Sociodemographic and Clinical Characteristics Associated with Genetic Testing among Cancer Survivors: Evidence from Three Cancer Registries

Cancer is a leading cause of death [1], with genetics playing a crucial role in both its development and treatment [2‒4]. Targeted therapies, based on an individual’s genetic makeup, have revolutionized cancer treatment and improved outcomes for patients [2, 5]. Pharmacogenomics (PGx), which identifies how a patient’s genetic profile influences their response to drugs, has gained increasing importance in cancer care [6]. In fact, the proportion of new drugs approved with pharmacogenomics labeling has nearly tripled, rising from 10.3% in 2000 to 28.2% in 2020 [7]. By utilizing genetic testing, healthcare providers can develop personalized treatment plans, leading to improved treatment outcomes and reduced side effects [3]. Incorporating genetic testing, including germline and tumor testing, into cancer care is vital for optimal cancer management, surveillance, family screening, and secondary prevention [3, 4].

Clinical guidelines recommend germline testing for cancer susceptibility genes and somatic tumor testing for pathogenic variants in patients diagnosed with cancer [2, 8, 9]. For example, the National Comprehensive Cancer Network (NCCN) guidelines suggest germline testing for individuals with abnormal tumor testing and those with a family history of Lynch syndrome-associated cancer [9]. Similarly, the American Society of Clinical Oncology (ASCO) recommends routine genetic sequencing for metastatic cancer patients with specific genetic alterations [2]. However, the extent to which genetic testing (both germline and tumor testing) is utilized in populations of patients diagnosed with cancer remains unclear. Some existing evidence examines genetic testing uptake and patterns in the general population or for direct-to-consumer genetic testing (e.g., AncestryDNA, 23andMe) [10‒12], suggesting that overall receipt of cancer genetic testing is low and there are disparities based on socioeconomic status and perceptions of cancer risk [12, 13]. Evidence specifically focused on cancer patients undergoing genetic testing for clinical purposes remains limited.

Although cancer patients report a wide variation in genetic testing utilization patterns, ranging from 5% of patients with colorectal cancer to 30% of those with breast cancer [14‒18], these estimates either relied on data from a single healthcare system [18] or older data (i.e., from 2005 to 2015) that do not reflect more recent guidelines for genetic testing [15]. Because the current evidence is limited, further studies are needed to provide updated information on the prevalence and patterns of genetic testing among cancer populations. To address this gap, we utilized data from three US cancer registries, with the objective of examining population-based rates of genetic testing and assessing patient-provider discussions regarding genetic testing. The goal is to gain a comprehensive understanding of potential barriers and opportunities to improve the integration of genetic testing into oncology practice.

This study used data from the Health Information National Trends Survey (HINTS) – Surveillance, Epidemiology, and End Results (SEER) pilot project conducted by the National Cancer Institute (NCI) from January to July 2021 [19]. HINTS is a nationally representative survey collecting data on Americans’ health-related information needs, access, utilization, behaviors, perceptions, and knowledge. The HINTS-SEER linked database was developed to address the limitation of small sample sizes of cancer survivors in individual HINTS administrations while adhering closely to established HINTS protocols. The major modification was the additional sampling of cancer survivors from three SEER cancer registries: the Iowa Cancer Registry, the New Mexico Tumor Registry, and the Greater Bay Area Cancer Registry [19]. Further details about HINTS-SEER are documented elsewhere [19, 20]. In this study, we included individuals diagnosed with invasive cancers, aged 18 years or older as of December 31, 2020 (or December 1, 2020, for the Greater Bay Area Cancer Registry), whose last contact with the SEER registry was no earlier than January 1, 2016, and whose date of diagnosis was prior to 2018 based on certified data. Survivors with a sole diagnosis of non-melanoma skin cancer were excluded. This study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline [21].

The outcomes of interest were cancer survivors’ self-reported utilization of genetic testing, including lifetime germline genetic testing for inherited cancer risk and tumor genetic profiling (also known as somatic testing) conducted as a part of cancer diagnosis and/or treatment. Germline genetic testing identifies inherited gene mutations that may increase cancer risk, such as BRCA 1/2 mutations for breast and ovarian cancer. Respondents were classified as having undergone germline testing if they answered “yes” to the survey question: “Have you ever had cancer genetic testing (for example, testing for inherited cancer syndromes like BRCA 1/2 or Lynch syndrome)?” To assess respondents’ experiences with tumor profiling, we used the HINTS survey question: “Was genetic testing on the cancer tumor or tissue done as part of your cancer diagnosis and/or treatment?” to determine whether testing occurred. We also used the question “Was genetic testing on the cancer tumor or tissue ever discussed with you?” to determine whether respondents discussed tumor profiling with their healthcare provider. By classifying responses to these questions as binary yes/no outcomes, we operationalized genetic testing experiences to quantify how many respondents underwent each type of testing and discussed tumor profiling with their providers.

