Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Dec 1.
Published in final edited form as: J Genet Couns. 2020 Jan 22;29(6):949–959. doi: 10.1002/jgc4.1219

Referencing BRCA in hereditary cancer risk discussions: In search of an anchor in a sea of uncertainty

Margaret Waltz 1, Anya E R Prince 2, Julianne M O’Daniel 3, Ann Katherine M Foreman 3, Bradford C Powell 3, Jonathan S Berg 3
PMCID: PMC7374021  NIHMSID: NIHMS1581322  PMID: 31967382

Abstract

As panel testing and exome sequencing are increasingly incorporated into clinical care, clinicians must grapple with how to communicate the risks and treatment decisions surrounding breast cancer genes beyond BRCA1 and BRCA2. In this paper, we examine clinicians’ practice of employing BRCA1 and BRCA2 to help contextualize less certain genetic information regarding cancer risk and the possible implications of this practice for patients within the context of an exome sequencing study, NCGENES. We audiorecorded return of results appointments for 14 women who participated in NCGENES, previously had breast cancer, and were suspected of having a hereditary cancer predisposition. We also interviewed these patients four weeks later regarding their understanding of their results. We found that BRCA1 and BRCA2 were held as the gold standard, where clinicians compared what is known about BRCA to the limited understanding of other breast cancer related genes. BRCA1 and BRCA2 were used as anchors to shape patients’ understandings of genetic knowledge, risk, and management, illustrating how the information clinicians provide to patients may work as an external anchor. Yet presenting BRCA1 and BRCA2 as a means of scientific reassurance can run the risk of patients conflating knowledge about certainty of risk with degree of risk after receiving a result for a moderate penetrance gene. This can be further complicated by misperceptions of the precision of cancer predictability attributed to these or other described “cancer genes” in public media.

Introduction

The concept of an anchor evokes imagery of security; a boat once adrift on the sea is now safely secured. Cognitive bias research has long identified ‘anchoring’ as a potentially problematic phenomenon where individuals’ perceptions of uncertain information is skewed towards the original anchoring information (Tversky and Kahneman 1974). Genetic sequencing can provide a sea of uncertain information (Skinner et al. 2018; Hunter 2016; Amendola et al. 2015), leaving individuals in need of a grounding anchor to help them understand the information. In this paper, we examine clinicians’ practice of employing BRCA1 and BRCA2 to help contextualize less certain genetic information regarding cancer risk and the possible implications of this practice for patients.

When people think of hereditary breast and ovarian cancer genetics, they often think of pathogenic variants in BRCA1 or BRCA2. Indeed, Angelina Jolie’s op-ed about her medical decisions following a positive genetic test created a bump, the Angelina effect, in discussion of and interest in genetic testing for these two genes (Liede et al., 2018). BRCA1 and BRCA2 are also among the earliest discovered cancer predisposition genes and have a high lifetime attributable risk for breast cancer (Kuchenbaecker et al., 2017). Subsequent research has identified numerous other genes that contribute to hereditary breast cancer risk, though most to a lesser degree (Willoughby et al. 2019).

While there is uncertainty about cancer risks even in the well-studied BRCA1 and BRCA2 genes (Domcheck 2019; Dean and Davidson 2018; Powers et al. 2018; Kuchenbaeker et al. 2017; Dean 2016; Hesse-Biber 2014), less is known regarding the magnitude of risk conveyed by many breast cancer genes other than BRCA1 and BRCA2 and, by extension, the risk-benefit ratio of medical interventions (Afghahi and Kurian, 2017; Graffeo et al., 2016; Robson et al., 2010). As panel testing and exome sequencing are increasingly incorporated into clinical care, less evidenced genes with unclear risk estimates may be included (Lee et al., 2018). Clinicians must therefore grapple with how to communicate the estimated risks and less evidenced management options surrounding many genes beyond BRCA1 and BRCA2.

We use the North Carolina Clinical Genomic Evaluation by NextGen Exome Sequencing (NCGENES), a larger study of patients with suspected hereditary genetic conditions, to examine how genetic counselors and clinical geneticists communicate results for breast cancer genes beyond BRCA1 and BRCA2 to patients. Previous research on anchoring in clinical settings has used surveys and interviews to reveal how patients can come into appointments with preconceived anchors that shape their understandings of clinical information (Nuccio et al. 2015; Senay and Kaphingst 2009; Gates 2004; Shiloh and Saxe 1989). Because of these prior beliefs or experiences to which patients anchor, it can be difficult for clinicians to change patients’ perceptions of risk (Shiloh and Saxe 1989; Wertz et al. 1986). In contexts outside of the clinic, research has shown that anchoring can also be external to individuals, emanating from “a starting point” that is given to them which then shapes their interpretation of information (Tversky and Kahneman 1974:1128; Furnham and Boo 2011; Woodward et al 2009; Brewer et al. 2007; Englich and Mussweiler 2001). Yet this research is often based on vignettes or experiments. This is the first study, to our knowledge, on external anchoring in a genetics clinic setting using audiorecordings of clinical encounters.

Methods

Procedures

This manuscript reports findings that are part of a larger study, NCGENES (Moore et al., 2017). The NCGENES study examined the utility of exome sequencing as a diagnostic tool for a clinically diverse group of 645 patients with suspected genetic conditions (Moore et al., 2017), including 125 adults with suspected hereditary cancer. By joining NCGENES, participants hoped to learn the cause of their illness, information on their family members’ risk, and to advance research or benefit others (Khan et al. 2016).

A subgroup of these patient participants were selected for a qualitative component of the study (Skinner et al., 2016) – the basis of this manuscript. The patients selected for the qualitative substudy were purposefully chosen from all enrolled patients to yield, as far as possible, equal numbers of pathogenic variants, variants of uncertain significance (VUS), and negative results across disease categories (e.g. hereditary cancer, intellectual disability, etc). This paper focuses on the fourteen patients within this substudy who previously had breast or ovarian cancer and were suspected to have hereditary breast cancer predisposition. Their in-person return of exome results appointments with NCGENES research clinicians were audio-recorded, and one of two women on study staff who are PhD level medical anthropologists with qualitative and ethnographic research training sat in during the session for observation. These clinician researchers included three clinical geneticists and five genetic counselors, all certified by their respective boards.

