Background

Pathogenic variants in the tumor suppressor breast cancer genes BRCA1 and BRCA2 increase the risk of female breast and ovarian cancers. Approximately 57% of women with an inherited BRCA1 pathogenic variant and 49% with a BRCA2 pathogenic variant develop breast cancer by 70 years of age [1], compared to a lifetime risk of 12.5% in women in the general population [2].

Testing for BRCA variants may impact cancer prevention or treatment decisions in patients at heightened risk of hereditary breast cancer [3]. One study assessing 220 breast cancer patients reported a significantly higher proportion of contralateral prophylactic mastectomy, which may positively impact survival, in patients who were aware of their BRCA variant status compared to patients who were not aware (76.4% vs. 14.7%) [4]. Alternatively, identifying pathogenic variant carriers allows risk reducing salpingo-oophrectomy which has been shown to improve survival [5, 6]. Currently, most patients with breast or ovarian cancer receive chemotherapy or hormone therapy as first-line treatment, regardless of BRCA status. As new, targeted therapies for breast cancer such as poly (ADP-ribose) polymerase (PARP) inhibitors are approved for use, knowledge of BRCA status is essential for understanding optimal treatment options for breast cancer patients.

National Comprehensive Cancer Network (NCCN) guidelines recommend genetic counseling and testing for BRCA1/BRCA2 status for patients at heightened risk of having pathogenic variants due to personal or family history. Testing for BRCA pathogenic variants has been available since 1996, and the rate of testing among breast cancer patients has been previously assessed. In 2005, Brown et al. reported that among 551 surveyed women with early onset breast cancer, only 45% had discussed genetic testing with their physicians and/or been referred to a genetic counselor; those with a family history of cancer (53%) and Ashkenazi Jewish women (81%) were more likely to have been referred to a genetic counselor [7]. Additionally, a 2011 study reported that only 29.1% of patients with breast cancer were referred for genetic counseling and/or testing by their physicians [8].

Despite the importance of BRCA testing for breast cancer treatment decisions, few recent studies have assessed the rate of testing among breast cancer patients at an increased risk for pathogenic variants. A study by Wood et al. found that, among breast cancer patients diagnosed between 2009 and 2011, 52.2% of patients with an increased risk for hereditary breast cancer were referred for genetic counseling and/or testing [8]. This proportion may have changed following 2013, due to expanded availability of BRCA testing. Understanding BRCA1/BRCA2 testing rates in breast cancer patients at a higher risk of having pathogenic variants, especially those treated in a community oncology setting where testing rates have traditionally been lower than in academic settings [9], will help inform strategies to improving awareness and testing of BRCA in eligible patients. The present study assessed the proportion of recently diagnosed, high risk, breast cancer patients treated in a US community oncology setting, who were tested for BRCA1/BRCA2 variants in accordance with NCCN guidelines. The proportion of tested patients who were positive for a BRCA1 or BRCA2 pathogenic or likely pathogenic variant was also assessed.

Methods

Data source, study design, and population

An online physician panel approach was used to recruit medical oncologists for this retrospective chart review study, conducted from June 2017 to April 2018. Cardinal Health Specialty Solutions sent open invitations to its panel of community-based oncologists who treat breast cancer in the US. Eligible oncologists were required to be able to participate in research approved by a central institutional review board (IRB) and to have treated at least one patient with breast cancer in the past year.

Oncologists who elected to participate in this study were blinded to the study sponsor and not informed of the study objective. They randomly selected patient charts based on a “random-letter” generating algorithm used to identify patient last names, to mitigate concerns about selection bias, and completed a web-based electronic case report form (eCRF) to collect secondary data from up to 10 breast cancer patients. To be eligible for this study, patients were required to have a confirmed diagnosis of breast cancer between January 2013 and October 2017 and an accompanying pathology report such that hormone receptor (HR) status (i.e., estrogen receptor [ER], progesterone receptor [PR]) and Human Epidermal Growth Factor Receptor 2 (HER2) status were known. They were required to be 18 years of age or older at initial breast cancer diagnosis, have information on family history of cancer, and have an increased risk of hereditary breast cancer. Criteria to identify increased hereditary risk were developed based on NCCN testing guidelines (as available from 2013 to 2017, across all years under study) and those commonly used to identify high risk patients in clinical practice (Additional file 1). Hereditary risks included personal and family history of breast cancer or other cancers, and Ashkenazi Jewish ancestry. Patient selection was designed to oversample triple negative breast cancer (TNBC) and metastatic breast cancer patients.

The primary outcomes of the study were BRCA testing status and results of testing. In addition, data on physician characteristics (e.g., primary medical specialty, practice setting and size, presence of a genetic counselor in their practice, and familiarity with NCCN guidelines) and patient characteristics (e.g., demographic information, breast cancer pathology, and personal and family history of breast and other cancers) were also collected through the eCRF.

Statistical analysis

Descriptive analyses were conducted to characterize physicians who participated in the study and to summarize demographic and clinical information for breast cancer patients meeting eligibility criteria, for whom data were collected.

