Leveraging the Shared Neurobiology of Placebo Effects and Functional Neurological Disorder: A Call for Research

The neuropathophysiology of FND is poorly understood and involves many complex, interweaving predisposing and precipitating factors (5). However, a growing number of neuroimaging studies are providing strong evidence suggestive of brain network dysfunction, most notably involving circuits of emotional regulation and self-agency (8). One brain region that has consistently emerged across FND studies is the amygdala, a critical mediator of stress and fear responses with strong connections across fronto-limbic networks. Reviews detailing how the amygdala and its connected brain regions may contribute to FND pathophysiology are described elsewhere (8), but we briefly highlight two main points. First, amygdala hyperactivity or lack of normal habituation in patients with FND may drive hypothalamo-pituitary-adrenal axis and sympathetic hyperarousal as evidenced by increases in stress markers such as cortisol (9). Second, abnormal functional connectivity has been demonstrated between the amygdala and brain regions involved in FND symptom generation (e.g., motor preparatory regions such as the supplementary motor area) across different FND subpopulations and study paradigms (8). Such findings have been suggested to provide a potential mechanism for how psychosocial stressors may trigger or exacerbate FND symptoms as seen clinically (8). It should be noted that the causal direction of these amygdala abnormalities is uncertain; however, many researchers believe that they may root from adverse childhood experiences during critical periods of brain development (10). Indeed, trauma is a known risk factor for FND, but its specific role in pathogenesis remains a source of debate (10).

Parallel to the growing FND literature, there has been a number of recent neuroimaging and neuropharmacological studies shedding new light on mechanisms of placebo effects. Syntheses of this literature propose a potential placebo network highlighting brain regions such as the dorsolateral prefrontal cortex, anterior cingulate cortex, anterior insula, ventral striatum/nucleus accumbens, temporal-parietal junction, periaqueductal gray matter, and amygdala (1, 11). Indeed, this set of regions overlaps closely with brain areas implicated in models of FND (8, 12). Focusing on the amygdala, the majority of clinical research providing a link between amygdala modulation and placebo effects comes from studies of patients with social anxiety disorder. During a social anxiety disorder treatment study with pre- and post-positron emission tomography neuroimaging, placebo responders and selective serotonin reuptake inhibitor responders showed shared bilateral amygdala deactivations (peaks in the right ventrolateral and left basomedial/basolateral regions), indicative of a common anxiolytic mechanism (13). Moreover, while investigating placebo responses in patients with social anxiety disorder, the G variant of the TPH2 G-703T polymorphism (serotonin-related tryptophan hydroxylase-2 gene) was found to mediate placebo-induced reductions of stress-related amygdala activity and predict clinical placebo response (14). The investigators suggested that this may implicate a modulatory role for serotonin in placebo-induced anxiolysis and represents one example of a growing list of genetic polymorphisms described in the “placebome.” (15) Finally, the relevance of placebo-based amygdala modulation generalizes beyond anxiety disorders and has been demonstrated to be “transferable” across placebo paradigms (16). Indeed, a recent review of placebo effect mechanisms highlighted the amygdala as one of the brain regions showing the most consistent reductions in activity during placebo-induced analgesia (11).

On the basis of the neurobiological considerations described above, a line of reasoning for harnessing placebo effects as treatment for patients with FND is quite straightforward. If amygdala hyperactivity and abnormal connectivity are critically implicated in FND pathophysiology, then leveraging placebo effects to modulate the amygdala and its connected brain network might improve symptoms and potentially modify the disorder. When this mechanistic rationale is combined with numerous anecdotal clinical reports of high placebo responsiveness in FND and limited existing treatment options, this presents a very promising therapeutic opportunity. However, the logistics of how to translate placebo effects to patients with FND in clinical practice are fraught with complicated ethical considerations. The ethical arguments for and against this have been extensively discussed elsewhere (6, 7) and primarily involve principles of informed consent, autonomy, nonmaleficence, beneficence, and justice.

