental exposure categorySpecific exposure
Energy balance Diet (n = 33), physical activity (n = 14) 
Drugs/treatment Chemotherapeutic drugs (n = 22), NSAIDs (n = 10) 
Lifestyle Smoking (n = 29), alcohol use (n = 18) 
Exogenous hormones Hormone replacement therapy (n = 8), oral contraceptive use (n = 2) 
Infection/inflammation Human papillomavirus (n = 3), hepatitis infection (n = 2) 
Reproductive factors Parity (n = 4), age at menarche (n = 4) 
Chemical environment Heavy metals (n = 3), occupational chemical exposures (n = 2) 
Social environment Socioeconomic status (n = 2), neighborhood characteristics (n = 2) 
Metabolomics Metabolites not otherwise specified (n = 3) 
Physical environment Sun exposure (n = 2), ionizing radiation (n = 1) 
Microbiome Microbiome not otherwise specified (n = 1) 

Abbreviation: n, number of grants.

To evaluate whether changes in research focus were observed over the 9-year period, we conducted a time–trend analysis. The number of newly awarded competing NCI grants remained stable over the time period (median, 1,160; r = 0.11). The distribution of relevant grants by each FY was also relatively consistent, with the greatest number in FY 2010 (15%). When restricted to studies of germline variation, we observed a decrease in the number of candidate gene applications. Although GxE applications using a GWAS approach remained constant throughout the same time period, the proportion of the funded applications using an agnostic approach (i.e., GWAS or NGS) increased (Table 3). Similar observations were noted when including all applications (n = 28, 27, and 18 candidate gene applications and two, six, and 13 whole-exome/whole-genome sequencing grants during the three time periods). Among the environmental exposures examined for trends, we observed less of an emphasis on energy balance applications over the 9-year time interval, while an increase in the number of applications examining drugs/treatments was found. In contrast, no changes were observed in funded methods-related applications over the time interval (Table 3).

Table 3.

Distribution of GxE interaction grants funded by the NCI from FY 2010 to 2018 classified into 3-year groups, by category.

CategoryNumber of grantsPercentage of grants
Geneticsa 
 Candidate   
  2010–2012 28 41.8% 
  2013–2015 24 35.8% 
  2016–2018 15 22.3% 
 GWAS   
  2010–2012 32.1% 
  2013–2015 10 35.7% 
  2016–2018 32.1% 
 Whole exome/whole genome   
  2010–2012 15.4% 
  2013–2015 38.5% 
  2016–2018 46.1% 
Environmental exposureb 
 Energy balance   
  2010–2012 22 44.0% 
  2013–2015 18 36.0% 
  2016–2018 10 20.0% 
 Lifestyle   
  2010–2012 13 32.5% 
  2013–2015 15 37.5% 
  2016–2018 12 30.0% 
 Drugs/treatment   
  2010–2012 11 25.6% 
  2013–2015 14 32.6% 
  2016–2018 18 41.8% 
Methods 
  2010–2012 10 34.5% 
  2013–2015 31.0% 
  2016–2018 10 34.5% 
CategoryNumber of grantsPercentage of grants
Geneticsa 
 Candidate   
  2010–2012 28 41.8% 
  2013–2015 24 35.8% 
  2016–2018 15 22.3% 
 GWAS   
  2010–2012 32.1% 
  2013–2015 10 35.7% 
  2016–2018 32.1% 
 Whole exome/whole genome   
  2010–2012 15.4% 
  2013–2015 38.5% 
  2016–2018 46.1% 
Environmental exposureb 
 Energy balance   
  2010–2012 22 44.0% 
  2013–2015 18 36.0% 
  2016–2018 10 20.0% 
 Lifestyle   
  2010–2012 13 32.5% 
  2013–2015 15 37.5% 
  2016–2018 12 30.0% 
 Drugs/treatment   
  2010–2012 11 25.6% 
  2013–2015 14 32.6% 
  2016–2018 18 41.8% 
Methods 
  2010–2012 10 34.5% 
  2013–2015 31.0% 
  2016–2018 10 34.5% 

aLimited to the 88 grants examining germline variation (germline-only and those assessing both germline and somatic variation).

bFor the energy balance category, grants were exploring factors that contribute to energy balance (e.g., diet and exercise), but not necessarily specifically assessing the balance of intake/output of energy.