Each patient participant then participated in one semi-structured follow up telephone interview. The interviews were conducted by one of the two previously mentioned medical anthropologists. The interviewers had observed the patients’ return of results appointment but had no prior relationships with the patients. Interviews took place four weeks after results were returned and, using a semi-structured interview guide, asked about patients’ understanding of their results, reactions to their results, and whether they had taken or planned to take any actions based on the results. Further details of these one-on-one interviews are published elsewhere (see Skinner et al. 2018 and Skinner et al. 2016). The interviews were audio-recorded. Field notes were taken by the interviewers after the interviews.

Prior to the return of results appointments, patient participants provided written consent for the observation and audiorecording of the appointments and for the telephone interviews to be scheduled four weeks later. Both the return of results appointments and interviews lasted, on average, one hour. The audiorecordings (and transcripts) were not returned to patients for comment or correction. Patients also did not comment on the findings presented below, but several co-authors are research clinicians from the study. Research activities were conducted between 2012–2016. The Institutional Review Board of UNC approved all procedures. All persons gave their informed consent prior to their inclusion in the study.

Patient Participants

These fourteen patient participants were suspected of having a hereditary cancer predisposition. All had previously been evaluated in the cancer genetics clinic, and all but one had prior genetic testing that included BRCA1 and BRCA2. The one un-tested individual had not had clinical testing due to out-of-pocket test costs at that time. Sociodemographic information of the patients is presented in Table 1. Six patients received negative results, 4 received results for pathogenic variants (BRCA1 [1], PALB2 [1], ATM [1] and BRIP1 [1]), and 4 received results for VUS or variants with uncertain contributions to the phenotype (BRCA1 [1], BARD1 [2], and PALB2 [1]). Participants’ cancer histories are presented in Table 2.

Table 1:

Sociodemographic characteristics of participants (N = 14)

Gender Female 100% (14)
Age Range/mean 33–66 / 53.8 years
Race/Ethnicity Non-Hispanic White 85.7% (12)
African-American 7.1% (1)
Asian 7.1% (1)
Education High school 12 years or less 21.4% (3)
Some college/associate’s degree 21.4% (3)
College degree 42.8% (6)
Advanced/professional degree 14.3% (2)
Open in a new tab

Table 2.

Cancer histories of participants (N = 14)

ID Prior Testinga Prior Resultb,c Cancer Historyd Family Historyd,e NCGENES Variant Interpretationb,c NCGENES Case-Level Interpretationf
1 BRCA1/2, TP53 Negative Breast and fallopian tube cancer in 50s; Dermatofibrosarcoma protuberans in 30s Breast (3)
Brain (1)
Lung (1)		 Negative Negative
2 BRCA1/2 Negative Breast cancer early onset Breast (3), one early onset PALB2 variant pathogenic Positive definitive
3 Multi-gene cancer panel ATM variant pathogenic Breast cancer in late 40s, 2 lung cancers in 60s Breast (2), one bilateral
Melanoma (2)
Lung (1)
NOTE: One relative had both breast cancer and melanoma		 ATM variant pathogenic, previously identified Uncertain contribution
4 BRCA1/2* Negative 3 primary breast cancers, first early onset Breast (3), one bilateral
Ovarian (1)
Prostate (1)
Colorectal (1)
NOTE: One relative had both breast and ovarian cancers, another had both prostate and colorectal cancers		 BRIP1 variant pathogenic Uncertain contribution
5 BRCA1/2 Negative Breast cancer early onset, melanoma, cervical cancer (age unknown) Breast (2), one in a male Negative Negative
6 BRCA1/2* Negative Breast cancer early onset Breast (3)
Colorectal (1)		 BARD1 variant likely pathogenic Uncertain contribution
7 None n/a Breast cancer in late 40s Breast (4), one early onset
Colorectal (2)		 PALB2 VUS and an incidental finding Uncertain variant
8 BRCA1/2 Negative Breast cancer early onset Breast (2) Negative Negative
9 BRCA1/2 Negative Breast cancer early onset Breast (3) Uterine (1) Negative Negative
10 BRCA1/2 Negative Ovarian cancer early onset Ovarian (2) Negative Negative
11 BRCA1/2 Negative 2 breast cancers in 50s and 60s, contralateral breast cancer in 50s Breast (2)
Ovarian (1)
Colorectal (1)
Ashkenazi Jewish ancestry		 Negative Negative
12 BRCA1/2; MLH1, PMS2** BRCA1 variant VUS 2 bilateral triple negative breast cancers, 1 early onset; colon cancer MSI-H in late 40s,
Immunohistochemistry missing for MLH1 and PMS2, no hypermethylation		 Breast (2), one early onset BRCA1 variant likely pathogenic, (reclassification of previously identified variant) Positive probable
13 BRCA1/2* Negative Invasive lobular breast cancer in 50s Breast (5), one bilateral and two early onset
Colorectal (1)		 BARD1 variant likely pathogenic Uncertain contribution
14 BRCA1/2 BRCA1 variant VUS 2 [bilateral] breast cancers, 1 early onset Breast (2), one early onset
Prostate (1)		 BRCA1 variant VUS Uncertain variant
Open in a new tab
a

All prior testing included sequencing and deletion/duplication analysis unless otherwise noted

b

Negative is defined as no clinically significant mutations identified

c

Variant of uncertain significance (VUS)

d

Early onset for breast cancer is under 45 years old. Early onset for ovarian cancer is under 50.

e

Number of first and second degree relatives with cancer

f

Case-level classifications with respect to uncertainty were made according to Strande et al. (2018)

*

Protein truncation testing of BRCA1/2 only

**

Sequencing and deletion/duplication analysis of BRCA1/2 and MLHI, and sequencing only of PMS2

Here we present data from transcribed audio-recordings of 14 return of result sessions with research clinicians and patient participants and 13 semi-structured interviews with patients four weeks after the return of results session. One patient could not be contacted for the post-results session interview.

