Deep Brain Stimulation for Alzheimer’s Disease: Ethical Challenges for Clinical Research

Overview

Decision-making capacity is fundamental to the principle of respect for patient autonomy and a necessary condition of adequate informed consent whether in regard to a complicated hospital procedure or routine outpatient care [43]. Over the past 70 years, the manner in which informed consent is obtained for research and treatment interventions has changed considerably. Although the Nuremberg Code (1946) established decision-making capacity as a necessary component of voluntary consent for research, it was decades later before there was any explicit concern for the process by which decision-making capacity is evaluated.

In the 1980s, psychiatrists Paul Appelbaum and Thomas Grisso created a standard for evaluating decision-making capacity that has legal precedent and continues to have broad support among the ethics and medical communities [44, 45]. Four categories of patient behavior (consistency of choice, understanding, appreciation, logical reasoning) are considered necessary and sufficient criteria for the establishment of decision-making capacity. Understanding and consistency in choice to participate are particularly relevant for a subject being evaluated for inclusion in a clinical trial of AD. Study designs are complex and participation has lengthy implications as DBS devices often remain implanted in the brains of subjects for years after trial completion. Potential DBS for AD research participants are highly likely to lack decision-making capacity at study entry or to lose decision-making capacity at some point during their participation due to several factors. First, AD is, by definition, a progressive disorder impairing cognitive domains essential to decision-making capacity. One study compared capacity to consent for drug treatment between patients with dementia, depression, or schizophrenia using the MacArthur Competency Assessment Tool for Treatment (MacCAT-T), a validated measure of decision-making capacity [46, 47]. Findings revealed a substantial difference between the groups, with the dementia group showing the most impairment overall. More than 50% of the dementia patients were impaired in understanding and reasoning. Other studies have shown that patients with even mild cognitive impairment may have significantly diminished capacity to consent to treatment [48, 49]. Second, patients with late-stage AD are likely to remain a target group for clinical trials of DBS, especially if alternative interventions have proved ineffective. Such research subjects would almost surely lack decision-making capacity to consent to trial participation. Therefore, DBS for AD trials must include a robust mechanism for both detecting loss of decision-making capacity and protecting the interests of incapacitated subjects.

Recommendations

I. Assessing Decision-Making Capacity to Consent to DBS for AD Clinical Trials

Because the stakes for participation are high, we believe it is incumbent upon investigators to ensure that patients know to what they are consenting and are capable of deciding based on a rational weighing of relevant information as it pertains to their values and preferences. The importance of establishing formal processes to assess potential subjects’ capacity to consent to such research was highlighted in the 2007 consensus statement of the U.S. National Institutes of Health (NIH) and Dana Farber Foundation conference on the Scientific and Ethical Issues involved in the use of DBS for Disorders of Mood, Behavior and Thought [50]. Furthermore, the Alzheimer’s Association consensus recommendations for obtaining informed consent on cognitively impaired adults stipulates that all research participants be evaluated for decision-making capacity using the Appelbaum and Grisso criteria [51]. For clinical trials involving more than minimal risk to subjects, IRBs may consider mandating the use of validated decision-making capacity assessment measures, such as the MacCAT-CR. The MacCAT-CR is a structured interview that incorporates information specific to Appelbaum and Grisso’s four decision-making capacity domains, and generates performance scores for each that have been shown to have high inter-rater reliability[52, 53].

Further, individual hospitals and medical centers should establish clear policies, practices, and training regimens related to assessing capacity. Use of an auxiliary “consenter,” as originally recommended by the 1977 Commission on psychosurgery, would also help to ensure clear and accurate understanding by patients/subjects. An auxiliary consenter would not be someone directly affiliated with the study but would meet with each patient to determine the person’s knowledge about study procedures, risks and benefits, treatment device, and study follow-up. Only after this determination of understanding is obtained would informed consent proceed.

AD subjects with demonstrated decision-making capacity at study entry remain at high risk for losing decision-making capacity at one or multiple time points throughout the trial. DBS for AD IRB protocols must therefore specify how reassessment of decision-making capacity will occur during the study trial and actions taken when enrolled participants are found to lose decision-making capacity prior to study completion [51, 54].

