Introduction

The identification of a germline BRCA1/2 mutation in a breast cancer patient is associated with potential benefits for herself as well as for her family members [1]. BRCA status could influence treatment decisions regarding local therapy and systemic therapy, and an oophorectomy performed after a diagnosis of breast cancer improves both breast cancer-specific survival and overall survival for mutation carriers [2,3,4,5]. Currently used selection criteria to merit genetic testing fail to detect up to half of the mutation carriers, and new methods for triaging patients to testing are needed [6, 7].

Over the last years, the rapid technical evolution and decreasing costs for genetic analyses have enabled testing on a much larger scale compared with what was previously feasible. Among other obstacles, the availability of genetic counselors is however a problem [7, 8]. Therefore, if genetic testing should be expanded to a larger number of individuals, then the counseling process needs to be simplified.

Recently, two large randomized trials on telephone genetic counseling for women at a high risk of BRCA mutations have been published [9, 10]. Telephone genetic counseling was cost effective and fulfilled the criteria for noninferiority to standard in-person counseling for all psychosocial and decision-making outcomes. The uptake of genetic testing was somewhat lower in the telephone arms: 28% versus 37% [9] and 84% versus 90% [10], respectively. A further simplification of the testing process is to offer written pre-test information instead of in-person or telephone counseling. No randomized trials on written pre-test information have been carried out in cohorts of unselected breast cancer patients, but a nonrandomized trial has suggested that it could be a feasible alternative to the standard procedure [11].

Taken together, currently available evidence suggests that the standard procedure of in-person genetic counseling could be modified in a cost-effective way without any perceivable negative impact on decisional conflict or psychosocial outcomes. The uptake rate is probably somewhat lower with written information or telephone counseling, but that drawback is outnumbered by the much larger number of patients that could be offered testing and by an increased number of identified mutation carriers.

We undertook a prospective, single-arm study of comprehensive BRCA1/2 mutation screening in a consecutive series of unselected patients with newly diagnosed breast cancer (BRCAsearch, ClinicalTrials.gov Identifier: NCT02557776). Pre-test information was provided by a standardized invitation letter, but the patients could contact a genetic counselor for telephone genetic counseling if they felt a need for that. Here, we report the uptake of genetic testing, the prevalence of BRCA1/2 mutations, the proportion of the mutation carriers that did not fulfill current criteria for BRCA testing, and how many of the patients that contacted us for questions, as well as some biological characteristics of the BRCA-associated breast tumors.

Materials and methods

Study cohort

Since late 2010, all patients with a newly diagnosed or strongly suspected invasive breast cancer in south Sweden are offered inclusion in an ongoing study called SCAN-B (Swedish Cancerome Analysis Network—Breast, ClinicalTrials.gov Identifier: NCT02306096) at the time of breast cancer diagnosis pre-surgery, or before start of preoperative medical treatment [12]. For consenting patients, a part of the tumor is sent to a research lab for RNA sequencing. SCAN-B is currently a biobank research study and the study results have no implications for the treatment decisions of individual participants. The main exclusion criteria for SCAN-B are the inability to understand written Swedish and severe psychological problems.

For reasons related to ethical permits and funding, only patients included in SCAN-B were eligible for inclusion in BRCAsearch. During the study period of BRCAsearch, 86% of all new invasive breast cancer cases at the participating hospitals were included in SCAN-B.

Inclusion criteria for BRCAsearch (all) are as follows: (i) the patient is included in the SCAN-B study; (ii) the patient is recently diagnosed with an invasive breast cancer; and (iii) the patient has signed an informed consent form for BRCAsearch.

Exclusion criteria for BRCAsearch (any) are as follows: (i) the patient is unable to understand the written information in Swedish; (ii) the patient is in a psychological state, due to chronic or temporary reasons, where one could suspect that information about the study substantially could be detrimental to the psychological well-being.

Summary of study procedure for BRCAsearch

  1. 1.