Data Analysis

Audio-recordings of the return of results appointments and telephone interviews were transcribed verbatim. The transcripts were read in their entirety by two authors (MW and AP) and analyzed using conventional content analysis (Hsieh and Shannon, 2005). After reading through the appointment and interview transcripts, these two authors met and discussed themes that emerged from the data, which included mentions of uncertainty and references to BRCA1 and BRCA2 in clinic discussions and patient interviews. Codes were developed out of these themes (“uncertainty” and “references to BRCA”) and applied to the transcripts. These authors jointly organized the coded material which was presented to and discussed by the research team. The research team identified the different ways in which BRCA1 and BRCA2 were referenced, including to communicate a spectrum of risk and management options as well as reassurances of scientific knowledge and certainty, which we discuss below.

Results

Regardless of the type of result patients received, BRCA1 and BRCA2 were used as the standard of certainty against which the uncertainties of other breast cancer genes (i.e. PALB2, ATM, BRIP1, and BARD1) were gauged, particularly in terms of genetic knowledge, risk, and management. Each return of results appointment included a disclosure of the result, discussion about what the result means for patient risk, and discussion about the implications of results. During each of these sections of the appointment, clinicians used BRCA1 and BRCA2 as anchors to shape patients’ understandings of their results, including: 1) to help convey why test results may not explain the whole risk story; 2) to introduce a continuum of risk that genetic results may show; and 3) as a comparison when discussing medical management recommendations. We discuss each of these uses of anchoring with BRCA1 and BRCA2 below.

Results Are Not the Whole Story

When disclosing results to patient participants, research clinicians often referenced BRCA1 and BRCA2 to help contextualize results for these lesser known and understood genes. For all three patients who received VUS or uncertain results for genes outside of BRCA1 and BRCA2, the uncertainties around their results were explained using these two genes. For instance, when returning an uncertain PALB2 to a patient (Participant 7), the clinician (Clinician 4) said:

All the genes we looked at in detail for you have been associated with increased cancer risk just like BRCA1 and BRCA2 have, but not all of them have been as well studied as BRCA1 and 2… We know a lot about those. When we widen up the more and more genes we look at, sometimes the less we know about them because they’re just more newly discovered. Genetics is not a very old science.

When returning an uncertain BARD1 result to another patient (Participant 6), this clinician (Clinician 4) stated:

And part of the reason that we’re uncertain about exactly how much it influences breast cancer risk is because it was just so recently discovered that we don’t have enough information yet to know like we do for BRCA… So we probably will have some more helpful information about BARD1 in the future.

Here, the clinician countered the uncertainty of the result with reassurance that there will come a time when more will be known about the patient’s risk with BARD1, much like what already exists for BRCA1 and BRCA2. Similarly, when returning an uncertain BARD1 result to a different patient (Participant 13), another clinician (Clinician 2) said:

Now BARD1 is one of the more recently discovered genes in women with breast cancer. The limitation to this information is that we just don’t know a lot about BARD1. There have been relatively few studies unlike BRCA1 and BRCA2…And so we pretty well understand what it means for a person to have a mutation in BRCA1 or BRCA2, but we don’t have that same understanding for BARD1.

Another clinician (Clinician 5) then told this patient that she should “keep in touch” with them because there is always “new information and new recommendations” coming out. This clinician continued, saying, “We may find something two years from now that would be even a better fit because maybe this is just part of the story and not the whole story.” The uncertainty of current genetic knowledge but the potential for more certainty in the future was reiterated by this patient in her interview. When asked in her interview if the clinicians thought the results explained her breast cancer, this patient (Participant 13) said:

They’re not sure… [The clinician] said they had to do more research on that because it’s something they hadn’t seen. So that’s why I say, you know, you don’t know. And until we can find an answer, I won’t really know.

Echoing the clinicians, this patient notes that her results do not currently explain her history of cancer, but alludes to the possibility of a fuller explanation in the future after more research has been done on BARD1.

For two of the three patients who received pathogenic variants results outside of BRCA1 and BRCA2, clinicians used the same device of referencing BRCA1 and BRCA2 to explain the uncertainties of their results, also saying that there have been relatively few studies in these more recently discovered genes in comparison to BRCA1 and BRCA2. One clinician (Clinician 2) stated when returning a PALB2 pathogenic variant to a patient (Participant 2) that:

The thing that’s a little bit challenging about PALB2 compared to BRCA1 and 2 is we don’t have as much information about exactly what degree of cancer susceptibility the PALB2 gene confers… It’s probably less than BRCA1 and 2,… we’re starting to learn more and more about it.

When explaining the uncertainties to another woman (Participant 4) with a BRIP1 pathogenic variant, the same clinician said:

It is as you might expect once you get outside of BRCA1 and 2 which are the genes we know the most about that you were already tested for…A lot of the other genes, there’s less information because they’re less common, and they’re not as often tested for. And so to a certain extent, there’s still some uncertainty about what actually this mutation means, but I think I’m – I’m fairly convinced that it is having something to do with the cancer in the family.

Even for those who receive pathogenic variant results, as in these examples, there is limited genetic knowledge of these lesser known genes. These limitations may prevent the pathogenic variants from being the whole story, like a positive BRCA1 or BRCA2 result might be. For example, it may be unclear for lesser known genes which types of cancers are definitively associated, or what the penetrance is. These knowledge gaps lead to uncertainties regarding the cause of these patients’ histories of cancer even when an apparently pathogenic variant is identified.

For six patients who received negative results, clinicians contextualized their results in terms of the current state of genetic knowledge. As a clinician (Clinician 2) told one patient (Participant 9) who received negative results, although the results were negative, that did not “exclude the possibility of a genetic susceptibility” as more may be known in the future about genetic susceptibilities to cancer. This possibility of receiving a different genetic test result in the future was referenced in other appointments. For instance, another clinician (Clinician 4) told a patient (Participant 11) that negative results are not necessarily “informative” like a positive result could be; however, “The hope is that these new technologies … will eventually give us a fuller understanding of the genetics in a person to be able to give that kind of answer.” Participant 1 was told by a clinician (Clinician 1), “So the fact that we did not find an answer doesn’t mean that there isn’t an answer out there to find, and it doesn’t mean we won’t.” Much like for patients with results for pathogenic and uncertain variants, clinicians noted the limitations of genetic knowledge for patients with negative results. These limitations and uncertainties mean that their results are not conveying the whole story about their cancer histories, but more of the story may be revealed in the future.