The development and outcome of many cancers are understood to be a product of genetic and environmental factors. Thus, NCI considers understanding the interaction between genetic variation and the environment as an important area of research. We conducted an analysis of the NCI extramural research grant portfolio of GxE interaction research from FY 2010 through 2018 to examine the distribution of cancer types, environmental exposures studied, and the approaches used to assess genetic variation. Not surprisingly, most of the funded grants in GxE research focused on breast, colorectal, and lung cancers, which are all common cancer types with a well-known environmental component (26–28). Approximately 70% of the grants evaluated in this study had a candidate gene target in the GxE analysis. The most commonly studied environmental factor categories included factors contributing to energy balance, drugs/treatment, and lifestyle. Although these findings are similar to those observed in our previous portfolio analysis based on projects active for FYs 2007 through 2009 (23), we noted some differences that reflect the incorporation of new methodologies and approaches (e.g., NGS and new analytic methods), as well as a shift in research focus (e.g., precision oncology and different distribution of cancer types).

The results of this portfolio analysis mirrored the advances in high-throughput genomic technologies. Grantees incorporated NGS technologies, such as whole-genome or whole-exome sequencing, to capture genetic data in their studies; moreover, this shift over time in the technologies being used was more pronounced in grant applications that were funded in FYs 2016 through 2018 compared with prior FYs. In addition, compared with the previous portfolio analysis, a greater number of grantees proposed studying GxE interactions in GWAS (26% vs. 8%). This area was suggested in our previous report (23) as a possible area of opportunity to investigate these associations. Not only have studies using this approach been recently published (29, 30), but grant applications focused on developing such analytic methods to study GxE on a genome-wide scale were observed in this grant portfolio. We also observed some changes as to which environmental exposures were more frequently included in the aims of the abstracted grants. For example, there was more interest in studying interactions with drugs and treatments than in the portfolio analysis of active grants funded in FYs 2007–2009 (40% vs. 29%), likely reflecting the excitement around precision oncology (31, 32). Within this environmental exposure category, chemotherapeutic agents were the most frequently studied class of drugs. Despite this interest, there were only a small number of randomized controlled trial studies (n = 7). For this portfolio analysis, we also expanded our evaluation to other factors related to the environment, namely, the microbiome, metabolomics, and social environment (socioeconomic status and neighborhood characteristics), that have garnered growing interest in cancer epidemiology studies (33–37). Compared with other environmental exposures, we did not observe many grants that included these factors. While these aspects are not considered to be traditional components of the environment in GxE studies, these factors do represent a broad and complex interplay of various external exposures with that of the host, and the observation of these grants may reflect the growing awareness and interest in these fields.

As noted above, investments made to study GxE interactions were mostly focused on common cancers; however, the distribution was different from the results of the prior portfolio analysis (23). Colorectal cancer was more frequently studied compared with the prior analysis (16% vs. 9%). This shift in research focus could be because the reported findings for colorectal cancer appear to show evidence for GxE interactions, particularly for diet and lifestyle factors that are more readily modifiable (38, 39). Also, recent reports of increased colorectal cancer incidence rates observed among younger patients (i.e., those under 50 years of age; refs. 40, 41) has drawn considerable interest among researchers as the etiology accounting for this phenomenon remains unclear, but the rapid increase in incidence is suggestive of an environmental etiology. However, while these circumstances may explain, in part, the percentage shift and focus of GxE research on colorectal cancer, the actual number of grants is still relatively small and could be due to chance.

Only 15% of the grant applications reviewed focused on examining racial/ethnic differences in GxE. The relatively small number of grants exploring health disparities in GxE research is consistent with the documented lack of diversity in cohort studies (42) and genomics research (43, 44). Increasing diversity in research studies could potentially increase the success of GxE research (4). Unfortunately, considerable effort, including policy changes and interventions, is required to help address the persistent underrepresentation of racial/ethnic and other understudied populations in research studies (42, 45, 46). More importantly, as genomic data are being used to inform precision medicine, research gaps may potentially exacerbate current health disparities (47).

