Maurie Markman, MD
In the rapidly evolving arena of molecular medicine, which prominently includes genetic cancer risk assessments and precision cancer medicine, there are several towering giants in the room which, unfortunately, rarely receive the essential attention they deserve. These are the closely related issues of (a) ensuring privacy with the aim of absolute security of all collected genetic data and (b) the process whereby decisions are made regarding the actual use of an individual’s germline and somatic genetic information.
The recent widely discussed news story regarding the arrest of a suspect in the Golden State Killer murder and assault spree that occurred over many years in California was based, at least in part, on information obtained from a DNA sample previously supplied by a relative for an entirely unrelated purpose—unrelated to any request for use by a law enforcement agency.1
This illustrates both areas of concern noted above. Although it is highly doubtful anyone would question the aim of the effort to find and bring to justice a person responsible for a truly horrendous series of crimes, the chief issue is the use of these data, or any DNA information, without the permission or at least the knowledge of that individual.
If it is considered acceptable for a governmental agency to use an individual’s germline genetic data without permission or without informing the person of the intended use, what other purposes might be proposed by the government or a commercial entity that fall far short of the worthy societal goal of attempting to capture a serial killer? To be clear, in this discussion there is no intent to challenge the actions of law enforcement in this process, but solely to use this event to acknowledge the importance of the discussion of the issue of genetic privacy and security.
In addition, there was the recent commentary describing a scientific report that suggested the potential utility of a “polygenic score” in an examination of educational attainment.2,3 As noted in the news story, “Research like this makes many people nervous. Linking social inequalities to DNA—isn’t this eugenics?”2
One can argue that the conduct of responsible research that benefits societal goals and employs DNA data should be encouraged, but meaningfully strong safeguards must be in place to ensure that intentional or unintentional uses of such data that may foster intolerance or divisiveness are neither encouraged nor permitted.
Potential Loss of Public Trust
One major danger in the documented abuse of genetic data through ill-conceived or poorly managed commercial or academic activities is the loss of individual and overall societal support for the potentially enormous value that thoughtful and scientifically valid examination of germline variants within the population can ultimately provide.
Consider, for example, the experience with BRCA
mutation testing in women, and more recently men, with certain diagnosed cancers or a family history of these malignancies. The presence of such a mutation is now widely recognized to substantially increase an individual woman’s risk for the development of ovarian or breast cancer, or both. More recent investigative efforts have revealed the relevance of BRCA
mutations in prostate cancer. What is even more relevant in this situation is not simply knowledge of such risk but rather clinical investigative efforts that have confirmed the highly favorable impact of specific prophylactic interventions (eg, bilateral oophorectomy), which can substantially reduce this risk and influence both cancer-specific and overall mortality.4
In the absence of a willingness by individuals to undergo this testing and agree to be followed for clinical outcomes, the now-documented benefits of this intervention would simply not have been known. In addition, knowledge of the presence of a germline BRCA mutation in an individual patient with ovarian and breast cancer has more recently been shown to suggest the use of an antineoplastic strategy (delivery of a PARP inhibitor) that is particularly beneficial in this clinical setting.5,6
Further, there is growing evidence of the potential clinical utility of knowledge of the presence of specific normal germline variants, which may directly impact cancer outcomes as well as the efficacy or toxicity of certain anticancer therapeutics. Again, if individuals are unwilling to agree to have their genetic profiles examined and be followed for outcomes because of fears of data security breaches, or unauthorized or misuse of the information, it will simply not be possible to learn if such testing is of clinical relevance. And that unfortunate result will surely be a major missed opportunity for future patients with cancer and our entire society.