A potentially more certain future, a time where there may be less uncertainty about these lesser known genes, was reiterated by three patients who received negative results in their interviews. One woman (Participant 1) who received a negative result focused on what is unknown now and what could be known in the future, saying, “You know, we only know it’s not genetic based on where science is today with it.” Another patient (Participant 8) echoed this sentiment in her interview, saying:

From what we do know we – we couldn’t find a relationship [between my genes and cancer]. So that was pretty much what I got out of it, and you know, maybe in the future as we get more data and as the technology improves as well, there could be a correlation identified. But at this point, there are some limitations, and…there was no relationship found.

While these two participants expressed that scientific or technological limitations could not currently demonstrate a genetic cause for their cancers, another patient (Participant 5) went a step further, saying that her DNA will continue to be monitored in conjunction with scientific advances. She commented, “I do like the idea that they’ll continue to check my genes later as they are finding more stuff.” Some patients with negative results, therefore, also thought that more could be learned in the future.

The Risk Continuum

For patients who received uncertain or pathogenic variant results for genes outside of BRCA1 and BRCA2, clinicians referenced BRCA1 and BRCA2 to help explain patients’ risk of breast cancer given the results. Clinicians often (5 out of 6 patients) situated the patients’ lifetime risk of breast cancer on a sort of continuum ranging from the average population risk (~10–12%) up to the estimated risk for female BRCA1 mutation carriers (~70%) (Kuchenbaeker et al., 2017). For instance, when returning an uncertain PALB2 result to a patient (Participant 7), a clinician (Clinician 4) explained to the patient why the result has been classified as uncertain. The clinician said one reason for the uncertain result is that:

PALB2 is actually considered a moderate risk cancer syndrome. So BRCA1 and 2, that’s a high risk cancer syndrome. It changes risk a lot. PALB2 is estimated to increase a woman’s risk between two and four times to get breast cancer. So if most people have about an eight percent chance of breast cancer, we’re talking a sixteen to thirty-something percent risk, which is higher than the general population.

For another patient (Participant 6) who received an uncertain BARD1 result, a clinician (Clinician 1) said:

When a person has a mutation in BRCA1, they have a really high risk of breast cancer… People with mutations in BARD1 have an elevated risk of breast cancer, but it’s not so dramatic.

This same type of continuum was also relayed to patients who received pathogenic variant results. For instance, when returning a pathogenic PALB2 variant to a patient (Participant 2), a clinician (Clinician 2) used BRCA to try to explain the patient’s risk given the finding, noting:

It’s a known breast cancer susceptibility gene. What’s unknown about it and what makes it a little bit challenging for family members is the degree of cancer risk it conveys. Every woman has a twelve percent lifetime risk for breast cancer. This probably increases that risk three to fourfold, something along those lines. We don’t – we don’t think it quite increases the risk to the same degree as BRCA1 does, for example.

In another appointment, a clinician (Clinician 1) returned a pathogenic ATM variant to a patient (Participant 3). The clinician introduced the ATM gene and said it may explain part of the patient’s family history. The clinician continued, asking the patient if she had heard of BRCA1 and 2. The patient said she has heard of BRCA, and the clinician said:

When somebody has a change in, for example, the BRCA1 gene like Angelina Jolie…that explains their family history. And it causes a high risk of cancer. Changes in the ATM gene, mutations in the ATM gene are associated with an increased risk but not a dramatically increased risk. It takes people’s lifetime risk, say eleven percent, and takes them to twenty-two percent. Something like that. So…it doubles it, but it’s not like some of those other genes like BRCA.

Here, the clinician used BRCA1 and BRCA2 to communicate what this result meant for the patient’s risk, but also used the comparison to BRCA1 and BRCA2 to clarify that this pathogenic variant result did not explain her family history, the whole story. Participant 4 was in a similar scenario, but had a pathogenic BRIP1 variant rather than ATM. She reiterated these points in her interview, saying:

I know [the clinician] said that the BRCA was much more likely [to increase risk], and the BRIP wasn’t as likely [to increase risk]. I do feel like we found part of a reason, you know… But that doesn’t necessarily mean…I still think there’s so many other factors, and nobody really knows I don’t believe… I don’t think there’s anybody [who] can sit down and tell me exactly why I had my cancer.

Before joining the study, she said doctors were unable to tell her the cause of her cancer, but she thought that “a combination of genes and environment” played a role. When asked if her thinking about what caused her cancer changed, she said, “Not at all…I think we have so much more to learn. There’s much more science being done at the moment in these areas.” The interviewer followed up, asking, “So is your thinking that maybe in the future genetics will be able to tell you more?” The patient responded, “Oh, I think – I do think so.” Like other patients, this participant echoed the clinicians and pointed to a more certain future where more may be known about these breast cancer related genes.

Ramifications of the Results

Finally, clinicians addressed what the results meant for patients and their family members, particularly as a catalyst for screening or preventive action. For one patient who received negative results, clinicians again used BRCA1 and BRCA2 to help contextualize the certainty and uncertainty underlying preventive recommendations. She (Participant 9) had a single mastectomy when she had breast cancer, and she was contemplating a prophylactic mastectomy for the other breast. She did not know how this new information should guide her decision making. The clinician (Clinician 2) said:

If we had found a BRCA1 or 2 mutation, and we could tell you ‘This means that your future risk for cancer is high enough that a mastectomy makes sense.’ That could have been useful, but what they [your doctors] need to do now is say ‘You know, based on other risk factors, based on your family history, there can be a guess about your future risk for cancer; [your risk] can be based on those things as opposed to based on a genetic result.’

She reiterated this information in her interview four weeks later when the interviewer asked if she had plans to see a doctor after receiving the results. She was not sure as she did not know how much new information the test result provided about the “two big gene mutations,” meaning BRCA1 and BRCA2, saying:

If I had tested positive for those mutations, an example of how they might change my treatment would be like… I had a mastectomy on my left side where I had breast cancer, but on my right side as far as we know it’s cancer free. So I still have that breast. But if I had the gene mutation, they might be more likely to say, “Prophylactically go ahead and have a mastectomy on the right side.” But because there’s not a genetic explanation that we know about, they wouldn’t necessarily recommend a mastectomy on the right side. That would be sort of totally up to me.