In our previous analysis of the NCI extramural research grant portfolio, we found that very few studies were funded to develop new analytic and computational methods related to GxE research. In response to this finding, NCI participated in two FOAs (48, 49) led by the National Institute of Environmental Health Sciences to encourage submissions of applications focusing on the development and application of novel GxE methods, with an emphasis on genome-wide data. We found that the number of grants funded to develop new analytic and computational methods remained stable over time from 2010 to 2018, although the specific focus of these grants has changed. Notably, compared with the previous analysis (2007–2009; ref. 23), there was a large increase in the number of grants which explored analytic methods (1.4% vs. 27.1%). These findings may suggest that the evolution of more high-throughput data has resulted in a corresponding evolution of analytic methods; yet, more methods development in this area is needed (19), including innovative analytic frameworks (e.g., machine learning; ref. 50) that can accommodate the increasingly larger volumes of data collected.

There were a few differences in the design of this analysis compared with the previous study (23). In this study, we included additional genetic (whole-exome/genome sequencing and targeted sequencing) and environmental terms (social environment, microbiome, and metabolomics) to reflect the evolution in how genetic variation and environmental exposures (14, 51, 52) are assessed. We also limited the genetic terms specifically to DNA-related measures in this analysis (the previous included RNA and proteomics). Finally, this portfolio analysis included only newly awarded grants compared with our previous analysis, which included a sample of all relevant active grants between FYs 2007 and 2009. Nevertheless, these differences in study design are unlikely to impact the interpretation of our study findings because of the length of time period explored.

However, there are limitations of this portfolio analysis. First, the analysis does not reflect all NCI-funded cancer researches devoted to GxE interaction (i.e., NCI intramural program and those evaluating cancer outcomes other than risk and survival, such as treatment toxicity). Another limitation is that, we relied primarily on the specific aims of the funded grants, which may not provide a complete picture of the scope of actual research being conducted. Furthermore, we were unable to assess how much of actual GxE research noted in the aims of the applications has been completed. Finally, due to small numbers, we were unable to statistically analyze temporal trends, and our observations are qualitative in nature. Nevertheless, we believe this analysis provides an overall picture of the scope of NCI support for GxE research over the past approximately 10 years.

The findings from our current portfolio analysis suggest that the focus of GxE cancer studies mirrors the evolution of the field. We observed a greater proportion of GxE research being conducted using agnostic approaches in GWAS compared with our previous analysis and a decrease in GxE focused on candidate genes, as well as increased incorporation of NGS technologies over the 9-year time period. We also found a greater emphasis on studying chemotherapeutic agents in this analysis. GxE in cancer research, however, continues to be concentrated in common cancer types and reflects racial/ethnic underrepresentation commonly observed in other research studies, suggesting possible opportunities in these areas. The increasing availability of new technologies for the characterization of genetic and environmental variation and application of these approaches to more diverse populations for studies of GxE in cancer has the potential to provide new insights into underlying mechanisms and health effects associated with the disease.

N.I. Simonds was supported by Scientific Consulting Group during the conduct of the study. No disclosures were reported by the other authors.

The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the NCI.

A.A. Ghazarian: Conceptualization, formal analysis, writing-original draft, writing-review and editing. N.I. Simonds: Conceptualization, formal analysis, writing-original draft, writing-review and editing. G.Y. Lai: Conceptualization, formal analysis, writing-original draft, writing-review and editing. L.E. Mechanic: Conceptualization, formal analysis, writing-original draft, writing-review and editing.