Demographic characteristics included age at the time of the survey (18–49, 50–64, 65–74, 75+), sex (male, female), race and ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, and others). Socioeconomic status was assessed through educational attainment (high school or less, some college, college graduate or higher), household income in the past year (<USD 50,000, USD 50,000–99,999, >USD 100,000), and current health insurance type (private, public, and uninsured). Detailed information on cancer was collected, including the year of diagnosis (before 2002, 2003–2010, 2011–2014, 2015–2017); HINTS-SEER cancer site (head and neck, gastrointestinal tract, respiratory system, skin-melanoma, breast, female reproductive system, male reproductive system, urinary tract, endocrine system, lymphomas/lymphoblastic leukemias, and others); time since diagnosis (less than or equal to 1 year, 2–5 years, 6–10 years, 11+ years); cancer stage (localized, regional, distant, unknown); and family cancer history. Chronic conditions included a history of diagnosis of diabetes, high blood pressure, heart condition, lung disease, and depression. Geographic information was also gathered on urban/rural designation (metro, micropolitan, small town, rural area). The SEER registry site (Iowa, New Mexico, California-Bay Area) was controlled for the differences in sampling procedures [19].

Statistical analyses were conducted using Stata 17 (StataCorp LLC.) in accordance with HINTS-SEER data analysis recommendations, accounting for complex sampling weights [19]. Jackknife and Balanced Repeated Replication methods were employed for variance estimation, ensuring the final estimates are representative of the three registries that participated in the study. According to HINTS-SEER reporting guidelines, data with counts below the recommended threshold of 25 were considered suppressed; thus, we report weighted numbers in this study to mitigate potential disclosure risks. Weighted sample sizes and percentages were calculated for participants’ demographic characteristics, socioeconomic status, lifestyle factors, cancer history, chronic conditions, and geographic locations for the overall sample and by the three primary self-reported genetic testing experiences: prior germline testing, prior tumor testing, and prior discussion of tumor testing with a healthcare provider. Wald χ² tests were used to examine bivariate associations between survivors’ characteristics and self-reported experiences with genetic testing. To investigate whether self-reported genetic testing experience differed by test eligibility criteria, we performed subgroup analyses by self-reported family cancer history (whether any first- or second-degree biological relatives had a history of cancer or not), cancer stage (localized, regional, or distant), and germline indication status (diagnosed with colon, pancreatic, ovarian, endometrial, prostate, melanoma vs. all others) [8, 9].

The study analyzed 1,234 cancer survivors (weighted population = 415,978) with an average age of 70.5 years at the time of the survey. Nearly 55% had been diagnosed more than 10 years ago. Of these survivors, 54.9% were female, 76.7% were non-Hispanic White, and 81.3% lived in metropolitan areas (Table 1). As shown in Table 2, 17.0% of survivors reported using germline genetic testing, 8.5% reported having tumor testing for cancer diagnosis and/or treatment, and 8.6% reported discussions about tumor genetic testing with their healthcare providers. The proportion of survivors who reported germline genetic testing was higher among younger survivors aged 18–49 years at the time of the survey (11.4 vs. 4.2% among those who had no test; p < 0.001), females (83.0 vs. 47.5%; p < 0.001), college graduates (69.7 vs. 54.7%; p < 0.001), and those with employer-sponsored health insurance (36.9 vs. 17.4%; p < 0.001). The proportion of survivors who reported tumor genetic testing was highest among those diagnosed in 2015–2017 (32.8 vs. 21.5% among those who had no test; p = 0.035) or within the past year (4.9 vs. 0.5%; p < 0.001). Similarly, the proportion of survivors diagnosed in 2015–2017 who reported discussing tumor testing was higher (33.1 vs. 21.4% among those who had no test discussion; p = 0.011) or those who lived in a metropolitan area (89.1 vs. 80.6%; p = 0.044) (Table 2).

When testing reports were stratified by germline indication and cancer type (Fig. 1), 20.9% of survivors with a possible germline testing indication reported undergoing testing, significantly higher than the 6.6% for those whose cancer does not have a germline test indication (p < 0.001). Reports of germline and tumor testing varied significantly by cancer type, with ovarian cancer survivors having the highest overall proportion of survivors reporting prior testing (43.0% germline, 21.3% tumor), followed by breast and endometrial cancer survivors. Overall, survivors with ovarian and breast cancers had a 7.0–36.4% higher prevalence of testing compared to those with other cancer types without germline indication (Fig. 1b). Survivors’ self-reported family history did not impact genetic testing experiences (Fig. 2a), even when stratified by germline testing indication (Fig. 2b, c). When examining testing experiences by cancer stage (Fig. 3), only differences in reports of discussing testing with a healthcare provider were significantly different by stage: localized (7.4%), regional (10.6%), and distant (17.3%) (p = 0.037; Fig. 3a). Among survivors without a potential germline testing indication, survivors with distant-stage cancer had higher reports of tumor testing (17.4%) and testing discussions (22.9%) compared to those with localized or regional disease (Fig. 3c).