For the other patients who received negative results, BRCA1 and BRCA2 were not referenced in terms of recommendations. Instead, clinicians focused on the importance of not interpreting negative results to mean the patients or their family members are not at increased risk for cancer. For Participant 5, for instance, the clinicians cautioned:

Clinician 1: You don’t wanta treat that [negative result] as “Hey. I’m at no increased risk for cancer.”

Clinician 7: Or someone who you might think – it’s probably your doctor. We know you know.

Clinician 1: Right. That’s the bigger danger. If you gave that to your doctor, your doctor thinks “Oh. She doesn’t need a mammogram. She was negative on this.”

While patients with negative results face the risk of undertreatment, clinicians commented on the lack of clear recommendations and the risks of over-interpretation or over-treatment for four of the six patients who received pathogenic or uncertain results outside of BRCA1 and BRCA2. The problem of a lack of clear recommendations came up when clinicians returned an uncertain BARD1 result to a patient (Participant 13). Clinician 2 said that if the patient’s family members were to be screened:

They may not be able to have as certain a clinical recommendation made. For example, if a young woman in the family were found to have a BRCA1 or BRCA2 mutation, we would make very specific recommendations about their screening and potentially prophylactic surgeries. With a BARD1 mutation in a healthy person or someone that hasn’t yet developed cancer, even if they may in some years from then, our recommendations will be a little bit more uncertain. We would probably recommend that they get increased screening, but it would be unclear whether – whether there would be benefit from a prophylactic surgery for example.

The uncertainty about the benefits of prophylactic surgery and the risk of over-interpretation and treatment becomes clear with the patient with a pathogenic PALB2 variant (Participant 2). Although the clinicians did not think the mutation increased her risk to the same degree as BRCA1 and BRCA2, they said that screening recommendations for her daughters should be treated the same as if it were BRCA1 or BRCA2. The patient extrapolated these screening recommendations to prophylactic surgery and her family’s situation. She was going to “highly encourage” her sister to be tested for the PALB2 variant, and said that if her sister “has the mutation, I hope she does mastectomies” in order to prevent her from having cancer and having to go through chemotherapy. The clinicians attempted to neutralize her response:

Clinician 6: There were times where people removed their appendices because there was a risk of appendicitis, and eventually the pendulum swung back and said ‘Actually you know you don’t need to do this.’ And I don’t want to compare it you know… Sleep on this…

Clinician 2: It’s fairly small numbers, [that] is the problem with our estimates…I think they quoted at about a threefold increased risk is what I’ve seen.

Patient: So threefold.

Clinician 6: I don’t know. [begins to look up reference]

Patient: Like a fifty to seventy-five percent. Is that what you mean?

Clinician 2: Well, probably more like a twenty to thirty percent lifetime risk…Something like that.

The patient’s desire for cancer prevention continued four weeks later during her interview, saying, “My sister is going to be tested for sure… She said if [the result] comes back, and she has it, she said she would agree to a double mastectomy… I was thrilled.” For this patient, the uncertainty regarding the magnitude of cancer risk and the caution of the clinicians did not abate her enthusiasm for her sister to take decisive preventive measures.

In summary, BRCA1 and BRCA2 are often used as the lens through which clinicians explain more recently discovered and less understood genes. Clinicians used BRCA1 and BRCA2 as anchors to introduce a continuum of genetic risk, to help convey why test results may be uncertain and not explain a patient’s whole risk story, and as a comparison when discussing medical management recommendations and possibilities of over-treatment.

Discussion

In this paper, we examined clinicians’ practice of employing BRCA1 and BRCA2 as an anchor to help contextualize less certain genetic information regarding cancer risk and the possible implications of this practice for patients. To communicate the uncertainties of genetic information outside of BRCA, we found that the research clinicians frequently went beyond simple reference of BRCA1 and BRCA2 as a means for explanation and reassurance. Throughout the return of results appointments, they often held these two genes out as the gold standard of what we know and continue to learn about BRCA1 and BRCA2 to then compare this with the uncertainty and limited understanding of risk in the other genes. Analyzing these clinician-patient discussions and clinicians’ use of anchoring, therefore, adds to the body of literature on uncertainty in genomic screening by illustrating how clinicians in return of results appointments express uncertainty, which has not been thoroughly studied (Park et al. 2019). This study also forecasts the need for future research to understand the implications of using BRCA1 and BRCA2 as anchors to express uncertainty, which may have the potential to both benefit or harm the interpretation of results information.

The references to BRCA1 and BRCA2 resonated with patients as evidenced by their reiteration of the uncertainties of their results during their interviews. Thus, in many instances, it appears that the use of BRCA1 and BRCA2 as an anchor benefited patients, helping them to ground their understanding of risk and the uncertainty of their results. These findings add to the existing literature on anchoring. While patients’ use of anchoring is typically considered to originate prior to a clinical encounter (Nuccio et al. 2015; Senay and Kaphingst 2009; Gates 2004), this study illustrates how the information clinicians provide to patients may work as an external anchor. Clinicians provided the starting point of BRCA1 and BRCA2, and built on patients’ previous knowledge of the genes. As noted, all but one of the patients in this study had undergone prior hereditary cancer risk assessment. Most assessments occurred before 2012 when testing beyond BRCA1 and BRCA2 for breast cancer risk was rare. The patients in the study likely had prior discussions of cancer risks associated with BRCA1 and BRCA2 with their clinical genetics team at that time. The employment of BRCA1 and BRCA2 in NCGENES therefore likely built on these prior conversations, and resulted in patients framing their perceptions of their results in comparison to BRCA1 and BRCA2.

