Effectiveness of Naltrexone for Treating Pathologic Skin Picking Behavior in an Adolescent with Prader–Willi Syndrome

To the Editor:

P athologic Skin Picking (PSP) is a behavioral complication frequently found in several psychiatric disorders. PSP can cause significant tissue damage and may lead to localized infection and septicemia, sometimes necessitating surgical repair and skin grafting (Lyell 1972; Neziroglu et al. 2008; Arnold et al. 1998; Odlaug and Grant 2008). Case reports of serious PSP causing epidural abscess and paralysis (Weintraub et al. 2000) or carotid artery exposure (O'Sullivan et al. 1999), or necessitating bilateral capsulotomy using gamma knife surgery (Kondziolka and Hudak 2008) have been reported in patients with a variety of psychiatric disorders such as body dysmorphic disorder, obsessive-compulsive disorder (OCD), borderline personality disorder, substance abuse disorders, and Prader–Willi syndrome (PWS) (Odlaug and Grant 2008). The estimated prevalence of PSP in PWS is between 78 and 95% (Dykens and Kasari 1997; Symons et al. 1999; Dykens and Shah 2003). This is a higher prevalence of PSP behavior than is found in the general population (1.4–5.4%) or in persons with other psychiatric disorders (Odlaug and Grant 2008).

PWS is a rare genetically determined neurodevelopmental disorder, with a prevalence of 1 in 10,000–15,000 births (Cassidy 1997), associated with abnormal or absent expression of imprinted, paternally expressed genes at the genetic locus 15q11–13. Three genetic subtypes are involved. The two main genetic subtypes are those with a deletion at the locus 15q11–13 involving chromosome 15 of paternal deletion (deletion 15 genetic subtype; ∼70% of cases) and those with maternal chromosome 15 uniparental disomy (disomy 15 genetic subtype; ∼25% of cases). The remaining 5% have a chromosomal rearrangement or a chromosome-15 imprinting center defect (Cassidy and Schwartz 1998). The syndrome is associated with a distinct physical dysmorphism, as well as with specific behavioral and psychopathologic symptoms. Psychiatric symptoms described in PWS include acute cycloid psychosis, obsessive-compulsive behavior, PSP, hyperphagia, and bipolar and pervasive developmental disorders (Descheemaeker et al. 2002). Cycloid psychosis is described as a constellation of affective symptoms and thought disturbances characterized by a subacute onset, fluctuating course, and mood instability. Varied symptoms have been described including agitation, auditory hallucinations, emotional turmoil, anxiety, confusion, mood swings, paranoid ideation, and obsessive ritual (Descheemaeker et al. 2002).

The exact etiology and neurobiologic basis for PSP in PWS is not known. Several models of PSP, not restricted to PWS, have been proposed. Models include 1) PSP as part of obsessive-compulsive behavior, 2) PSP as a learned behavior and, 3) PSP as an addiction behavior involving the reward pathways of the brain. Several authors have conceptualized PSP as a symptom of obsessive-compulsive behavior. For example, the motor driven symptoms and compulsive picking rituals seen in PSP are often seen in other disorders considered to be part of the obsessive-compulsive spectrum such as Tourette's syndrome or OCD (Bienvenu et al. 2000; Stein et al. 2006). It has been noted that compulsive behaviors may arise in association with cerebral damage, as a relatively involuntary behavior with no apparent purposeful function, and associated with dopamine and serotonin neurotransmitter dysfunction, suggesting important roles for these neurotransmitters in the etiology of PSP. Furthermore, interventions effective in treating obsessive-compulsive behaviors (serotonin reuptake inhibitors [SSRIs] and response prevention) may also be effective in the treatment of PSP (King 1993; Hellings and Warnock 1994; Bloch et al. 2001). However, important differences also exist between PSP and OCD. Considering clinical characteristics, OCD is thought to have a relatively even gender distribution whereas PSP has a large preponderance of females. Whereas treatments effective for OCD may be helpful in PSP, SSRIs and response prevention are generally found more effective for OCD than PSP. Finally, unlike OCD, in PSP the act of picking is rarely driven by intrusive or obsessive thoughts or to relieve anxiety symptoms (Odlaug and Grant 2010).