The relevant ethical considerations vary, depending on the approach for delivering placebo effects in FND, which can be broadly divided into three main categories: deceptive use of placebo, “open-label”/honest placebo, and extracting principles of placebo effects to strengthen the therapeutic encounter. Most clinicians and treatment providers would likely consider the latter two options as ethically permissible. In fact, we suggest that applying principles of placebo effects is already firmly embedded in current best-practice models of FND diagnosis delivery and explanation (17). Most notably, this includes the fundamental component of emphasizing reversibility and potential for symptom resolution (17), which promotes strong positive expectations and cognitive framing akin to placebo effects. Fully transparent, “open-label” administration of placebo is conceptually challenging but has shown preliminary promise in chronic pain conditions (18). This stream of placebo research is still in its relative infancy, and dialogue for how to best introduce placebo in this context and proposed mechanisms of action continue to evolve. Finally, arguments on well-intentioned but deceptive or misleading use of placebos, including “pure” (completely inert, e.g., sugar pill or saline injection) or “impure” (biologically active but probably not useful, e.g., subthreshold dosages of medications), are much more controversial. Most portray this approach as prohibitively problematic as a result of violations of autonomy and consent (7); however, others question the lines defining “deceit” and that treatment providers may be able to defend themselves by saying that “this treatment works through various networks in the brain, the particulars of which no one fully understands.’’ (19) As summarized by Rommelfanger (6), the American Medical Association recommends against the deceptive use of placebo on the basis of defining placebo as a substance that has “no specific pharmacologic effect upon the condition being treated.” However, as described previously, placebo effects could legitimately have a “specific” effect for FND and thus raises questions as to whether such recommendations should apply in this situation.

Regardless of one’s opinion on these complex ethical issues, we believe that it would be very beneficial to begin conducting research to investigate these important questions and concerns in an empirical manner. We outline three critical steps: 1) assess feasibility and acceptability by the relevant parties (including qualitative feedback from FND patients, families, treatment providers, etc.); 2) demonstrate target engagement (i.e., can we successfully modulate networks implicated in FND pathophysiology with placebo) and preliminary efficacy; and 3) delineate the durability of placebo effects for FND (i.e., short-term symptom relief versus long-term modification of the disease course) and at what stage(s) of management patients may benefit the most. Finally, this line of research should learn from and leverage existing data on placebo research among patients with related functional syndromes, such as irritable bowel syndrome (20).

There are many important caveats that need to be emphasized alongside this discussion. First, by no means do we believe or suggest that placebo-based treatment will be a quick-fix that replaces existing multidisciplinary care pathways for patients with FND. Instead, it should be considered as an adjunctive and perhaps synergistic addition to current psychoeducation, physical and occupational therapy, and psychotherapy practices. Second is the concern that this line of work could encourage counterproductive viewpoints portraying FND as a “fake” disorder. Such ideology needs to be rectified by the robust evidence demonstrating that placebo effects can improve symptoms in many disorders across medicine, ranging from asthma to angina and from migraine to Parkinson’s disease (2). Do we now interpret these disorders as illegitimate? No, and there is no reason why the same should not hold for FND. We also advocate that this reasoning should discourage attempts to provocatively diagnose FND based solely on positive placebo responses. Large responses may provide a supportive piece of evidence, but the lack of specificity necessitates caution in such interpretations. Third, we must emphasize the paucity of current empirical evidence that placebo effects may be effective as part of FND treatment. As previously discussed, there have been no prospective placebo-focused FND treatment studies, and much of the available information is based on expert opinion, anecdotal examples of high placebo responsiveness in clinical settings, and inferential reports from open-label treatment studies that attribute large and rapidly observed responses (inconsistent with the treatment’s therapeutic mechanism) to potential placebo effects (6). There are also very limited randomized placebo-controlled trials on FND, in which patients randomly assigned to placebo may receive a 50% (not 100%) expectation of receiving active treatment. In one small pilot study of antidepressant medication for psychogenic nonepileptic seizures (21), placebo effects (inert oral pill) were not observed, and more data in this context are clearly needed. Finally, there are many important topics related to this subject matter that have not been covered in this article. Most notably, nocebo effects (negative psychosocial context surrounding a treatment leading to adverse effects) (1), which may provide a model relevant to FND pathogenesis and require further study.

In conclusion, we provide a mechanism-based rationale that supports the potential use of placebo effects for the treatment of FND. This builds on recent reviews of this topic (6, 7), and we are hopeful that this collective push may break the inertia of conducting research in this field. This line of study has many obstacles and will not be easy; however, with the growing evidence substantiating the neuromodulatory capacity of placebo effects, we believe it would be a disservice to our patients if we continue to avoid or ignore this potentially valuable treatment opportunity.