While this practice can help patients understand the risk and uncertainty related to their results, this study also illustrates that this practice of anchoring uncertain results to more certain genetic information may not always be useful to patients. Using BRCA1 and BRCA2 as a reference of certainty and promise has the potential to confuse the dialogue between clinicians and patients. Instead of a helpful context, the comparatively high risks associated with BRCA1 and BRCA2 may interfere with the adjustment of patients’ perceived risk after receiving a result for a moderate penetrance gene. This is likely to be pronounced for patients without prior knowledge of the complexities of genetic testing for cancer predisposition. But it may be even more important for those who, like the patients in our study, have been previously counseled about BRCA1- or BRCA2-related risk and are receiving “updated” testing. As shown by the patient with the pathogenic PALB2 variant, patients who previously received genetic testing for BRCA1 and BRCA2 may struggle to grapple with the nuances of the results in comparison. Therefore, it is important for clinicians to be mindful of this as more patients are offered the option for testing for other breast cancer related genes, and for future research to further examine the impact of clinicians’ comparisons to BRCA1 and BRCA2 on patients’ understanding of their results.

Presenting BRCA1 and BRCA2 as a gold standard of scientific assurance can also present the risk of patients conflating knowledge of certainty of risk with degree of risk. Although current research seems to support the notion that the breast cancer risk attributable to more recently discovered cancer genes is significantly less than that predicted for BRCA1 or BRCA2, that may not always be the case in the future. And even with the wealth of data about BRCA1 and BRCA2 in particular, there are still uncertainties. Original estimates assessed a lifetime risk for breast cancer as high as 87%, but subsequent analysis lowered estimates for cumulative risk to around 70% (Kuchenbaeker et al., 2017; Powers et al., 2018). As Domcheck notes, this risk estimate is “not absolute,” leaving women to face uncertainty in decisions over possible risk-reduction, and/or surveillance strategies (Domcheck, 2019; Hesse-Biber 2014; Dean 2016). In addition to medical uncertainties, there are also existential uncertainties related to pathogenic BRCA1 and BRCA2 mutations (Dimillo et al. 2013). Those who receive positive BRCA1 and BRCA2 results work to manage the long-term uncertainties of whether they might get cancer, which can provoke fear and anxiety during frequent screening appointments and worries of what would happen to their families if they did get cancer (Dean 2016; Dean and Davidson 2018). Therefore, it can be problematic to present BRCA1 and BRCA2 as fixtures of certainty as this practice could potentially skew societal perceptions of risk for breast cancer related genes, including BRCA1 and BRCA2, through this anchoring.

Moreover, the mere reputation of risk certainty for BRCA1 and BRCA2 may contribute to over- and mis-treatment of patients, especially for those who received variants of uncertain significance in these genes. Kurian and colleagues found that up to 50 percent of surgeons reported managing a BRCA1 or BRCA2 VUS result as if it were pathogenic (Kurian et al., 2017). It is feasible that the same may be true of pathogenic or uncertain variants in genes associated with more moderate cancer risks. Additionally, patients’ desire for clear, preventive, medical recommendations can be complicated by scientific uncertainty about less evidenced genes, and it can be further complicated by misperceptions of the precision of cancer predictability attributed to these genes in public media, as with Angelina Jolie (Borzekowski et al., 2014) or other described “cancer genes” (Caulfied, 2018; Marcon et al., 2018). Indeed, given the ubiquity of discussion of BRCA in the media and the potential for mixed messages, this misperception may occur regardless of what is communicated by the clinicians (Bylund et al. 2012). Despite this, it is still important for clinicians to consider how their use of BRCA1 and BRCA2 as anchors may impact patient understandings of uncertainty.

Study Limitations

The limitations of this study should be considered, particularly that the sample was mostly white and highly educated which may limit the generalizability of the results. It is also unclear how individuals without experience undergoing prior hereditary cancer risk assessment would have interpreted their results. Additionally, study activities were conducted prior to the advent of large panel testing as a first line test. It is, therefore, likely that additional themes will be identified in future research on the return of results for breast cancer related genes and clinicians’ use of anchoring.

Research Recommendations

These findings indicate a need for further research on clinicians use of anchoring in return of results discussions as well as how the practice of referencing BRCA1 and BRCA2 impacts patients’ perception of certainty in regards to cancer risks, risk magnitude, and management recommendations for non-BRCA breast cancer genes, particularly for those without background knowledge of BRCA1 and BRCA2. Similar phenomena might occur in other areas of genetics with other well-known conditions serving as anchors for explanation of genetic findings with less certain implications. In addition, the current influence of gene “name recognition” from popular press and social media on patients’ perceptions of risk magnitude should be investigated as it informs genetic counseling both in terms of content and in shifting patient perceptions regarding hereditary cancer.

Practice Implications

While the use of BRCA1 and BRCA2 as anchors may confuse the dialogue with patients and their perceptions of risk, these comparisons of lesser known and understood breast cancer related genes to BRCA1 and BRCA2 may be unavoidable in some cases. There are generally few clinical guidelines for risk management of carriers of moderately penetrant hereditary breast cancer genes (NCCN, 2019; Tung et al., 2019). In these cases, situating the risk and, therefore, the screening considerations along the continuum from general population risk to that of BRCA1- or BRCA2-related hereditary breast and ovarian cancer predisposition may be an important means for clinicians to communicate uncertainties. Yet as exome sequencing and multi-gene panel testing are increasingly incorporated into clinical care, more women and men will receive results conferring uncertain elevations in cancer risks, leaving them and their healthcare team with ambiguity about how best to proceed with medical management (Lu et al., 2019; Taylor et al., 2018; Couch et al., 2017; Graffeo et al., 2016). As a result, it is crucial to elicit patients’ baseline perceptions and take-away understandings to reduce misperceptions or confusion among patients. Given the presently inevitable uncertainties of hereditary cancer genetics, it is also imperative for clinicians to clearly convey the risk and outcome uncertainties for patients – even for those who receive positive results for BRCA1 and BRCA2.

Conclusion

The dual images of an anchor as grounding and as potentially skewing perceptions are an apt analogy for how BRCA1 and BRCA2 are used as a reference in hereditary cancer genetics. On the one hand, the well-established genes can be a useful guide to help patients and clinicians contextualize the meaning of their results. On the other, this use of BRCA1 and BRCA2 as a reference may overpromise the possibility of future certainty regarding cancer risk and management and may confuse patients regardless of their baseline knowledge of BRCA1 and BRCA2. Therefore, during return of results appointments, clinicians need to be cognizant of how anchors are being used or provided and what patients may be taking away from the anchors. As exome sequencing and multi-gene panel testing are increasingly incorporated into clinical care, further research into best practices for communicating risk and uncertainty is critical to prevent potential confusion and over- or mis-treatment.

