To increase transparency, PeerJ operates a system of 'optional signed reviews and history'. This takes two forms: (1) peer reviewers are encouraged, but not required, to provide their names (if they do so, then their profile page records the articles they have reviewed), and (2) authors are given the option of reproducing their entire peer review history alongside their published article (in which case the complete peer review process is provided, including revisions, rebuttal letters and editor decision letters).
I believe you have adequately addressed the reviewers' suggested minor revisions, either by incorporating clarifications into the manuscript or by explaining why changes were unnecessary or not feasible.
Your manuscript received very positive reviews; however, the reviewers noted minor, yet important, revisions that should be addressed. Please address the following key areas in your revisions:
(1) Clarify the methods used in developing the interview guide, identifying themes, coding, assessing inter-coder reliability, and analyzing the data. Please also provide clarification on the timing of the interviews as they relate to return of the results.
(2) To the extent feasible, incorporate additional data and discussion of the predictor variables the reviewers identified as important (e.g. age, those with children, and those who viewed the counseling-like materials).
(3) Revise your discussion to include the suggested relevant literature, to clarify limitations of your data, and to address the additional aspects the reviewers noted as important in the interpretation or significance of your findings (e.g. how BRCA data specifically are informative or distinguishable from WGS/WES or PGS generally).
There are several tables in this manuscript; at the very least, I think Table 7 can be deleted and summarized in the text.
Methods – Please describe the development of the interview guide – were the questions adapted from other studies (several studies referenced about patients’ responses to BRCA1/2 testing), particularly how were the six options constructed to ascertain emotional response, was the guide piloted for understandability and clarity, was the interviewer allowed to ask follow-up questions not included in the guide, etc.
There is no information about how the authors analyzed the data. I’m assuming the interviews were recorded and transcribed. How did they identify themes, how did they code the data, etc? Were participants’ comprehension/recall of results compared to actual results?
While gender and BRCA-carrier status are the major variables analyzed, can the authors provide more insight about the effect of age – presumably responses were different from individuals in their 20’s compared to those in their 70’s? Also, is it possible to link some of responses? For example, was the woman who misunderstood that she would be getting a definitive answer about carrier status and then her anxiety was increased following genetic counseling session also one of the respondents who indicated being ‘moderately upset, surprise, or shocked.’
Could the authors comment on how long after the receipt of results that the interview was conducted? This could impact the many of the comments by the consumers, as well as comprehension and behaviors.
RESULTS – some of the sections seem a bit redundant; some reorganization could limit data from being repeated and shorten this section overall. For example, in the section “Actions taken or planned in response…” , the sharing of data is mentioned again. The quotes only appear in the latter half of Results section – can it be more balanced throughout, such as including some in the Emotional Response section? Maybe have a section on Familial Implications?
Regarding comment on perceived worth of PGS, I think it should be clarified whether their response was based on the BRCA finding or overall. There certainly may have been other data that would influence this response.
This manuscript presents some interesting data about consumers’ responses to receipt of BRCA1/2 mutations from their purchase of the 23andMe Personal Genome Services. Some further information on their methods and analyses would be helpful as suggested.
The authors identify BRCA mutation+ customers who were at least 18 years of age. How many persons under 18 years of age were identified as mutation+ ? Perhaps this is the subject of another manuscript, but the way in which parents dealt with this would be very interesting.
Before viewing BRCA results, customers are encouraged to read counseling-like materials. Do the investigators know how many persons do so and was this examined as a predictor variable?
One wonders if there was a correlation between those who were unaware that BRCA testing was included in the PGS and those who declined to view, or who were unable to recall, their results.
Were there any instances where the PGS finding turned out to be inaccurate on confirmatory testing?
Among mutation-positive individuals who learned this information for the first time, how many did NOT seek medical advice and what were the reasons they did not do so?
The study would have been stronger had validated measures of emotional distress been used, such as the Impact of Events Scale. However, the questions asked are similar to those in the IES and likely have much face validity.
The greatest and perhaps only significant limitation of this incredible report is the possibility that the persons who declined to respond to the request for interview have fundamentally different experiences than those who agreed to the interviews. The Discussion begins with a statement that is true among those interviewed, but since the investigators cannot know whether those they did not interview had the same characteristics (as they themselves point out later in the Discussion), they should modify their statement here and in the abstract to reflect this important limitation.
It is my understanding that the penetrance of BRCA mutations in the absence of family history is not fully understood. Therefore it may be that mutation positive customers who were given risk information as part of the PGS were given higher risk information than was strictly accurate or risk information of more ambiguous accuracy than was fully acknowledged. This issue does not diminish the importance of the paper, but should be discussed.
The investigators explore the emotional impact of the customers themselves, but not the emotional impact upon the 30 family members who underwent testing secondarily. There is an important difference. While at least some of the customers sought out this testing, their relatives did not. Perhaps this information was not readily attainable, but it is important to acknowledge that persons who find out they are at risk without having chosen proactively to be tested might have more significant negative emotional consequences.
This is an extremely important paper addressing one of the most urgent questions in the field of genetics and genomics: the impact of receiving unexpected highly penetrant genetic risk information. Data about this is desperately needed as society and the medical establishment face the rapid integration of genome sequencing. Moreover, in this study 25 persons who did not know they were carrying a BRCA mutation were given unexpected and potentially life-saving information and 30 additional relatives were screened, with an added discovery of 13 positive cases.
The authors are to be congratulated for exploring an issue that could have resulted in information contrary to their commercial interests. Instead, they have discovered findings that may influence the fundamental nature of how screening occurs in the future.