The proportion of high risk patients who were tested in accordance with NCCN BRCA testing guidelines was calculated overall for the total study population and for risk groups defined based on NCCN testing guidelines released during the study period (2013–2017) (Additional file 1). Patients could belong to more than one high risk group. Stratified analyses by year were conducted to assess trends over the study period. The proportion of these patients who tested positive for a BRCA1/BRCA2 pathogenic variant was also assessed in each risk group.

Stratified analyses were performed in patients with metastatic and triple negative breast cancer, patients with metastatic and HER2 negative (−) breast cancer, African American patients, patients treated in practices with a genetic counselor, and patients treated in practices without a genetic counselor.

All analyses were conducted in SAS 9.4 (SAS Institute, Cary, N.C.).

Results

Physician and patient characteristics

A total of 63 oncologists elected to participate in this study, of whom the majority were general oncologists (97%). The participating oncologists had, on average, 16 years of practice and were proportionally distributed across the Northeast, Midwest, South, and West. They reported treating an average of 154 breast cancer patients in the previous year. All oncologists indicated familiarity with NCCN BRCA guidelines, and 67% had a genetic counselor as part of their practice (Table 1).

Table 1 Physician and Practice Characteristics
Full size table

Among 410 patients included in the study, the mean age at breast cancer diagnosis was 52 years, the majority (95%) were female, White (74%), and 19% were of Ashkenazi Jewish ancestry (Table 2). At diagnosis, 126 patients (31%) had TNBC and 124 patients (30%) had metastatic breast cancer, of whom 105 (85%) were HER2(−) and 34 (27%) were triple negative. Patients with metastatic and triple negative breast cancer appeared to have a lower mean age at diagnosis (51 vs. 57 years) and were less likely to have been previously treated for ovarian cancer (3% vs. 5%) compared to patients with metastatic and HER2(−) breast cancer (Additional file 1).

Table 2 Demographic and Clinical Characteristics in All Patients
Full size table

BRCA testing and variant status

Among the 410 patients included in this study, 384 (94%) were tested for BRCA1/BRCA2 variants; 24 were not tested, and 2 had an unknown testing status (Table 3). For 75% of the 24 patients who were not tested, physicians deemed that genetic testing would have had no impact on medical management (data not shown). The testing rate was 100% among female patients with at least one close blood relative with a BRCA pathogenic variant (n = 68), female patients with a diagnosis of breast cancer before age 50 who had an additional primary breast cancer (n = 12), and female patients with a personal history of ovarian cancer (n = 18). Testing rates over the study period were assessed by year, and tended to range from 88 to 95% from 2013 to 2017, suggesting there was no major change in testing rates over the study period (Additional file 1).

Table 3 Testing for BRCA Pathogenic Variants in Accordance with NCCN Guidelinesa in All Patients
Full size table

Female patients with male-relative related risk factors had slightly lower testing rates. For example, female patients diagnosed at age ≤ 50 years with at least one close blood relative with prostate cancer (Gleason score ≥ 7), had a lower testing rate of 87% and female patients with at least one close male blood relative with breast cancer had the lowest testing rate of 78% (Table 3). Testing rates were also stratified by year to assess trends over the course of the study period. Due to the small sample size for each of the high risk groups in each year, no clear trends were discernable (Additional file 1).

The proportion of patients who tested positive for a BRCA pathogenic variant in various risk groups ranged from 22 to 78% (Table 3). Among the 68 patients who had at least one close blood relative with a known BRCA1/2 pathogenic variant, 53 (78%) had a positive result, whereas 62 out of the remaining 316 patients (20%) had a positive result.

Almost all (97%) metastatic and triple negative breast cancer patients were tested for BRCA pathogenic variants, and 52% tested positive (Table 4). Among patients with metastatic and HER2(−) breast cancer, 93% underwent BRCA testing, of whom 40% tested positive for a BRCA pathogenic variant. The testing rate for BRCA pathogenic variants in African American patients was slightly lower (91%) than for the overall population. In patients treated in practices with a genetic counselor, 95% were tested for BRCA pathogenic variants, compared to 91% of patients treated in practices without a genetic counselor (Table 4).

Table 4 Testing for BRCA Pathogenic Variants by Patient and Physician Characteristics
Full size table

Solo practitioners and physicians in small private community practices (1–5 physicians) had the lowest BRCA testing rate (92%) (Additional file 1). Additionally, the testing rate was lower for physicians with a greater number of years of experience (21 or more years of experience: 90%) compared to physicians with fewer years of experience (0–11 years of experience: 99%), and for those who treated ≤50 breast cancer patients in the past year (88%) compared to physicians who treated over 50 patients (> 92%).

Conclusions

This study demonstrates high rates of BRCA testing among recently diagnosed breast cancer patients at high risk of hereditary disease, treated by community-based oncologists who elected to participate in this study. Certain risk groups, such as female patients with male-relative related risk factors, had a lower testing rate compared to others, such as patients with a family history of BRCA pathogenic variants.