ACKNOWLEDGMENTS

The authors would like to thank Jean Cadigan, Michael Flatt, and Debra Skinner for helpful feedback during the development of this paper. Many thanks also to those who participated in the NCGENES study. Support for this article was funded by the National Institutes of Health (NIH) Grants U01HG006487, 2P50HG004488, and R00HG008819 (for A.E.R. Prince). The views expressed are those of the authors alone, and do not necessarily reflect the views of NIH.

Footnotes

Publisher's Disclaimer: This is the peer reviewed version of the following article: Waltz M, Prince AER, O’Daniel JM, Foreman AKM, Powell BC, Berg JS. Referencing BRCA in hereditary cancer risk discussions: In search of an anchor in a sea of uncertainty. J Genet Couns. 2020;00:1–11., which has been published in final form at https://doi.org/10.1002/jgc4.1219. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions.

Conflict of Interest Statement

Margaret Waltz, Anya E.R. Prince, Julianne O’Daniel, Ann Katherine M. Foreman, Bradford Powell, and Jonathan Berg declare that they have no conflict of interest.

Human Studies and Informed Consent

The Institutional Review Board of UNC approved all procedures. All persons gave their informed consent prior to their inclusion in the study. All procedures performed were in accordance with the ethical standards of the responsible committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Animal Studies

No non-human animal studies were carried out by the authors for this article.