    An invitation letter (Appendix 1) was given to the patient by the nurse at the regular visit to the surgeon a week after primary surgery or by the oncologist at the time of information about neoadjuvant chemotherapy. The invitation letter contained information about the study as well as possible implications of genetic testing, an informed consent form, psychosocial questionnaires, contact information to a genetic counselor, and a referral form for a blood sample for DNA extraction. The patient was invited to contact a genetic counselor for pre-test telephone counseling if she felt a need for more information.

  2. 2.

    If consent was given and the blood sample was sent in, BRCA1 and BRCA2 were analyzed (full sequencing; detailed in Appendix 2).

  3. 3.

    For patients who had not returned the consent form, one reminder was sent per mail approximately 2–4 months after the invitation letter.

  4. 4.

    Noncarriers were informed about the test result through a letter. Mutation carriers were telephoned and given a time for an appointment within one week at the Oncogenetic unit, Department of Clinical Genetics, Lund, Sweden, or, for patients who were at the time under adjuvant/neoadjuvant chemotherapy treatment, a visit was scheduled at the Department of Oncology.

According to the study protocol, patients that fulfilled the Swedish BRCA testing criteria should receive the invitation letter for BRCAsearch, but should also be referred for an assessment at the Department of Clinical Genetics, in order to enable testing of genes other than BRCA1 and BRCA2.

BRCAsearch enrolled patients from three hospitals in south Sweden: Helsingborg Hospital (February 2, 2015–August 26, 2016), Kristianstad Hospital (March 2, 2015–August 26, 2016), and Skåne University Hospital, Malmö (November 2, 2015–August 26, 2016). The targeted accrual was > 500 patients consenting to analysis, which was reached by August, 2016. The study flowchart is presented in Fig. 1. Information on BRCA sequence variant has been submitted to the Breast Cancer Information Core (https://research.nhgri.nih.gov/bic/).

Fig. 1
figure 1

Flowchart of patient inclusion for BRCAsearch and genetic analyses

Full size image

Results

Uptake of genetic testing

The invitation letter was given to 818 patients with newly diagnosed breast cancer at three participating hospitals during the period Feb 2, 2015–Aug 26, 2016. The mean age at diagnosis was 63.6 years (median: 65.4 years; range: 26–94 years). Through Jan 31, 2017, five hundred and forty-two (66.2%) of them consented to analysis of BRCA1 and BRCA2 (Fig. 1). Consenting patients were younger at diagnosis than nonconsenting patients (mean: 61.8 vs. 67.1 years).

For nonconsenting patients, the study protocol dictated that a reminder should be sent by mail 2–4 months after the invitation to the study. However, due to practical reasons, the median time from the invitation to the study to the reminder was in fact 138 days (4.5 months). Out of 542 patients who consented to analysis, 459 (84.7%) consented without a reminder and 83 (15.3%) consented following a reminder.

Time from breast cancer diagnosis to test result

The patients received the invitation letter at the first postoperative visit to the surgeon, usually a week after surgery. Patients treated with neoadjuvant systemic therapy instead received the invitation letter at the first visit to the medical oncologist (See “Materials and methods”). The median time from study invitation to study consent was 15 days. The median turnaround time for the mutation analyses at the laboratory was 48 days. The median time from the completion of the analysis until the letter was sent to the patient was 26 days. Consequently, the median time from breast cancer diagnosis to the delivery of the test result was approximately 3 months.

Mutation analysis and prevalence of BRCA mutations

Eleven pathogenic mutations were found (BRCA1, n = 2; BRCA2, n = 9) in 542 tested patients, corresponding to a mutation prevalence of 2.0% (Agresti–Coull 95% confidence interval, CI: 1.1–3.6%). In addition, there were two patients who were assessed for eligibility that were already known BRCA1/2 mutation carriers. Including those two patients in the analysis yielded a mutation prevalence of 2.4% (CI: 1.4–4.1%). Despite violating the study protocol, eight patients were referred for clinical genetic counseling and testing instead of receiving the invitation letter. As of yet, six of them have not been tested for BRCA mutations and two have tested negative.