II. Protections for Subjects Deemed to Lack decision-making capacity

Under federal regulations for the protection of research participants, known as the Common Rule, research involving vulnerable subjects require additional safeguards [55]. While these regulations do not specify how IRBs and researchers might implement such safeguards, an evolving scholarship in the area has defined the relevant considerations [51, 56].

The Alzheimer’s Disease Association Consensus Recommendations on obtaining research consent for cognitively impaired adults specify three conditions necessary to justify research on participants who lack decision-making capacity [51].

First, IRBs must decide if the proposed clinical trial offers a reasonable chance of personal health benefits to subjects. If not, then the ethics principle of Nonmaleficence dictates that such studies must involve no more than minimal risk. The Common Rule defines Minimal risk as a research protocol in which “the probability and magnitude of harm or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests” [55]. By this definition, trials involving DBS for AD would involve significantly more than minimal risk. An exception to this general rule is made if “the condition causing cognitive impairment is the object of study, so that the research could not otherwise be carried out without the involvement of individuals unable to consent, and the individual has given a research advance directive that authorizes a proxy to provide consent for enrollment in research of this kind” [51].

Second, out of respect for subject autonomy, clinical trials enrolling individuals who lack decision-making capacity must solicit consent from an appropriate surrogate decision maker who is capable of employing substituted judgment when rendering his or her decision on behalf of the prospective research participant. Substituted judgment is a standard for decision-making on behalf of another that asks the question of the surrogate, “How would he/she have decided in this instance if possessed with sufficient decision-making capacity?” The surrogate is thus expected to be someone in close relationship with the subject and aware of the subject’s relevant values and preferences. However, there are several instances in which surrogate decision making (SDM) is either inadequate or undesirable. First, a significant minority of subjects may not wish to defer a decision to participate in research to a surrogate [57]. Additionally, studies have shown that surrogate decision makers often cannot accurately infer patient preferences [58, 59]. Moreover, in many jurisdictions, SDM is limited to providing research consent for studies involving no more than minimal risk when there is no expected personal health benefit to participation [51, 60]. Lastly, individuals with mild cognitive impairment with intact decision-making capacity but without readily available surrogates should not be universally excluded from prospective participation in AD research [60]. Advance research directives have been proposed as a mechanism for dealing with these challenges by allowing competent individuals to prospectively consent to future participation in research in the event they lose decision-making capacity to consent [54, 60].

Third, respect for persons dictates that researchers must seek participant assent and exclude individuals expressing dissent when absence of decision-making capacity precludes formal consent [51]. The concept of Assent grew out of regulations for research with children and is defined by the Common Rule as “a child’s affirmative agreement to participate in research” [55]. Assent is now also sought for adults with cognitive impairment, and we believe IRBs should define standards for assessing and documenting assent or dissent in studies of DBS for AD.

Overview

Patients with probable AD, desperate for relief and without effective alternatives, are likely to consider interventions such as DBS a last resort for treatment. Such desperation could so alter a person’s perspective that they view the primary goal of the study to be the provision of personal health benefits rather than generalizable knowledge about the efficacy of DBS (primary goals of research). This phenomenon, whereby a patient conflates the goals of research with those of clinical treatment, is termed therapeutic misconception. Left unchecked, it is likely to distort patients’ understanding of the risks and benefits of DBS.

In one widely cited study in which end-stage oncology patients were offered participation in a phase I medication trial, 75% of respondents reported that their chance of benefit was at least 50%, a figure the authors note is “10 times greater than the chance of benefit usually resulting from phase I trials” [61]. For the desperately ill, for whom enrollment in a research trial often represents a last chance for cure, decisions about enrollment are often made on the basis of hope or trust rather than a careful consideration of risks and benefits [61].