REFERENCES

  1. Afghahi A, Kurian AW (2017). The changing landscape of genetic testing for inherited breast cancer predisposition. Current Treatment Options in Oncology, 18(5),27. doi: 10.1007/s11864-017-0468-y [DOI] [PubMed] [Google Scholar]
  2. Amendola LM, Lautenbach D, Scollon S, Bernhardt B, Biswas S, East K, … & Wynn J (2015). Illustrative case studies in the return of exome and genome sequencing results. Personalized Medicine, 12(3), 283–295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Borzekowski DL, Guan Y, Smith KC, Erby LH, Roter DL (2014). The Angelina effect: immediate reach, grasp, and impact of going public. Genetics in Medicine, 16(7), 516–21. doi: 10.1038/gim.2013.181 [DOI] [PubMed] [Google Scholar]
  4. Brewer NT, Chapman GB, Schwartz JA, & Bergus GR (2007). The influence of irrelevant anchors on the judgments and choices of doctors and patients. Medical Decision Making, 27(2), 203–211. [DOI] [PubMed] [Google Scholar]
  5. Bylund CL, Fisher CL, Brashers D, Edgerson S, Glogowski EA, Boyar SR, … & Kissane D (2012). Sources of uncertainty about daughters’ breast cancer risk that emerge during genetic counseling consultations. Journal of Genetic Counseling, 21(2), 292–304. [DOI] [PubMed] [Google Scholar]
  6. Caulfield T (2018). Spinning the genome: Why science hype matters. Perspectives in Biology and Medicine, 61(4), 560–571. doi: 10.1353/pbm.2018.0065 [DOI] [PubMed] [Google Scholar]
  7. Couch FJ, Shimelis H, Hu C, Hart SN, Polley EC, Na J, … & Feng B (2017). Associations between cancer predisposition testing panel genes and breast cancer. JAMA Oncology, 3(9), 1190–1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dean M & Davidson L (2018). Previvors’ uncertainty management strategies for hereditary breast and ovarian cancer. Health Communication, 33(2), 122–130. doi: 10.1080/10410236.2016.1250187 [DOI] [PubMed] [Google Scholar]
  9. Dean M (2016). “It’s not if I get cancer, it’s when I get cancer”: BRCA-positive patients’(un)certain health experiences regarding hereditary breast and ovarian cancer risk. Social Science & Medicine, 163, 21–27. 10.1016/j.socscimed.2016.06.039 [DOI] [PubMed] [Google Scholar]
  10. DiMillo J, Samson A, Thériault A, Lowry S, Corsini L, Verma S, & Tomiak E (2013). Living with the BRCA genetic mutation: an uncertain conclusion to an unending process. Psychology, Health & Medicine, 18(2), 125–134. [DOI] [PubMed] [Google Scholar]
  11. Domchek SM (2019). Risk-reducing mastectomy in BRCA1 and BRCA2 mutation carriers: A complex discussion. JAMA, 321(1), 27. doi: 10.1001/jama.2018.18942 [DOI] [PubMed] [Google Scholar]
  12. Englich B, & Mussweiler T (2001). Sentencing under uncertainty: Anchoring effects in the courtroom 1. Journal of applied social psychology, 31(7), 1535–1551. [Google Scholar]
  13. Furnham A, & Boo HC (2011). A literature review of the anchoring effect. The Journal of Socio-Economics, 40(1), 35–42. [Google Scholar]
  14. Khan CM, Moore EG, Leos C, & Rini C (2016). Patient hopes for diagnostic genomic sequencing: roles of uncertainty and social status. European Journal of Human Genetics, 24(6), 803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gates EA (2004). Communicating risk in prenatal genetic testing. The Journal of Midwifery & Women’s Health, 49(3), 220–227. [DOI] [PubMed] [Google Scholar]
  16. Graffeo R, Livraghi L, Pagani O, Goldhirsch A, Partridge AH, Garber JE (2016). Time to incorporate germline multigene panel testing into breast and ovarian cancer patient care. Breast Cancer Research and Treatment, 160(3), 393–410. doi: 10.1007/s10549-016-4003-9 [DOI] [PubMed] [Google Scholar]
  17. Hesse-Biber S (2014). The genetic testing experience of BRCA-positive women: Deciding between surveillance and surgery. Qualitative Health Research, 24(6), 773–789. [DOI] [PubMed] [Google Scholar]
  18. Hsieh HF, Shannon SE (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 1277–1288. [DOI] [PubMed] [Google Scholar]
  19. Hunter DJ (2016). Uncertainty in the era of precision medicine. New England Journal of Medicine, 375(8), 711–713. [DOI] [PubMed] [Google Scholar]
  20. Kerr A, Ross E, Jacques G, Cunningham-Burley S (2018). The sociology of cancer: a decade of research. Sociology of Health & Illness, 40(3), 552–576. doi: 10.1111/1467-9566.12662 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kuchenbaecker KB, Hopper JL, Barnes DR, Phillips KA, Mooij TM, Roos-Blom MJ, … Olsson H (2017). Risks of breast, ovarian, and contralateral breast cancer for BRCA1 and BRCA2 mutation carriers. JAMA, 317(23), 2402–2416. doi: 10.1001/jama.2017.7112 [DOI] [PubMed] [Google Scholar]
  22. Kurian AW, Li Y, Hamilton AS, Ward KC, Hawley ST, … Katz SJ (2017). Gaps in incorporating germline genetic testing into treatment decision-making for early-stage breast cancer. Journal of Clinical Oncology, 35(20), 2232–2239. doi: 10.1200/JCO.2016.71.6480 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lee K, Seifert BA, Shimelis H, Ghosh R, Crowley SB, … Couch FJ (2018). Clinical validity assessment of genes frequently tested on hereditary breast and ovarian cancer susceptibility sequencing panels. Genetics in Medicine. doi: 10.1038/s41436-018-0361-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Liede A, Cai M, Crouter TF, Niepel D, Callaghan F, Evans DG (2018). Risk-reducing mastectomy rates in the US: a closer examination of the Angelina Jolie effect. Breast Cancer Research and Treatment, 171(2), 435–442. doi: 10.1007/s10549-018-4824-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lu HM, Li S, Black MH, Lee S, Hoiness R, Wu S, … & Tian Y (2019). Association of breast and ovarian cancers with predisposition genes identified by large-scale sequencing. JAMA Oncology, 5(1), 51–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Marcon AR, Bieber M, Caulfield T (2018). Representing a “revolution”: how the popular press has portrayed personalized medicine. Genetics in Medicine, 20(9), 950–956. doi: 10.1038/gim.2017.217 [DOI] [PubMed] [Google Scholar]
  27. Moore EG, Roche M, Rini C, Corty EW, Gimary Z, … Berg JS (2017). Examining the cascade of participant attrition in a genomic medicine research study: Barriers and facilitators to achieving diversity. Public Health Genomics, 20(6), 332–42. doi: 10.1159/000490519 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology: Breast Cancer Screening and Diagnosis (Version 3.2018) https://www.nccn.org/professionals/physician_gls/pdf/breast-screening.pdf. Accessed February 8, 2019. [DOI] [PubMed]
  29. Nuccio R, Hashmi SS, Mastrobattista J, Noblin SJ, Refuerzo J, Smith JL, & Singletary CN (2015). Influence of anchoring on miscarriage risk perception associated with amniocentesis. Journal of Genetic Counseling, 24(2), 278–284. [DOI] [PubMed] [Google Scholar]
  30. Park J, Zayhowski K, Newson AJ, & Ormond KE (2019). Genetic counselors’ perceptions of uncertainty in pretest counseling for genomic sequencing: A qualitative study. Journal of genetic counseling, 28(2), 292–303. [DOI] [PubMed] [Google Scholar]
  31. Powers B, Pal T, Laronga C (2018). Considerations in testing for inherited breast cancer predisposition in the era of personalized medicine. Surgical Oncology Clinics of North America, 27(1), 1–22. doi: 10.1016/j.soc.2017.08.003 [DOI] [PubMed] [Google Scholar]
  32. Robson ME, Storm CD, Weitzel J, Wollins DS, Offit K (2010). American Society of Clinical Oncology policy statement update: genetic and genomic testing for cancer susceptibility. Journal of Clinical Oncology, 28(5), 893–901. doi: 10.120/JCO.2009.27.0660 [DOI] [PubMed] [Google Scholar]
  33. Senay I, & Kaphingst KA (2009). Anchoring-and-adjustment bias in communication of disease risk. Medical Decision Making, 29(2), 193–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Shiloh S, & Saxe L (1989). Perception of risk in genetic counseling. Psychology and Health, 3(1), 45–61. [Google Scholar]
  35. Skinner D, Raspberry KA, King M (2016). The nuanced negative: meanings of a negative diagnostic result in clinical exome sequencing. Sociology of Health & Illness, 38(8), 1303–17. doi: 10.1111/1467-9566.12460 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Skinner D, Roche MI, Weck KE, Raspberry KA, Foreman AKM, Strande NT, … & Henderson GE (2018). “Possibly positive or certainly uncertain?”: Participants’ responses to uncertain diagnostic results from exome sequencing. Genetics in Medicine, 20(3), 313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Strande NT, Brnich SE, Roman TS, & Berg JS (2018). Navigating the nuances of clinical sequence variant interpretation in Mendelian disease. Genetics in Medicine, 20(9), 918–926. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Taylor A, Brady AF, Frayling IM, Hanson H, Tischkowitz M, Turnbull C, & Side L (2018). Consensus for genes to be included on cancer panel tests offered by UK genetics services: guidelines of the UK Cancer Genetics Group. Journal of Medical Genetics, 55(6), 372–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tung N, Domchek SM, Stadler Z, Nathanson KL, Couch F, Garber JE, … & Robson ME (2016). Counselling framework for moderate-penetrance cancer-susceptibility mutations. Nature Reviews Clinical Oncology, 13(9), 581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tversky A, Kahneman D (1974). Judgment under uncertainty: Heuristics and biases. Science 185(4157), 1124–1131. [DOI] [PubMed] [Google Scholar]
  41. Wertz DC, Sorenson JR, & Heeren TC (1986). Clients’ interpretation of risks provided in genetic counseling. American Journal of Human Genetics, 39(2), 253. [PMC free article] [PubMed] [Google Scholar]
  42. Willoughby A, Andreassen PR, & Toland AE (2019). Genetic testing to guide risk-stratified screens for breast cancer. Journal of Personalized Medicine, 9(1), 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Woodward HE, Taft CT, Gordon RA, & Meis LA (2009). Clinician bias in the diagnosis of posttraumatic stress disorder and borderline personality disorder. Psychological Trauma: Theory, Research, Practice, and Policy, 1(4), 282. [Google Scholar]

RESOURCES