Germline Testing Holds Promise But Demands a Cautious Approach

Maurie Markman, MD, editor in chief, is president of Medicine and Science at Cancer Treatment Centers of America and clinical professor of medicine, Drexel University College of Medicine

Maurie Markman, MD

It is difficult to overstate the excitement within the medical and clinical research communities regarding the potential benefits of our rapidly increasing understanding of the role of specific germline variants in health and disease. Much of the past and current focus of such efforts in oncology has been on defining the risk for the development of malignant disease or determining whether an individual patient with cancer might be an appropriate candidate for a specific antineoplastic agent “targeted” to a particular mutation contained within the malignancy.

Yet an equally impressive opportunity may exist in understanding how unique germline variants affect the metabolism (activation or inactivation) or elimination of individual agents or drug classes. Such understanding could lead to a truly revolutionary change in the way cancer is treated, most importantly by increasing the opportunity for clinical benefit while reducing the risks resulting from the anticipated adverse effects of new and existing antineoplastic agents.

Unfortunately, limited data currently support this concept in oncology, and considerable controversy surrounds even the most robust investigational evidence regarding the impact of specific germline variants on clinical outcomes. A case in point is whether to change the dose of tamoxifen in breast cancer management based on how the germline genetic profile relates to the metabolism of the agent.¹

However, increasing experience in other areas of medicine provides strong support for the potential utility of this strategy within the oncology domain. For example, a recent report examined the clinical benefits of genotyping variants of cytochrome P450 2C19 in the selection of medications administered to prevent recurrent thrombotic events following percutaneous coronary intervention and stent placement.² In a randomized trial involving 2488 patients, investigators observed noninferiority for the genotyping experimental arm versus the control. Results showed a statistically significant decrease in the primary toxic event (bleeding) for patients managed with prospective genotyping and drug selection at least partially based on this factor versus standard treatment, or 9.8% versus 12.5%, respectively (HR, 0.78; P = .04).

Although examples of potential clinical benefit of genotyping in oncology can also be cited (eg, prevention of severe genetically associated fluorouracil toxicity³), the community needs to more robustly define the overall utility of such approaches by relative costs (money, time, and effort) and the impact on meaningful and clinically measurable outcomes, such as reduction in serious adverse events, increase in objective response rates, and improvement in time to disease progression.

However, along with recognizing the current limited knowledge base supporting a genotyping strategy, it is relevant to also acknowledge several additional potentially problematic aspects of a patient’s decision to undergo genetic testing, whether in an investigatory setting or during routine clinical care. In my opinion, practitioners will need to carefully consider and successfully confront these issues to ensure patient and societal acceptance of routine germline testing when data on the therapeutic effectiveness, beyond the current indications for risk assessment and direct antineoplastic therapeutics (eg, presence of BRCA mutations), increasingly become available.

The first critical issue relates to ensuring patient privacy and the safety of an individual’s genetic (and other personal) data. Reports have noted the ability of investigators with bioinformatics expertise to track suspects via family DNA without the knowledge of the individuals whose genetic data were used for that purpose.⁴ The implications of undertaking such efforts outside a controlled environment (eg, law enforcement) are quite sobering, and rogue groups with access to sophisticated bioinformatics algorithms could certainly obtain highly personal information (including germline data) for nefarious purposes.

Just as disturbing, a recent report noted that investigators were able to potentially identify individuals with deidentified brain MRI scans simply by reviewing the facial profiles contained within those images.⁵

Although resolution of this concern may be difficult, it is critical that the research, regulatory, and information technology communities work together to truly ensure the safety and privacy of personal data. Without such assurances, many individuals who could benefit from obtaining their genetic profiles will likely be unwilling to participate in either research studies or routine medical practice requiring this highly personal information.

The second concern relates to the willingness and ability of healthcare providers to carefully and clearly explain to patients the meaning of their molecular testing results.⁶ Although clinicians may obtain genotyping results for a very specific purpose (eg, selection of a drug in the management of a particular illness), the implication of the findings may have relevance for multiple other agents in different disease settings and for family members.

Further, practitioners may discover normal uncommon variants for which current data cannot support specific recommendations. However, with additional data involving more patients, the clinical relevance of the specific molecular variant may become clear. Will the healthcare team be able to provide follow-up to individual patients if such information becomes available? Should we be satisfied with a negative answer to this question?

There is no reason that the issues highlighted here cannot be resolved with careful thought and focused attention on the implications of the concern. Yet companies developing products in this arena and individuals hoping to develop generalizable knowledge or obtain grant funding may not consider these matters to be their direct concern. Such beliefs would, in the opinion of this commentator, be most unfortunate and strongly discouraged by appropriate regulatory bodies and society.

References

1. Province MA, Goetz MP, Brauch H, et al; International Tamoxifen Pharmacogenomics Consortium. CYP2D6 genotype and adjuvant tamoxifen: meta-analysis of heterogeneous study populations. Clin Pharmacol Ther. 2014;95(2):216-227. doi: 10.1038/clpt.2013.186.
2. Claassens DMF, Vos GJA, Bergmeijer TO, et al. A genotype-guided strategy for oral P2Y12 inhibitors in primary PCI. N Engl Med. 2019;381(17):1621-1631. doi: 10.1056/NEJMoa1907096.
3. Schwab M, Zanger UM, Marx C, et al; German 5-FU Toxicity Study Group. Role of genetic and nongenetic factors for fluorouracil treatment-related severe toxicity: a prospective clinical trial by the German 5-FU toxicity study group. J Clin Oncol. 2008;26(13):2131-2138. doi: 10.1200/JCO.2006.10.4182.
4. May T. Sociogenetic risks—ancestry DNA testing, third-party identity, and protection of privacy. N Engl J Med. 2018;379(5):410-412. doi: 10.1056/NEJMp1805870.
5. Schwarz CG, Kremers WK, Therneau TM, et al. Identification of anonymous MRI research participants with face-recognition software. N Engl J Med. 2019;381(17):1684-1686. doi: 10.1056/NEJMc1908881.
6. Appelbaum PS, Stiles DF, Chung W. Cases in precision medicine: should you participate in a study involving genomic sequencing of your patients? Ann Intern Med. 2019;171(8):568-572. doi: 10.7326/M19-1414.