In this population-based study linking cancer registry and survey data, 17.0% of cancer survivors reported undergoing germline testing, while 8.5% reported tumor testing. However, discussions about tumor testing with healthcare providers were reported by only 8.6% of survivors. Despite increasing recognition of the clinical utility of genetic testing to guide cancer treatment and prevention [2, 8, 9], testing prevalence remains low across most cancer types [12, 22, 23]. An exception was observed in ovarian and breast cancer, where nearly half of the survivors had undergone germline testing, which may reflect the clinical utility of identifying actionable BRCA and other mutations to guide treatment and screening recommendations in these cancers [14, 24, 25]. There has been a significant uptick in germline testing among patients with ovarian and breast cancer, reflecting a growing recognition of its importance [24, 25]. Specifically, the proportion of ovarian cancer patients receiving genetic testing within a year of diagnosis surged from 23.0% in 2011 to 52.9% by 2020 [24]. Similarly, the percentage of breast cancer patients undergoing genetic testing within the same timeframe also saw a marked increase, climbing from 37.0% in 2011 to an impressive 67.9% in 2020 [24].

Our findings showed that some cancer types have a lower prevalence of routine genetic testing, suggesting that different cancer sites may adopt genetic testing at varying rates [26]. Although there was an observable trend regarding discussions about tumor testing as the cancer stage advanced, this trend did not seem to translate into the uptake of genetic testing overall. However, in our subgroup analysis by germline testing indication, there was a significant difference observed for those with distant-stage cancer, who showed a higher prevalence of tumor testing and discussions about the testing, compared to their counterparts with localized or regional stages. Such findings are consistent with the trend of targeted therapy adoption. It was reported that about 20% of patients with non-small cell lung cancer with targetable variants are potential beneficiaries of targeted therapies to improve overall survival by years [27]. These findings suggest a nuanced landscape in the adoption of genetic testing across different cancer types and stages [26, 28]. While genetic testing can provide valuable information, it is important to note that it may not be clinically indicated or recommended for all patients, particularly those with early-stage cancer without inherited genetic predisposition [29].

Precision medicine holds great promise for advancing cancer care when clinically appropriate. However, it is also true its integration into clinical practice appears to face many barriers [26, 29‒31]. One major obstacle is the limited knowledge and training in germline and somatic testing among physicians, including both primary care providers and oncologists [30, 31]. The lower utilization of germline testing seen among cancer types (e.g., colorectal, pancreatic, prostate, melanoma) in this study may suggest opportunities for more appropriate use of germline testing, consistent with previous study findings [16]. It may also shed light on the importance of aligning somatic testing with germline testing to facilitate clinical decision-making. The 2019 updates to certain oncology practice guidelines recommend that germline testing follow tumor sequencing in patients meeting various criteria to determine whether particular variants are of germline or somatic origin. Walker and colleagues [32] reported that tumor somatic testing following germline testing increased the detection of actionable alterations by 5-fold, which may be translated into better survival outcomes for drug-targetable patients. Taken together, growing evidence suggests the need for expanded education and training opportunities in genomics and molecular diagnostics for physicians across specialties [30, 31]. Targeted efforts to enhance their genomic literacy and ability to interpret and act upon genetic test results would be critical for realizing the potential of precision oncology care and ensuring equal access to these advances for all patients [29, 33]. Additionally, the paucity of access to genetic counselors, lack of insurance coverage for genetic counseling services, and high out-of-pocket costs pose significant barriers [34, 35]. Many physicians are hesitant to order germline testing without a formal genetic counseling consultation and informed consent process, given the complex documentation requirements [36]. Thus, addressing workforce shortages and inadequate reimbursement for genetic counseling would be vital for facilitating the responsible adoption of genomic medicine in clinically appropriate situations [36, 37].