A model based on social learning theory postulates that PSP is a functional and learned (operant) behavior, acquired through an individual's history of interaction with the social and nonsocial environment. Studies have shown that any kind of self-injurious behavior (SIB), including PSP can be maintained by positive social reinforcement (e.g., it may be reinforced or maintained by attention from others, or access to food or leisure material), positive nonsocial reinforcement (e.g., sensory stimuli directly produced by this response), negative social reinforcement (escape from or avoidance of academic, work, and similar tasks; demands of other people; distraction from negative internal affect), and/or negative nonsocial reinforcement (escape from or avoidance of high levels of arousal or discomfort) (Iwata et al. 1994).

An addiction behavior model has also been proposed to explain PSP behaviors. This model posits effects of dopamine-regulated reward pathways and/or opiate-mediated central nervous system (CNS) effects on PSP. Compulsive skin picking may be conceptualized in some individuals as a dysregulation of the dopaminergic reward circuitry. Dopaminergic agonists such as cocaine, stimulants, and methamphetamine increase dopamine in the ventral striatum (Goodman et al. 1990) and individuals who abuse cocaine or methamphetamine often report uncontrollable picking of their skin to the point of tissue damage (Bergstrom 2008). Opioid antagonists such as naltrexone modulate dopamine within the ventral striatum indirectly through GABAergic neurons (Pierce and Kumaresan 2006), and there exists evidence of some benefit in treating PSP in uncontrolled studies with use of these agents (Lienemann and Walker 1989). Opiate-mediated CNS effects may also be important. It has been postulated that self-picking behavior releases endogenous opioids, and elevated plasma metenkephlin levels have been observed in patients who self mutilate (Coid et al. 1983). There are several case reports of attenuation of these symptoms with opioid antagonists such as naloxone (Richardson and Zaleski 1983; Sandyk 1985). There is also a case report of skin picking behavior in a child with PWS that was successfully treated with naltrexone, an opioid anatagonist with a long half life (Benjamin and Buot-Smith 1993).

The purpose of our letter is to describe the case of an adolescent boy with PWS and severe skin-picking behaviors successfully treated with naltrexone added to ongoing risperidone.

Case Report

AB is a 15-year-old Caucasian male with PWS and mild mental retardation (full scale intelligence quotient [IQ]=65), living at home with his parents, who entered treatment for affective symptoms and thought disturbance suggestive of cycloid psychosis. The patient would become physically and verbally aggressive toward his family and treatment staff, seemingly lose his sense of reality, become internally preoccupied, and demonstrate suspicious ideas of reference, accusing staff and family of taking his comics and toys away. During these episodes he appeared confused, talking about fictional characters in his comic books as if they were real. These episodes usually lasted 30 minutes to a few hours or more, but rarely >1–2 days. Symptoms suggestive of cycloid psychosis were treated with risperidone, gradually titrated to a total dose of 5 mg/day over 6 months, with resultant improvement in psychotic, obsessive, and mood symptoms.

Despite clinical improvement in some symptoms on risperidone, severe skin-picking behaviors continued, largely involving his nose and chin, but also his fingers and forehead, leading to frequent skin infections often necessitating treatment with topical antibiotics such as Neosporin (combination of polymyxin B, neomycin, Bacitracin). Sertraline was added to his treatment regimen and gradually titrated over 2 months to a total dose of 200 mg/day, in an attempt to treat PSP. No improvement was noted after a 6 month trial and sertraline was gradually tapered away. A behavior modification plan based on contingency reinforcement and a token economy system was instituted with parental approval and support in the home, and bandages were placed on the patient's nose and chin. AB received reinforcement if bandages stayed in place the whole day. He was also given a rubber band to manipulate or snap as an alternative behavior to skin picking if he felt the urge to do so. Although AB seemed to enjoy snapping the rubber band, it was not effective in decreasing his skin-picking behavior. Behavioral therapy continued over 2 years with frequent modifications of reinforcers and consequences without improvement in skin-picking behaviors.