The authors thank Scott Rogers (NCI DCCPS, EGRP) for his assistance with the grant portfolio analysis. This work was supported by the extramural research program of the NCI.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Carbone
M
,
Amelio
I
,
Affar
EB
,
Brugarolas
J
,
Cannon-Albright
LA
,
Cantley
LC
, et al
Consensus report of the 8 and 9th Weinman Symposia on gene x environment interaction in carcinogenesis: novel opportunities for precision medicine
.
Cell Death Differ
2018
;
25
:
1885
904
.
2.
Dempfle
A
,
Scherag
A
,
Hein
R
,
Beckmann
L
,
Chang-Claude
J
,
Schafer
H
. 
Gene-environment interactions for complex traits: definitions, methodological requirements and challenges
.
Eur J Hum Genet
2008
;
16
:
1164
72
.
3.
Murcray
CE
,
Lewinger
JP
,
Gauderman
WJ
. 
Gene-environment interaction in genome-wide association studies
.
Am J Epidemiol
2009
;
169
:
219
26
.
4.
Ritz
BR
,
Chatterjee
N
,
Garcia-Closas
M
,
Gauderman
WJ
,
Pierce
BL
,
Kraft
P
, et al
Lessons learned from past gene-environment interaction successes
.
Am J Epidemiol
2017
;
186
:
778
86
.
5.
Simonds
NI
,
Ghazarian
AA
,
Pimentel
CB
,
Schully
SD
,
Ellison
GL
,
Gillanders
EM
, et al
Review of the gene-environment interaction literature in cancer: what do we know?
Genet Epidemiol
2016
;
40
:
356
65
.
6.
Hunter
DJ
. 
Gene-environment interactions in human diseases
.
Nat Rev Genet
2005
;
6
:
287
98
.
7.
Le Marchand
L
,
Wilkens
LR
. 
Design considerations for genomic association studies: importance of gene-environment interactions
.
Cancer Epidemiol Biomarkers Prev
2008
;
17
:
263
7
.
8.
Cumberbatch
MGK
,
Jubber
I
,
Black
PC
,
Esperto
F
,
Figueroa
JD
,
Kamat
AM
, et al
Epidemiology of bladder cancer: a systematic review and contemporary update of risk factors in 2018
.
Eur Urol
2018
;
74
:
784
95
.
9.
Rudolph
A
,
Chang-Claude
J
,
Schmidt
MK
. 
Gene-environment interaction and risk of breast cancer
.
Br J Cancer
2016
;
114
:
125
33
.
10.
Yang
T
,
Li
X
,
Montazeri
Z
,
Little
J
,
Farrington
SM
,
Ioannidis
JPA
, et al
Gene-environment interactions and colorectal cancer risk: an umbrella review of systematic reviews and meta-analyses of observational studies
.
Int J Cancer
2019
;
145
:
2315
29
.
11.
McAllister
K
,
Mechanic
LE
,
Amos
C
,
Aschard
H
,
Blair
IA
,
Chatterjee
N
, et al
Current challenges and new opportunities for gene-environment interaction studies of complex diseases
.
Am J Epidemiol
2017
;
186
:
753
61
.
12.
Patel
CJ
. 
Analytical complexity in detection of gene variant-by-environment exposure interactions in high-throughput genomic and exposomic research
.
Curr Environ Health Rep
2016
;
3
:
64
72
.
13.
Ritchie
MD
,
Davis
JR
,
Aschard
H
,
Battle
A
,
Conti
D
,
Du
M
, et al
Incorporation of biological knowledge into the study of gene-environment interactions
.
Am J Epidemiol
2017
;
186
:
771
7
.
14.
Rotunno
M
,
Barajas
R
,
Clyne
M
,
Hoover
E
,
Simonds
NI
,
Lam
TK
, et al
A systematic literature review of whole exome and genome sequencing population studies of genetic susceptibility to cancer
.
Cancer Epidemiol Biomarkers Prev
2020
;
29
:
1519
34
.
15.
Flanagan
JM
. 
Epigenome-wide association studies (EWAS): past, present, and future
.
Methods Mol Biol
2015
;
1238
:
51
63
.
16.
Armitage
EG
,
Ciborowski
M
. 
Applications of metabolomics in cancer studies
.
Adv Exp Med Biol
2017
;
965
:
209
34
.
17.
Wu
L
,
Shi
W
,
Long
J