Marked differences emerged in genetic testing experiences by age, gender, socioeconomic status, and geographic classification (urbanity). Younger survivors were twice as likely to receive germline testing compared to seniors aged 75+ which may reflect increasing awareness and acceptance of genetic testing among younger generations [38]. However, implicit age-related biases among ordering providers cannot be ruled out as another factor influencing lower testing prevalence in older patients [39]. While early age of onset is an established indicator for genetic testing, seniors who develop cancer may still benefit from testing to guide therapy or inform family risk [28, 29, 40]. Females had four times the testing rate of males. Although breast cancer survivors stand out for their high rates of genetic testing, gender-based differences may be due to variations in healthcare-seeking behavior, perceived risk, and concerns for personal and familial health between men and women [41‒43]. However, the higher rate among females might also be influenced by socioeconomic factors impacting healthcare access, provider biases, and social norms rather than individual behaviors alone [41‒43]. Disparities were also noted based on socioeconomic status and geographic location, with those having higher educational attainment and private insurance and those residing in metropolitan areas also more likely to have genetic testing, suggesting potential disparities in access and awareness across socioeconomic and geographic boundaries [12, 13, 44]. Taken together, developing tailored education materials and strategic outreach efforts to promote genetic testing in underserved communities, including those with lower educational attainment, underinsured or uninsured status, and limited proximity to testing facilities, is crucial for enhancing equitable access and utilization among diverse patient populations [12, 44, 45].

Limitations of the study include the potential underrepresentation of certain demographic groups (non-Hispanic Black) and geographic regions (Iowa, New Mexico, California-Bay Area), limited generalizability to all US cancer survivors, and the lack of information on reasons for not undergoing genetic testing, such as patient or provider preferences and insurance coverage issues. Moreover, the cross-sectional design did not allow us to directly disentangle the effects of age and time period on genetic testing prevalence. As genetic testing has become more widely available and adopted over time [2, 8, 9], the observed differences in testing by age group may be confounded by secular trends. An important limitation to note is that nearly 55% of the study participants were diagnosed more than 10 years ago, reflecting testing practices from over a decade ago when genetic testing was less common and accessible. For example, guidelines recommending genetic testing for certain cancer types, such as ovarian cancer, were not widely implemented until the late 2010s [44]. This older time period of diagnosis could introduce recall bias and may not accurately represent current genetic testing utilization rates. Although these data provide a valuable historical context and baseline that could be leveraged to better understand the changing genetic testing landscape, future longitudinal studies would be needed to separate the independent influences of patient age more rigorously versus the year of cancer diagnosis on testing utilization. Another limitation point was that the survey was administered during the COVID-19 pandemic when the healthcare system was significantly strained, and many providers restricted services to only essential visits. The pandemic introduced many confounding variables that may have impacted general access to care, including genetic counseling and testing [46]. However, only a small fraction (1.04% weighted) of the cancer survivors in this study were diagnosed less than 1 year prior, suggesting the impact of these pandemic-related disruptions on our findings was likely minimal. Nevertheless, our study findings indicate that there is still inadequate access to and utilization of cancer genetic testing. Integrating genetic counselors and simplifying testing protocols could help ensure appropriate patients are offered testing [47]. Furthermore, educating providers on emerging genetic testing guidelines for cancer management across diagnoses may promote more equitable adoption [48]. Additional research should examine barriers to genetic testing uptake among specific survivor subgroups to tailor implementation strategies.

Our findings have revealed that a higher proportion of cancer survivors who were young, female, with college degrees, and private insurance had germline testing, particularly those with cancer types with germline indications, including ovarian and breast cancers. Additionally, our study shows that survivors diagnosed in recent years or with advanced-stage cancer were more likely to discuss tumor testing with their healthcare providers. However, the overall prevalence of genetic testing seemed to be low, suggesting potential barriers that may prevent equitable access for many cancer survivors. It is worth noting that these findings do not necessarily indicate the current overall rate of testing is problematic or concerning, as genetic testing should be tailored based on individual patient factors; the observed socioeconomic disparities highlight potential inequities in access and awareness. Genetic testing can provide valuable information to guide treatment decisions, but its optimal utilization depends on numerous patient-specific factors. Therefore, further research is needed to better understand the barriers and educational needs of cancer patients, caregivers, and providers to optimize genetic testing strategies, helping them make informed decisions about genetic testing and its potential benefits for their cancer treatment and survivorship.

This study used de-identified and publicly available data from the National Cancer Institute (NCI) and was approved as an exemption by the Institutional Review Board at the University of Florida. Patient consents were not required as this study was based on publicly available data.

The authors have no conflicts of interest, financial or otherwise, to disclose.

This study was partly supported by the UF Health Cancer Center Pilot Grant (#CCPS202103). The funders had no role in the design of the study; the collection, analyses, or interpretation of data; the writing of the manuscript; or the decision to publish the results.

Y.-R.H., T.J.G., and D.B. conceived the study. Y.-R.H., R.W., and G.C. carried out the analysis and drafted the first manuscript. M.K., S.V., J.B., T.J.G., and D.B. provided critical feedback and helped shape the manuscript. All authors discussed the results and edited the manuscript. All authors critically read and revised the manuscript and gave final approval for publication.

The data used in this study are from the Health Information National Trends Survey (HINTS), sponsored by the National Cancer Institute (NCI). HINTS data are publicly available and can be accessed directly through the NCI’s HINTS website: https://hints.cancer.gov/.