In their introduction, the authors cite Wolf et al, 2010 where consensus statement was created for dealing with incidental genomic findings in biobanking. They should also address the more relevant publications in which geneticists have been describing approaches to the return of incidental information in clinical sequencing and some of the limits of interpreting that information, ie Berg et al, GIM 13:499, 2011; Green et al, GIMS 14:405, 2012; Berg et al, GIM, 10.1038/gim.2012.112
“Not a single mutation-negative participant interpreted the negative BRCA test results as indicating no or significantly lower than average risk for these cancers” is a very important observation, suggesting that at one level, mutation-negative individuals are not falsely reassured. A related question that remains unanswered is whether someone who has a family history of breast/ovarian cancer might feel that obtaining negative results on the PGS screen of the 3 mutations was sufficient testing. The authors should acknowledge that this question is not answerable with their approach.
The authors do not provide information about the educational level of participants (in demographics Table 1), which would be useful if available.
Is there information on which participants had children? This would be helpful to interpret the results regarding men’s concerns about passing mutations to their daughters. Did these men already have children? What about the women?
The authors do not specify what type of software (if any) was used to code and analyze the data.
The authors also do not specify how the four coders 'reconciled' their coding. Can they provide intercoder reliability scores (if used), or further information on how reconciliation was achieved?
This article makes an important contribution to the literature by providing much needed empirical evidence from individuals who have received 'high impact' genetic information via DTC testing. While the data is encouraging, the small sample size and broad conclusion that population wide screening should be considered require further qualifications. Therefore my main comments are in relation to the discussion and conclusions drawn, and I suggest the authors acknowledge/address the issues below more fully:
1. The evidence does indicate that there did not appear to be detrimental emotional effects for the individuals who received mutation positive results, however it is important to note that many of the carriers (22/32) had family histories, had already been tested, had cancer themselves...etc and therefore were aware of their increased risk prior to testing. It is likely that this pre-existing sense of risk helped prepare them for results and led to less stress or shock. Those individuals who are completely unprepared or unaware of their risk may find results more distressing as indicated by the fact that 3 out of 4 who were ‘moderately upset’ were not expecting the results and reported feeling shock or surprise. If the author’s recommendation is population wide screening, more research is required about the implications for individuals who are completely unprepared.
2. The authors note that 23andMe customers are not representative of the population. More should be said here, in particular regarding the fact that this may be a very unique, proactive, and highly educated group (it would be nice to see information on educational level if available) that does not necessarily reflect the larger population.
3. The authors note that BRCA mutation testing is valuable because it is clinically actionable. This is true, however it is important not to understate the difficult decisions and deliberations for women considering undergoing prophylactic bilateral mastectomy and oophorectomy (especially at a young age). There remain considerable uncertainties about the effectiveness of screening (MRI vs mammography, ovarian screening) especially for women below 25. If population screening is being recommended, it is likely more young women will be tested and these women may find the available options more difficult to cope with emotionally.
4. The authors report the surprising and 'remarkable' number of carriers who reported feeling 'neutral' after receiving their mutation positive results. There does not appear to be enough evidence to determine if individuals were indeed 'neutral' or if these were the individuals who already were known mutation carriers, had strong family histories, or perhaps felt no change to their pre-existing feelings which are unknown.
5. One of the concerns about DTC testing is that it will put an undue burden on healthcare system (as mentioned by the authors in the introduction). This needs to be addressed further, as 60% of individuals who were mutation carriers did seek further medical advice and are advised to do so by 23andMe. The authors suggest that this high number may be related to the actionability of the results, however this may put a strain on the health care system especially if population wide screening is being recommended.
6. Related to the point above, it would be useful for the authors to address the fact that physicians were not necessarily prepared to handle DTC information, gave inappropriate advice, or did not know what to do with the reports. This points to a need for physician education as well.
7. The authors touch upon, but do not go any further, in discussing the presence of these mutations in non-Ashkenazi individuals, and the deficiencies of using self identified Ashkenazi ancestry to determine who should be tested. Do the authors have a suggestion for how to deal with this? Do they suggest testing all individuals in the population, only self-identified Ashkenazi Jewish individuals, Sephardi Jews as well…etc? This also relates to an additional, and unaddressed, point regarding the fact that other populations, such as Latino/as and African Americans may not benefit from this population screening despite the fact that they may also carry particular founder mutations. Is there potential to increase health disparities?
8. The authors suggest that the results of this study will be important especially as whole genome/exome sequencing is increasingly available. This study does provide important data about individual responses to DTC testing for serious genetic conditions, however BRCA testing is significantly different from much of the information that is currently available from whole genome/exome studies where results may be of unknown significance or create a very small and multi-factorial increased risk. BRCA mutations are highly penetrant and transmitted in an autosomal dominant fashion, and management options are available. The authors could make this distinction more explicit.
9.The authors report that males, rather than females, had greater concern about passing on mutations to their daughters. This interpretation appears overly simplistic. There is increasing literature about the impact of mutation status on reproductive decision making for women (Ormondroyd et al, Julian-Reynier et al 2012, Fortuny et al 2009). I think it is slightly misguided to conclude that this is a greater concern for men than women. Rather I think it is the primary or initial concern for men who do not have to consider risk management options for breast/ovarian cancer, whereas women are faced with their own risk management as well as their children’s risks.
10. The fact that men were less likely to seek advice from a physician than women and felt uncomfortable sharing information with relatives speaks to a common and recognized ‘gendered’ burden of health care and sharing of information which often falls to women.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.