The partial µ-opioid agonist buprenorphine in autism spectrum disorder: a case report

Autism spectrum disorder (ASD) is defined by functional impairments in social interaction, disruptions in social communication, and repetitive stereotyped behaviors [1, 2]. Patients typically experience an impaired ability to interpret and appropriately respond to social cues, as well as difficulties in initiating social contacts. This results in an impaired ability to cope with demands of daily life, and to reach long-term goals [3, 4]. ASD is associated with a plethora of coexisting psychiatric comorbidities, including attention deficit hyperactivity disorder (ADHD) [5, 6]. Although the degree of impairment varies considerably, ASD is generally associated with poor quality of life [7, 8], and increased morbidity and mortality [9]. With a worldwide prevalence of 1–2% [10‐12], this represents a significant contribution to global disease burden.

There are currently no medications with regulatory approval or evidence for efficacy to treat any of the core symptoms of ASD. Two medications, risperidone and aripiprazole, are approved by the US Food and Drug Administration (FDA) for treatment of associated behavioral problems (e.g., aggression, severe temper tantrums) in ASD [13]. Although certain individuals may experience positive effects, these drugs are also associated with adverse psychological and physiological effects and their overall effectiveness has been questioned. In the absence of medications with an ability to target core symptoms of ASD, children and adults with this condition are frequently prescribed multiple psychiatric medications of questionable value [14‐16]. Identifying medications targeting the core symptoms of ASD is therefore a major unmet medical need.

Here, we report a case of unexpected beneficial effects in ASD associated with long-term use of the partial µ-receptor agonist buprenorphine administered at a low dose. We then discuss this case in the context of the literature on the role of endogenous opioid systems for social cognition. With this as a basis, we outline opportunities for future research.

M is a single, 43-year-old Caucasian male who lives partly with his aging parents and partly in an apartment provided by social services. He was first seen by author CS after his mother scheduled an appointment for him. The presenting complaint was that, after having achieved considerable functional improvement over more than a decade, the patient had experienced a functional decline in the months prior to the appointment. M and his mother, who accompanied him for the visit, jointly presented the history on multiple visits with author CS, an experienced consultant specializing in neuropsychiatric developmental disorders.

In addition to obtaining the history and carrying out a chart review, the patient was evaluated using an extensive battery of physical and neuropsychological assessments, based on international recommendations for diagnosing neurodevelopmental disorders. For the assessment battery, see Table 1. In addition to ASD-related assessments, this battery covers a broad range of psychiatric differential diagnoses, evaluates hazardous or harmful use of alcohol or drugs, assesses level of function in a broad range of domains, and assesses the patient’s perceived health-related quality of life.

The patient was referred for a second opinion to author MH, a psychiatry professor and expert on addiction medicine, who reviewed medical records and the results of the initial evaluation. MH then met with the patient and his mother. Based on the totality of information, he determined that off-label prescription with low doses of buprenorphine could be justified, provided the patient with a prescription (2 mg once daily), and referred him back.

The patient encouraged the treatment team to share his experience with others through a scientific publication. To strengthen this reporting, it was agreed that a controlled discontinuation attempt would be carried out, allowing a formal evaluation of M both on and off buprenorphine. A protocol for this was developed and approved by the Swedish Ethics Authority (dnr. 2019-05335), and written informed consent was obtained. Based on the half-life of buprenorphine, M was instructed to discontinue his daily buprenorphine dose of 2 mg for 14 days, to allow for tests and assessments of psychosocial functioning off medication. As a rescue plan, the patient was to resume medication if he experienced unacceptable worsening.

We describe here a patient with high-functioning ASD who experienced marked, stable, and independently confirmed long-term functional improvement with low daily doses (0.8–2.0 mg) of buprenorphine. Buprenorphine is a high-affinity partial agonist at µ-opioid and an antagonist at κ-opioid receptors. At the doses used here, it is estimated that buprenorphine has only a modest, perhaps 20%, central µ-opioid receptor occupancy (RO), and negligible activity at kappa receptors [28]. Thus, the clinical effects we report are likely to be associated with a low level of pharmacological µ-opioid receptor activation.

Endogenous opioid systems, and µ-opioid receptors in particular, have long been hypothesized to play a role in social processes and attachment [29]. Data indicate that a role for µ-opioid signaling in social function is phylogenetically conserved. For instance, µ-opioid receptor activation underpins the rewarding properties of social play in juvenile rats [30], and µ-opioid signaling is necessary for pair-bonding in the socially monogamous prairie voles [31]. In nonhuman primates, functional genetic variation in the gene encoding the µ-receptor moderates distress-related behaviors upon maternal separation of infants [32]. In healthy humans, a single dose of an opioid drug sharpened social preferences [33], increased visual attention to eyes [34], memory for happy faces [35], and ratings of positive social stimuli [36]. For negative social stimuli, acute opioid drug effects dampen attention to [36, 37] and perception of [38] negative facial expressions. Conversely, acute opioid antagonism is reported to reduce behavioral and brain responses to socially rewarding stimuli [34, 39]. However, it is also equally true that opiates in high doses can blunt sociality, resulting in introvert behaviors and anhedonia [40].

Research on the role of opioids for social attachment in particular was pioneered by Jaak Panksepp [41]. By extension from these data, Panksepp then proposed that autism could be caused by excessive brain opioid activity [42]. This has, however, not been supported by data; antagonist treatment has shown modest or no effects on core symptoms in clinical trials of children on the autism spectrum [43]. In fact, mice genetically modified to lack the µ-opioid receptor (rather than having an excessive activation of it) show severe social deficits [44] as well as stereotyped and perseverative behaviors necessary to fulfill ASD criteria [45]. These knockout mice also show a number of comorbid symptoms of ASD such as aggression, anxiety, and motor clumsiness, and are currently used as a mouse model of autism [46].

An attempt to reconcile these seemingly contradictory observations is offered by the µ-opioid receptor balance model [46], which posits that both excessive and deficient μ-receptor activity may negatively influence social behavior through an inverted-U relationship (Fig. 1). Accordingly, the model predicts that both opioid agonist and opioid antagonist treatment may be able to improve social functioning, depending on an individual’s opioid tone before treatment.

Our observations in the case of M are consistent with this possibility. Since beginning to work on this case, we have been informally approached by colleagues who have made similar observations. Importantly, however, buprenorphine and other opioids are controlled substances that may be abused, diverted, and result in tolerance and addiction [47]. Given the current severe and lethal opiate epidemic [48], prescription of scheduled substances such as opioid should always be made by experienced clinicians and after careful assessment of individual potential risks and benefits.

Clearly, our noncontrolled clinical observations do not amount to evidence of efficacy. ASD is a heterogeneous condition, highly prevalent with other psychiatric and neurological disorders that may present with symptoms ranging from discrete social challenges to severe and global functional impairments [2]. Clearly, a case observation as described above does not allow for generalizability or treatment recommendations, in neither adult nor pediatric populations. However, our results do allow us to generate the hypothesis that low-dose buprenorphine may be a future therapeutic option in specific cases of ASD. Given the extensive unmet medical needs in this condition, putting this hypothesis to test through a randomized controlled trial appears warranted.