,
Guo
X
,
Michailidou
K
,
Beesley
J
, et al
A transcriptome-wide association study of 229,000 women identifies new candidate susceptibility genes for breast cancer
.
Nat Genet
2018
;
50
:
968
78
.
18.
DeBord
DG
,
Carreon
T
,
Lentz
TJ
,
Middendorf
PJ
,
Hoover
MD
,
Schulte
PA
. 
Use of the “exposome” in the practice of epidemiology: a primer on -omic technologies
.
Am J Epidemiol
2016
;
184
:
302
14
.
19.
Gauderman
WJ
,
Mukherjee
B
,
Aschard
H
,
Hsu
L
,
Lewinger
JP
,
Patel
CJ
, et al
Update on the state of the science for analytical methods for gene-environment interactions
.
Am J Epidemiol
2017
;
186
:
762
70
.
20.
Patel
CJ
,
Kerr
J
,
Thomas
DC
,
Mukherjee
B
,
Ritz
B
,
Chatterjee
N
, et al
Opportunities and challenges for environmental exposure assessment in population-based studies
.
Cancer Epidemiol Biomarkers Prev
2017
;
26
:
1370
80
.
21.
Hutter
CM
,
Mechanic
LE
,
Chatterjee
N
,
Kraft
P
,
Gillanders
EM
,
NCI Gene-Environment Think Tank
. 
Gene-environment interactions in cancer epidemiology: a National Cancer Institute Think Tank Report
.
Genet Epidemiol
2013
;
37
:
643
57
.
22.
Mechanic
LE
,
Chen
HS
,
Amos
CI
,
Chatterjee
N
,
Cox
NJ
,
Divi
RL
, et al
Next generation analytic tools for large scale genetic epidemiology studies of complex diseases
.
Genet Epidemiol
2012
;
36
:
22
35
.
23.
Ghazarian
AA
,
Simonds
NI
,
Bennett
K
,
Pimentel
CB
,
Ellison
GL
,
Gillanders
EM
, et al
A review of NCI's extramural grant portfolio: identifying opportunities for future research in genes and environment in cancer
.
Cancer Epidemiol Biomarkers Prev
2013
;
22
:
501
7
.
24.
Nazaroff
WW
. 
Embracing microbes in exposure science
.
J Expo Sci Environ Epidemiol
2019
;
29
:
1
10
.
25.
Walker
DI
,
Valvi
D
,
Rothman
N
,
Lan
Q
,
Miller
GW
,
Jones
DP
. 
The metabolome: a key measure for exposome research in epidemiology
.
Curr Epidemiol Rep
2019
;
6
:
93
103
.
26.
Barta
JA
,
Powell
CA
,
Wisnivesky
JP
. 
Global epidemiology of lung cancer
.
Ann Glob Health
2019
;
85
:
8
.
27.
Coughlin
SS
. 
Epidemiology of breast cancer in women
.
Adv Exp Med Biol
2019
;
1152
:
9
29
.
28.
Keum
N
,
Giovannucci
E
. 
Global burden of colorectal cancer: emerging trends, risk factors and prevention strategies
.
Nat Rev Gastroenterol Hepatol
2019
;
16
:
713
32
.
29.
Kim
S
,
Wang
M
,
Tyrer
JP
,
Jensen
A
,
Wiensch
A
,
Liu
G
, et al
A comprehensive gene-environment interaction analysis in ovarian cancer using genome-wide significant common variants
.
Int J Cancer
2019
;
144
:
2192
205
.
30.
Li
Y
,
Xiao
X
,
Han
Y
,
Gorlova
O
,
Qian
D
,
Leighl
N
, et al
Genome-wide interaction study of smoking behavior and non-small cell lung cancer risk in Caucasian population
.
Carcinogenesis
2018
;
39
:
336
46
.
31.
Garraway
LA
. 
Genomics-driven oncology: framework for an emerging paradigm
.
J Clin Oncol
2013
;
31
:
1806
14
.
32.
Senft
D
,
Leiserson
MDM
,
Ruppin
E
,
Ronai
ZA
. 
Precision oncology: the road ahead
.
Trends Mol Med
2017
;
23
:
874
98
.
33.
Fearnley
LG
,
Inouye
M
. 
Metabolomics in epidemiology: from metabolite concentrations to integrative reaction networks
.
Int J Epidemiol
2016
;
45
:
1319
28
.
34.
Rattray
NJW
,
Deziel
NC
,
Wallach
JD
,
Khan
SA
,
Vasiliou
V
,
Ioannidis
JPA
, et al
Beyond genomics: understanding exposotypes through metabolomics
.
Hum Genomics
2018
;
12
:
4
.
35.
Sinha
R
,
Ahsan
H
,
Blaser
M
,
Caporaso
JG
,
Carmical
JR