Characteristics of mutation carriers and BRCA-associated tumors

The mean age at diagnosis of the eleven mutation carriers previously not identified was 59.2 years (BRCA1: 52.1 years; BRCA2: 60.9 years). Four of them had previously been diagnosed with a breast cancer in the contralateral breast; consequently, the mean age at first breast cancer diagnosis was lower (53.1 years; median 49 years). Six out of 11 fulfilled the Swedish BRCA testing criteria [7], and 9 out of 11 fulfilled the NCCN BRCA testing criteria [13]. Tumor characteristics of BRCA-associated tumors are listed in Table 1. Regarding molecular subtypes (based on immunohistochemical surrogates according to modified St Gallen criteria [14]), no luminal A-like tumors were seen.

Table 1 Characteristics of mutation carriers (n = 11)
Full size table

Questions and genetic counseling

Only a small minority of the patients contacted us for questions related to genetic counseling (n = 14) or practical questions (n = 19). Out of 542 tested patients, eleven (2.0%) contacted us for questions related to genetic counseling and nineteen (3.5%) contacted us for practical questions, such as where to draw the blood sample and how long time the analysis would take.

Discussion

We offered genetic testing of BRCA1 and BRCA2 to > 800 patients with newly diagnosed breast cancer and provided written pre-test information instead of in-person pre-test genetic counseling. Approximately 2/3 of the patients consented to testing. Very few contacted us for telephone genetic counseling or practical questions, suggesting that a majority of the patients felt that the written pre-test information was sufficient for them to make a decision on testing.

The prevalence of BRCA mutations among unselected breast cancer patients could be assessed in different ways within a single cohort, yielding slightly different results. Among the 542 patients tested within our study, the prevalence was 2.0%. Including the two patients that were already known mutation carriers at the time of breast cancer diagnosis, the prevalence was 2.4%. Factors increasing the likelihood of finding a mutation—such as younger age and family history of cancer—are probably more common among tested patients than among patients not consenting to analysis. Consequently, the prevalence of BRCA mutations among all breast cancer patients might be somewhat lower than these estimates. Another factor that needs to be taken into account is the uptake of prophylactic mastectomy among known mutation carriers. In the catchment area of the participating hospitals in our study, 102 known living female mutation carriers had undergone prophylactic mastectomy by Feb 2, 2015, when the study started (BRCA1, n = 67; BRCA2, n = 35), potentially preventing incident breast cancer in some mutation carriers during the period of time that the BRCAsearch study was open.

In countries with a high incidence of breast cancer, such as Sweden, the median age at diagnosis is higher than that in countries with a lower incidence of breast cancer. Since the median age of BRCA-associated breast cancer is lower than sporadic breast cancer, the overall prevalence of BRCA mutations among unselected breast cancer patients in high-incidence countries is probably lower than that in low-incidence countries. Very few studies, if any, have been carried out with comprehensive analysis of BRCA1 and BRCA2 in truly unselected breast cancer patients in high-incidence countries without strong founder mutations. Instead, data have been extrapolated from studies on selected cohorts and from studies that have used panel testing of previously identified mutations instead of full-length sequencing of both genes. The 2–3% prevalence of BRCA mutations found in our study is broadly in line with the prevalence estimates of these previous studies carried out in comparable populations [6, 11, 15, 16].

For ovarian cancer patients, the approval of PARP inhibitors for treatment of BRCA-deficient tumors has sparked much activity regarding methods of streamlining the process of genetic counseling, testing, and delivery of test results [17]. Also for breast cancer patients, there is currently an increasing interest in simplified protocols for genetic testing. The first randomized trial of written pre-test information in hereditary breast cancer was published in 2016 [18]. In this relatively small Australian study, 135 women with newly diagnosed breast cancer and a high (> 10%) likelihood of carrying a BRCA mutation were offered treatment-focused genetic testing (TFGT; testing prior to surgery that could inform surgical decisions). The patients were randomized to either standard care or a brief educational pamphlet instead of pre-test genetic counseling. Following testing, all patients received the test result at a face-to-face appointment. The intervention arm was cost effective and noninferior to the standard arm on the primary outcome decisional conflict [18].