At a follow-up treatment review and planning conference different treatment options were discussed with the patient's guardian and treatment staff, and a clinical trial of naltrexone was agreed upon. The frequency of AB's daily skin picking was the outcome measure to be followed. Naltrexone was initiated at 25 mg/day for 7 days and then increased to 50 mg/day. Within 1 week of reaching naltrexone 50 mg/day the patient's skin picking decreased. Within 4 weeks of starting naltrexone 50 mg/day this behavior had all but disappeared. AB reported no side effects from the naltrexone and tolerated the medication very well. During the naltrexone trial risperidone was continued at the same constant dose of 5 mg/day. After a period of successful naltrexone therapy, treatment was abruptly discontinued when the naltrexone supply was interrupted at his local pharmacy. Within 3 days of stopping naltrexone problematic skin picking returned. After a 5 day gap, AB restarted naltrexone and within 3–4 days of treatment at 50 mg/day, his skin picking again stopped. He has continued to do well over the past 15 months on a combination of naltrexone 50 mg/day and risperidone 5 mg/day.

Discussion

We present a case of an adolescent male with PWS whose severe skin picking resolved on naltrexone 50 mg/day and risperidone 5 mg/day after sertraline and risperidone plus behavioral therapy failed to be effective over 2 years. Effectiveness was demonstrated after 4 weeks on the drug with continued good results over the previous 15 months. During this time, the patient has not reported any side effects or long-term tolerability issues with naltrexone. Our case adds to previous reports suggesting naltrexone may be helpful in treating problematic skin- picking behavior in patients with psychiatric disorders (Benjamin and Buot-Smith 1993). Accidental discontinuation of naltrexone for 5 days caused by lack of supplies at the pharmacy leading to concomitant return of symptoms, and their subsequent improvement with restarting it, increases confidence that naltrexone was effective.

The usual dose of naltrexone is 50–100 mg /day for treatment of alcohol and opioid dependence (Anton 2008). For treatment of PSP, our case suggests that the lower dose end of this dose range is effective. For maladaptive behaviors, it is generally recommended that a trial of behavioral therapy be attempted before the introduction of medication. The common adverse events occurring early in treatment with naltrexone are generally gastrointestinal, including nausea, vomiting, and abdominal pain or discomfort. These side effects, along with headache and fatigue, make up the majority of adverse event reports in clinical trials. Such effects may occur in up to 30% of naltrexone patients (Anton 2008). Naltrexone is contraindicated in patients with liver failure, and in overdose this agent can cause serious hepatocellular injury. Naltrexone is also currently available as a depot preparation for opioid dependence and alcohol dependence in adults. This preparation has shown superior efficacy over oral naltrexone in treatment of these disorders (Lobmaier et al. 2011). Although speculative, it is possible that the depot preparation of naltrexone may prove a useful option for patients with severe skin-picking behavior when compliance with oral medication is an issue.

The effectiveness of naltrexone in our case suggests support for an addiction hypothesis of PSP behaviors. Both behavioral therapies to influence learned behavior and medication that influences serotonin neurotransmission were ineffective, suggesting a lack of support for learning theory or obsessive-compulsive spectrum disorder in the etiology of PSP in our case. Given the mechanism of action of naltrexone as an opiate antagonist with regulatory influence on dopamine-mediated neurotransmission in the striatum, our case suggests that PSP and addictive behaviors may share some common neurobiologic mechanisms. As such, further research on the effects of opiate antagonists in the treatment of PSP is warranted.