Personalized Medicine in Oncology: Ethical Implications for the Delivery of Healthcare

Nathalie Egalite; Iris Jaitovich Groisman; Beatrice Godard


Personalized Medicine. 2014;11(7):659-668. 

In This Article

An Ethical Analysis of Personalized Medicine in Oncology

Emphasizing the Particularities of Cancer Care

Ethical analyses of personalized medicine have identified dominant ethical issues, but have underemphasized the clinical context in which cancer care is delivered. As we have argued, beyond the obstacles to the implementation of personalized medicine, there remain ethical issues embedded in clinical decisions. Among these decisions are the choice of therapy, the extent to which genetic testing and NGS should be involved, the ability to provide the drug and its affordability, how to address the inequalities in outcomes and resources between patients and how to provide the best care that is specific to a particular patient given his or her circumstances.

Individuals diagnosed with cancer are in a particularly precarious situation. With patients facing grim outcomes (possibly terminal), 'last chance' therapies can be especially appealing, even though the gains in survival may be negligible. Patients may not be as health literate or cognisant of the aims of cancer therapies or understand the distinctions between expectations for treatment (curing, prolonging life, alleviating symptoms and so on). The desperation felt by cancer patients amounts to a vulnerable situation that may intensify the burden caused by the often excessive costs of the targeted cancer therapies prescribed. A study undertaken by Buckmaster and Gallagher of breast cancer survivors found that participants felt emotional stresses: a perceived loss of control due to their condition and the psychological impact of having experienced cancer.[26] Consequently, we must be careful that the discourse of individual responsibility and patient empowerment that is at times favored by proponents of personalized medicine does not serve to further burden patients.

The same study showed that relatives were similarly impacted by having witnessed cancer. In our view, analyses of personalized medicine have neglected to satisfyingly capture the familial context in which cancer care is delivered. Genetic information is inherently familial in nature; results on cancer susceptibility could have far-reaching health implications for both the individual patient and their family. However, while genetic testing can present ethical dilemmas for the family no matter the illness, the severity of cancer can take a psychosocial toll on families in a much more profound manner.[27] Queries into a patient's family history have long been a fixture of the patient–provider healthcare interaction, and more recently, considerable efforts have been made to provide adequate oncogenetic counseling for relatives. However, to what extent does the healthcare system have the ability to incorporate and promote relatives' autonomous decisions?

Critically Assessing the Scientific Advances Behind Personalized Medicine in Oncology

In light of an evaluation of the benefits of targeted therapies for patients, it is important to undertake a critical assessment of the scientific progress that supports personalized cancer medicine, as well as its implications. Personalized medicine benefits from biomarkers as tools in order to identify subgroups of patients in the quest for tailoring therapy and improving treatment responses.[4] Although pharmacogenomics has been regarded as a key element in achieving this goal, the use of pharmacogenomic information in oncology clinical practice has been so far restricted.[4] Individual variation of drug responses based on a combination of genetic and environmental factors and patient characteristics is hoped to be understood with the proliferation of NGS technologies. Meyer and colleagues propose that only the concurrence of changes identified by genomic, epigenomic, transcriptomic, proteomic and metabolomic integrated analysis through mathematical modeling could provide a valid personalized approach.[4] A comparable type of analysis could also be applied to the prevention of cancer development. Tumors may show high rates of progression and or recurrence.[28,29] Hence, an ideal personalized medicine would be one including quantitative analysis of cancer dormancy,[28] analysis of histological normal tissue "well before the onset of dysplasia" in order to detect early procancer genetic changes concealed in normal tissue[29] and information from early-stage therapies, as they may influence cell subtype proliferation, thereby affecting the long-term clinical outcomes.[30] It could be argued that the use of genomic technologies to identify oncogenetic variants in order to predict drug response is essentially stratification. Debate continues with regards to stratifying patients into subgroups and whether this is the best we can hope for or rather just a stop on the way to individually tailored health assessments.[31] Weighing between the successes (better health outcomes and increased quality of life) and disappointments (high costs and rare market incentives to create drugs), we need to critically decide whether the use of resources is justified.

Emphasizing Fairness in Resource Allocation in the Delivery of Personalized Healthcare

Targeted therapies in oncology will certainly improve health outcomes, but also, they will almost inevitably bring burdens for patients, clinicians and payers. Indeed, the personalization of therapy is especially appropriate, given that the response to nonspecific cancer chemotherapy is suboptimal in most patients. It can achieve cost savings by avoiding the use of treatments that are ineffective in particular patients. Technological advances and the declining costs of genetic analysis now make it possible to screen for thousands of genetic mutations quickly and inexpensively. Nevertheless, although the new targeted cancer therapies undoubtedly offer the best treatment option, these agents can be very expensive, establishing a high bar based on economics for making the decision to treat. In addition, the predictive tests needed for personalized disease management may not be accessible to the majority of the world's population because of cost constraints for the less prosperous and/or cost–effectiveness for the more wealthy.[7] Thus, these limitations maintain the use of similar therapies for related and even different types of tumors. If the development of target-specific medications is accompanied by appropriate regulatory structures promoting the enrichment of clinical trials with the molecular profiling of study participants, the 'reduction' of the drug market triggered by therapeutic specificity could in fact decrease "the costs of [drug] failures at later developmental stages",[7] and the disappointment of unmet needs in oncology therapeutics.