,
Chan
AT
, et al
Next steps in studying the human microbiome and health in prospective studies, Bethesda, MD, May 16–17, 2017
.
Microbiome
2018
;
6
:
210
.
36.
NIH Human Microbiome Portfolio Analysis Team
. 
A review of 10 years of human microbiome research activities at the US National Institutes of Health, Fiscal Years 2007–2016
.
Microbiome
2019
;
7
:
31
.
37.
Zahnd
WE
,
McLafferty
SL
. 
Contextual effects and cancer outcomes in the United States: a systematic review of characteristics in multilevel analyses
.
Ann Epidemiol
2017
;
27
:
739
48
.
38.
Murphy
N
,
Moreno
V
,
Hughes
DJ
,
Vodicka
L
,
Vodicka
P
,
Aglago
EK
, et al
Lifestyle and dietary environmental factors in colorectal cancer susceptibility
.
Mol Aspects Med
2019
;
69
:
2
9
.
39.
Ahmed
FE
. 
Gene-gene, gene-environment & multiple interactions in colorectal cancer
.
J Environ Sci Health C Environ Carcinog Ecotoxicol Rev
2006
;
24
:
1
101
.
40.
Ahnen
DJ
,
Wade
SW
,
Jones
WF
,
Sifri
R
,
Mendoza Silveiras
J
,
Greenamyer
J
, et al
The increasing incidence of young-onset colorectal cancer: a call to action
.
Mayo Clin Proc
2014
;
89
:
216
24
.
41.
Siegel
RL
,
Miller
KD
,
Goding Sauer
A
,
Fedewa
SA
,
Butterly
LF
,
Anderson
JC
, et al
Colorectal cancer statistics, 2020
.
CA Cancer J Clin
2020
;
70
:
145
64
.
42.
Martin
DN
,
Lam
TK
,
Brignole
K
,
Ashing
KT
,
Blot
WJ
,
Burhansstipanov
L
, et al
Recommendations for cancer epidemiologic research in understudied populations and implications for future needs
.
Cancer Epidemiol Biomarkers Prev
2016
;
25
:
573
80
.
43.
Bustamante
CD
,
Burchard
EG
,
De la Vega
FM
. 
Genomics for the world
.
Nature
2011
;
475
:
163
5
.
44.
Popejoy
AB
,
Fullerton
SM
. 
Genomics is failing on diversity
.
Nature
2016
;
538
:
161
4
.
45.
Alcaraz
KI
,
Wiedt
TL
,
Daniels
EC
,
Yabroff
KR
,
Guerra
CE
,
Wender
RC
. 
Understanding and addressing social determinants to advance cancer health equity in the United States: a blueprint for practice, research, and policy
.
CA Cancer J Clin
2020
;
70
:
31
46
.
46.
Vuong
I
,
Wright
J
,
Nolan
MB
,
Eggen
A
,
Bailey
E
,
Strickland
R
, et al
Overcoming barriers: evidence-based strategies to increase enrollment of underrepresented populations in cancer therapeutic clinical trials-a narrative review
.
J Cancer Educ
2020
;
35
:
841
9
.
47.
Martin
AR
,
Kanai
M
,
Kamatani
Y
,
Okada
Y
,
Neale
BM
,
Daly
MJ
. 
Clinical use of current polygenic risk scores may exacerbate health disparities
.
Nat Genet
2019
;
51
:
584
91
.
48.
Department of Health and Human Services.
Analysis of genome-wide gene-environment (GxE) interactions (R21).
Available from
: https://grants.nih.gov/grants/guide/pa-files/PAR-13-382.html.
49.
Department of Health and Human Services.
Methods and approaches for detection of gene-environment interactions in human disease (R21).
Available from:
https://grants.nih.gov/grants/guide/pa-files/par-11-032.html.
50.
Li
J
,
Li
X
,
Zhang
S
,
Snyder
M
. 
Gene-environment interaction in the era of precision medicine
.
Cell
2019
;
177
:
38
44
.
51.
Dennis
KK
,
Marder
E
,
Balshaw
DM
,
Cui
Y
,
Lynes
MA
,
Patti
GJ
, et al
Biomonitoring in the era of the exposome
.
Environ Health Perspect
2017
;
125
:
502
10
.
52.
Visscher
PM
,
Wray
NR
,
Zhang
Q
,
Sklar
P
,
McCarthy
MI
,
Brown
MA
, et al
10 years of GWAS discovery: biology, function, and translation
.
Am J Hum Genet
2017
;
101
:
5
22
.