The overall BRCA testing rate of 94% in this study was higher than, though generally in line with, testing rates found in similar studies conducted since 2013. A 2016 study by Rosenberg et al. of 897 women aged ≤ 40 years at breast cancer diagnosis treated in an academic or community medical center reported a testing rate of 87% by 1 year post-diagnosis [10]. The investigators noted that the frequency of BRCA testing among women diagnosed with breast cancer between 2006 and 2013 increased from 77 to 95% [10]. Similarly, Chen et al. found that the BRCA testing rate increased by 57% in 2013, compared to 11% average annual increases during the preceding 3 years [11]. The current study found that testing rates from 2013 to 2017 remained relatively stable. A recent analysis by Katz et al. conducted in breast cancer patients at an elevated risk of pathogenic variant and treated in a community-based setting, found that approximately half received genetic testing, a lower figure than reported in this study [12]. Unlike this study, the previous analysis by Katz et al. did not include any patients with metastatic breast cancer, and did not report how many women had TNBC. An overall positive test result was observed in 78% of patients who had at least one close blood relative with a known BRCA pathogenic variant, and in 20% of patients with no close blood relatives with a BRCA pathogenic variant. High positive rates of about 40% were also observed in patients with TNBC (who were oversampled) and metastatic and HER2(−) breast cancer. Katz et al. observed variants of uncertain significance or pathogenic variant in 22.7% of tested women. In this study, however, physicians were not asked how they characterized a result of variants of uncertain significance on the CRF. High rates of pathogenic variants among TNBC patients have previously been reported [13, 14].

This study was designed to identify the BRCA testing rate among patients meeting NCCN criteria for testing, thus we would expect to estimate a higher testing rate than for the general population of breast cancer patients. Additionally, the high testing rates observed in this study are applicable to the physicians who elected to contribute data. These physicians had extensive clinical experience, inferred by the fact that they treated an average of 154 breast cancer patients in the previous year, and that they had an average of 16 years of clinical practice. A large proportion also had a genetic counselor within his or her practice. All patients included in this study had documented family history of cancers, which is key to identifying those at a greater risk of hereditary breast cancer. The BRCA testing rate may be lower in patients with unknown, or without documented, family history. Wood et al. found that oncologists participating in the 2011 American Society of Clinical Oncology (ASCO) Quality Oncology Practice Initiative (QOPI), documented a complete family history for less than 40% of the breast cancer patients they treated [8]. It is important to promote education of elements of a complete documented family history, such as family history of cancers with age of diagnosis in first- and second-degree relatives. Lastly, this study required the availability of HER2 receptor status, which may have excluded patients with ductal carcinoma in situ (DCIS) as these patients are frequently not tested for HER2. Conclusions regarding testing rates cannot be applied to these patients.

Several key milestones in 2013 may have impacted BRCA testing rates, including the expanded availability of next-generation sequencing (NGS) technology, the revocation of the Myriad BRCA test patent by the US Supreme Court that resulted in a greater number of laboratories offering the test, and the publication of a celebrity editorial and resulting increased publicity around BRCA testing [15]. Previous studies have estimated a 37–64% relative increase in the rate of BRCA testing immediately following the publication of the celebrity editorial [16, 17]. Furthermore, the Affordable Care Act passed in 2010 gave the US population greater access to healthcare in addition to full coverage of BRCA testing for many individuals who were not previously eligible. This study focused on patients diagnosed with breast cancer during and after 2013, therefore, the high testing rate may be explained by changes in clinical practice and societal attitudes that encouraged BRCA testing among high risk patients.

Among community-based physicians who had extensive clinical experience and the majority of whom had a genetic counselor as part of their practice, the reported rate of BRCA testing was high. These findings are encouraging as targeted breast cancer treatments for patients with BRCA pathogenic variants, such as olaparib, have been approved. Among metastatic and HER2(−) breast cancer patients with a germline BRCA pathogenic variant, olaparib was found to significantly prolong progression-free survival compared with standard of care chemotherapy in a phase III clinical trial [18]. Talazoparib, another targeted treatment, was also recently found to improve progression-free survival over standard chemotherapy among advanced breast cancer patients with a germline BRCA pathogenic variant [19].

Our study demonstrates that a high level of compliance to guidelines in a community setting is possible with a delivery model for genetic counseling and testing. Steps in this direction can help improve treatment for breast cancer patients. Additional training may also be helpful, for example, although not assessed in this study, previous research has indicated that physicians who were trained in clinical genetics in medical school or through continuing medical education (CME) were more likely to have ordered or referred a patient for genetic testing [20]. When physicians maintain high testing rates in accordance with NCCN guidelines, patients identified to have a BRCA pathogenic variant have the option of using a targeted therapy, such as olaparib or talazoparib. In fact, multi-gene panel testing utilizing next generation sequencing has enabled physicians to identify pathogenic variants in genes other than BRCA 1 and 2, and further research is needed to determine the rates of multi-gene panel testing and adherence to NCCN recommendation for additional high-risk genes (though this was not within the scope of the current study). Improvements in BRCA 1 and 2 testing rates are needed for patients with metastatic breast cancer, with risk factors related to male blood relatives, and patients of African American heritage, to ensure that all breast cancer patients can make informed treatment decisions.