Patients with AD and their families may feel a similar desperation when considering DBS. For example, a subject consenting to a sham-control trial of DBS for AD may falsely believe the investigators will choose to place her in a particular treatment arm based on the severity of her illness, ignoring the fact of randomization. The same patient may also strongly believe that restoration of her memory impairment is likely to occur following device implantation. Erroneous assumptions such as these can lead to an underestimation of the risks of participation relative to potential benefits. One study of DBS for treatment-resistant depression involving 31 people found that a majority (64.5%) held erroneous beliefs about the likelihood of personal benefit from the study [62].

Recommendations

Despite the ubiquity of therapeutic misconception in clinical trials and its potential to compromise informed consent, there have been no published studies investigating possible interventions for prevention [63, 64]. Moreover, the lack of a consistent definition of therapeutic misconception hampered attempts at measuring its prevalence in different contexts or in testing interventions to reduce it [65]. In particular, researchers and study participants frequently disagree about the extent to which individual therapeutic benefit may be counted as a study purpose [65]. The development of several new consensus definitions of therapeutic misconception is a promising step forward. Henderson et al. convened a workshop of researchers in medicine, sociology, philosophy, anthropology, law, and bioethics, for the purpose of developing a shared understanding of the phenomenon [65]. The resulting definition reflects the group consensus regarding the fundamental purpose of clinical research while recognizing that therapeutic benefit may accrue to individual participants as an important secondary purpose: “Therapeutic Misconception exists when individuals do not understand that the defining purpose of clinical research is to produce generalizable knowledge, regardless of whether the subjects enrolled in the trial may potentially benefit from the intervention under study or from other aspects of the clinical trial” [65]. More recently, Appelbaum and colleagues developed a 10-item therapeutic misconception scale by surveying 220 research subjects across four academic medical centers in the United States. Beliefs associated with therapeutic misconception were assessed via a 28-item Likert questionnaire that was validated against a semi-structured interview, and data was subjected to factor analysis [66]. The final scale includes statements that suggest mistaken beliefs in one or more of three dimensions: “the degree of individualization of treatment, likely benefit from a study, and the overall purpose of the study” [66]. With a consistent definition of therapeutic misconception, studies examining the effectiveness of interventions should proceed. During the consent process, providing information about study methods, risks, and the use of placebo or sham treatments may not be enough to undermine the belief in direct personal benefit. Rather, it may be necessary to directly inform patients during study consent that “scientific goals will have priority over therapeutic goals” [67]. It has also been suggested that research subjects should understand both how and why clinical research and treatment differ [63]. Another intervention that may prove efficacious in reducing therapeutic misconception is the provision of a “cooling off” period, in which subjects are given adequate time to evaluate their options after learning all the relevant information [61, 68].

Overview

At the conclusion of a clinical research trial investigating the use of DBS for AD, should subjects who clearly benefit from the intervention be provided continued treatment? If so, who is responsible for payment and providing the clinical care necessary for such treatment? These questions lie at the crux of the ethical debate over post-trial access to treatment (hereafter referred to as Post-Trial Access), and are particularly salient in the case of DBS due to its high cost, long duration of treatment, and use of implanted hardware that may remain in a patient’s body for years after the initial research trial ends.