Some authors consider that jurisdictions will need to determine whether they will rely wholly on economic evaluation or on health policy that has economic policy as a key underpinning.[8] Other authors question whether every individual faced with cancer has a just claim to receive treatment with one of more of these targeted cancer therapies at society's expense.[32] However, issues related to regulation provoke ethical concerns. The lack of consistency and national coverage are problematic. These issues limit physicians' access to information regarding appropriate testing, and they may experience challenges in adapting such knowledge within their local environment in the absence of supportive policies and structures, further exacerbating inequalities in access for patients.[31] Results from the studies by Machini et al.[21] and McGowan et al.[22] with proponents of personalized medicine demonstrate that some ethical challenges within cancer genomics are broader challenges of personalized medicine, such as the costs of integrating new technologies into the healthcare system and differential access to healthcare resources for patients and clinicians. While these are recognized ethical challenges, they point to systemic issues that are often ignored or lost in the face of technocratic and technological challenges that garner more attention. According to McGowan et al., the burden of managing the costs of genomic technologies within the healthcare system looms large, and there is little consensus regarding how to efficiently incorporate genomics into healthcare.[22] The authors report that some leaders in clinical oncology are taking active measures in order to address reimbursement challenges, although proposed solutions are nevertheless fraught with practical and moral complexity.

In his 'Just Caring' analysis, Fleck wonders whether everyone facing cancer has a moral right to the most effective cancer care available, no matter the cost and no matter whether a particular individual has the personal ability to pay for that care.[23,33] His answer is that limits are morally and economically necessary for several reasons: first, referring to Callahan's work, because "health care needs are unlimited because new medical interventions create new health care needs";[34] second, the explosion of new medical technologies has driven the problem of escalating healthcare costs; and third, medicine can increasingly do more by way of offering life-prolonging interventions, but these interventions are generally extremely costly and very marginally beneficial. In conclusion, the practical implication of the 'Just Caring' problem is that the need for healthcare rationing and priority setting is inescapable. Rather, Fleck defends the view that "access to needed health care should be thought of as a matter of social justice," again for several reasons: the vast majority of cancer therapies yield "gains in life expectancy for most patients that are measurable in weeks or months. From the perspectives of cost–effectiveness analysis and the fair allocation of limited social resources this outcome is problematic." This would suggest that we ought to find some way as a society to make wiser and fairer choices that personalized medicine could allow. In many countries, medical authorities are forbidden by law from using cost–effectiveness in order to make coverage decisions. This means that uninsured and underinsured patients would not have assured access to the future possible benefits of these combinatorial cancer therapies, all of which leads Fleck to ask: "Should a just and caring society accept such a state of affairs as just enough?"

The prospect that "access to needed health care should be thought of as a matter of social justice" is also advocated by other authors for the same reasons, and more broadly for integrating genomic knowledge into healthcare delivery more equitably. Indeed, integrating genomic knowledge into healthcare delivery is hoped to result in optimal therapy for the individual patient and significant cost savings by administering the right treatment at the right dose at the right time. Many efforts are being made to this end. Therefore, should the use of individual genetic information in the delivery of healthcare be a cause for concern? The risk that the use of such information leads to psychosocial stress or stigmatization or discrimination by third parties has been widely mentioned. By contrast, the risk that individuals' genetic information leads to healthcare inequities (potentially discriminatory) has been less discussed. This is the argument made by McClellan et al. when they ask "whether exclusion based on genetic information could result in inequitable access to health care, where equal medical need does not result in equal access".[24] In fact, many cancer risk prediction models are based on cancer prevalence rates in particular populations. With regards to breast cancer, for instance, concerns over accuracy among African–Americans, Hispanics and Asians have led to questions from the medical community as to whether individuals in these groups receive optimal care when medical decisions are based on risk assessment scores.[35] Another limitation relates to the implementation of these models into medical practice. Countries worldwide have established particular risk thresholds before a patient is eligible for cancer risk reduction measures.[18,23,24] Often the highest value is attributed to the treatment rather than the diagnostic. Although the use of risk assessment scores in order to prioritize patients for targeted cancer therapies is consistent with evidence-based medicine and can be justified in the context of publically funded healthcare systems, many patients would still not have access to the future possible benefits of these targeted therapies. However, these patients would be paying increased taxes in order to support such access for particular individuals or populations, and the scientific advances underlying personalized medicine are the result of enormous public investment. This is also without accounting for people who lack access to quality healthcare. As an example, Burke et al. have reported that a lack of adequate healthcare access has been shown to contribute substantially to differential rates in prostate cancer mortality among African–American men.[36] Similarly, genomic testing in order to assess breast cancer risk can provide little benefit for American–Indian women who experience significant barriers to receiving standard mammography and receive much less intensive follow-up tailored to their genetic risk. As Burke et al. point out, genomic refinements of risk are largely irrelevant for individuals whose life chances are severely constrained because of their social circumstances.[36] McClellan et al. highlight another source of inequality arising from using genetic information in order to determine access to healthcare that is related to how family history is collected, while the accuracy of a family history can significantly affect the validity of the resulting risk assessment.[24] For patients, knowledge, socioeconomic and cultural barriers are factors that have been shown to influence patient accuracy of their family history. Older age, lower education and membership in a minority group have also been shown to be associated with a lower accuracy of the personal and family history of cancer. For healthcare providers, their knowledge of cancer incidence among minority populations is purported to be a barrier to the accurate collection of family histories. Consequently, some populations may be excluded from personalized therapies or sent for superfluous testing. McClellan et al. and Burke et al. highlight the need to equip healthcare providers with tools targeting underserved groups and to develop therapeutic alliances in which clinicians and patients can address these inequities and work together to implement the best healthcare for the patient.[24,36] We also believe that addressing these inequities is an important part of the personalized medicine agenda, and failure to consider them would increase the likelihood that they could be perpetuated by the use of genetic information as part of medical decision-making.