No randomized trials on written pre-test information have been carried out in cohorts of unselected breast cancer patients. In a nonrandomized Norwegian study, that is very similar to our study in design, all newly diagnosed breast cancer patients were offered BRCA testing [11]. Written pre-test information was used instead of face-to-face counseling, and the uptake of testing was 45.4%, which is considerably lower than the uptake of testing in our study. The reasons for this difference in uptake are not clear, but could potentially be attributed to differences in the public knowledge of hereditary breast cancer, the wording in the written information, or other aspects of the study procedures. In the Norwegian study, the invitation letter was given to the patients at the time of diagnosis in order to have a result ready at time of primary surgery, a situation where the patient is faced with a lot of potentially overwhelming information regarding the diagnosis and treatment of the cancer. Also, in contrast to the Norwegian study, we sent a reminder to nonconsenting patients. Following the reminder, the uptake in our study increased from 56.1 to 66.2%.

If given a choice, it seems like a majority of patients eligible for testing would opt for a simplified mode of genetic counseling and delivery of test results. In a study from The Netherlands, 233 breast cancer patients referred for genetic counseling could choose between standard care and an intervention called “DNA-direct.” In DNA-direct, pre-test face-to-face genetic counseling was replaced by telephone and written and digital information. Of note, patients with psychological problems or difficulty with Dutch text were excluded. 161 patients (59%) opted for DNA-direct, of whom 90% were satisfied and would choose DNA-direct again (including 6 out of 8 BRCA mutation carriers) [19]. In another study from The Netherlands, counselees undergoing predictive testing of BRCA mutations or Lynch syndrome mutations all received standard pre-test genetic counseling, but were offered the choice to receive the test result either in-person or by a letter. A majority (69%) opted to receive the test result by a letter [20].

Over the last two decades, > 1500 families with pathogenic germline BRCA1/2 mutations have been identified in Sweden. Most of these families have fulfilled clinical criteria for mutation testing. Of the mutations found, ~ 1030 are BRCA1 and ~ 490 are BRCA2 (ratio ~ 2:1). Founder mutations have been identified in Sweden but are not common; families carrying the five most recurrent mutations (all in BRCA1 and found in 45 or more families) account for ~ 23% of the total number of families with a BRCA mutation (Å Borg, personal communication). In our present study, we notice that a majority of the mutations were BRCA2 mutations. The different ratio could potentially be explained by the higher penetrance of ovarian cancer and younger age at breast cancer diagnosis in BRCA1 carriers, making those families more likely to fulfill current BRCA testing criteria.

We also notice that none of the BRCA-associated tumors were of the luminal A-like subtype. Since treatment decisions regarding adjuvant chemotherapy are currently based mainly on molecular subtyping, 9 out of 11 mutation carriers received chemotherapy despite the fact that mutation status was not known at the time of primary treatment decisions. The median time from the date of breast cancer diagnosis to the delivery of the test result in our study was 3 months, meaning that most mutation carriers were informed about the test result while they were undergoing treatment with neoadjuvant or adjuvant chemotherapy. At the outset of the study, the intention was that the first information to the mutation carriers would be conveyed by a clinical geneticist. As the study progressed, it turned out that a more appropriate approach was to have a medical oncologist deliver the test results, since issues regarding additional systemic therapy (adjuvant PARP inhibitor trial), bilateral prophylactic mastectomy instead of postoperative radiotherapy for node-negative patients, and oophorectomy as a part of breast cancer treatment were often addressed already at this point of time.

There are limitations to our study. First, the nonrandomized study design precludes any solid comparisons with standard genetic testing procedures. Second, by only including patients that had previously been included in the SCAN-B biobank research study, 14% of all consecutive breast cancer patients were not eligible for inclusion and were not invited to participate in BRCAsearch. However, the exclusion criteria for SCAN-B and BRCAsearch were very similar, diminishing the importance of this selection bias, but also decreasing the proportion of the patients assessed for eligibility that were not given the invitation letter for BRCAsearch.

In summary, written pre-test information was a feasible way of streamlining the process of genetic testing in newly diagnosed breast cancer patients. Future studies evaluating potential psychosocial impacts of proactive and simplified procedures of genetic testing are needed, preferably with randomized study designs. Also, the penetrance of BRCA1/2 mutations identified within cohorts of unselected patients needs further study.