Most of the literature addressing these questions focuses on clinical drug trials, but nevertheless examine reasons for providing or withholding Post-Trial Access that can be applied to the use of DBS. A recent review of the literature found 75 peer-reviewed publications that take a position on the matter of Post-Trial Access for investigational drug trials [69]. Reasons used to justify providing or withholding Post-Trial Access fall into four general categories: (1) moral (justice and roles/relationships between stakeholders), (2) practical, (3) legal, and (4) interests/incentives considerations [69]. Only one study concluded that there is no obligation to provide Post-Trial Access to research subjects [70]. The majority of studies (60%) conclude that there is sometimes an obligation to provide Post-Trial Access, when the strength of compelling reasons to provide access outweighs those reasons to withhold. In the case of DBS studied as an intervention for severe treatment-resistant psychiatric and neurologic illness, we believe the relevant moral and legal reasons, as well as stakeholder interests weigh in favor of providing Post-Trial Access to subjects shown to benefit from the trial intervention. Moral reasons to provide Post-Trial Access acknowledge the last-resort nature of DBS and the lack of effective alternative treatments available to study participants. The medical ethics principle, Non-maleficence, or “do no harm,” is embodied in the Federal Common Rule, and requires clinical research to minimize harm relative to potential study benefits [71]. We believe denying a patient access to the only intervention known to alleviate her suffering is tantamount to violating this sacrosanct principle. Furthermore, as long as device hardware remains implanted in the brains of former study subjects, device maintenance and regular consultation with clinical specialists is required to avoid abandoning patients. Legal guidance, until recently, has been scant or nonexistent. However, in 2000, the World Medical Association revised the landmark Declaration of Helsinki, first published in 1964 and which remains an international source of guidance for the ethical conduct of clinical research. One addition to the 2000 revision is particularly relevant to the provision of Post-Trial Access in DBS: “At the conclusion of the study, every patient entered into the study should be assured of access to the best proven prophylactic, diagnostic, and therapeutic methods identified by the study” [72].

Finally, taking into account the incentives or interests of stakeholders further strengthens the arguments for providing Post-Trial Access to DBS and related treatment for patients shown to derive benefit. The trial sponsor has a primary interest in maximizing profit through the successful adoption of the study intervention. Without the trust and participation of patients suffering from the psychiatric condition the intervention is designed to treat, research and clinical development of DBS would come to a halt. We believe an important way for sponsors to create trust is by providing Post-Trial Access to former study subjects who show meaningful clinical improvements in health and disease burden with the use of DBS. Former research subjects have an interest in continuing to receive beneficial treatment when no adequate treatment alternatives exist, at a cost that is not prohibitive. Researchers have an interest in generating meaningful data. The provision of Post-Trial Access that allows for the collection of longitudinal data on effectiveness and safety would improve the quality of clinical post-trial DBS data. In 2013, St. Jude Medical, a manufacturer of DBS hardware, discontinued a large study of DBS for depression due to lackluster results. Nevertheless, Helen Mayberg, an Emory University psychiatrist and DBS pioneer, has continued an independent trial of the St. Jude device because her data shows benefit to a subset of participants [73]. She also noted that insurance will cover part of the medical costs related to the device as long as the trial remains open [41]. Lastly, governments have an interest in the health of its citizens and in protecting the interests of those most vulnerable.

Recommendations

Once we accept the position that Post-Trial Access should be provided to former study subjects who benefit from DBS, the challenge is to assign responsibility for treatment. To date, several hundred patients have undergone DBS as part of psychiatric research trials, and for most, post-trial access will be prohibitively expensive without some form of third party payment. In our correspondence with a prominent neurosurgeon conducing DBS for AD research, we learned that the sponsoring medical center often picks up the tab for the costs of post-trial office visits and device programming when a patient’s health insurance refuses coverage. In fact, grant reviewers may condition research funding on a commitment by the applicants’ host institution to provide necessary DBS maintenance after the trial’s conclusion. We believe that no individual stakeholder should be burdened with sole responsibility for the costs of DBS and necessary ancillary treatment (i.e., DBS battery replacement, DBS device technical support and maintenance, neurosurgery consultations, DBS programming, psychopharmacology, psychotherapy, etc.). Patients, together with sponsors, investigators, health care systems, insurance, governments, and non-profit organizations must partner to share responsibility and negotiate continued access arrangements prior to study enrollment. An example of a successful collaboration in this model is the HIV Netherlands, Australia, Thailand Research Collaboration (HIV-NAT) co-payment and sliding scale drug fund program [74]. A partnership between the Thai Red Cross AIDS research Center in Thailand, the National Centre in HIV Epidemiology and Clinical Research in Sydney, and the International Antiviral Therapy Evaluation Centre in Amsterdam, HIV-NAT sponsored clinical trials of anti-retroviral medications in Thailand and provided funding for several years of post-trial access to study medications [74]. Similar creative funding solutions will be necessary to provide Post-Trial Access for subjects in effective DBS trials.