Abstract
Introduction
The most commonly used medications in ASD are serotonin reuptake inhibitors (SRIs) (Aman et al. 2005), which are commonly used to target compulsive behaviors, but evidence of their efficacy in children with ASD is at best equivocal in double-blind trials (Gordon et al. 1992; Gordon et al. 1993; McDougle et al. 1996; McDougle et al. 2000; Hollander et al. 2005; King et al. 2008). Single studies support the possible efficacy of other medications for compulsive behavior in ASD. A secondary analysis of the randomized, controlled trial of risperidone in children with autism showed significant improvement in compulsive behavior (McDougle et al. 2005). Unfortunately, the use of risperidone is associated with significant risk of adverse effects (McCracken et al. 2002). One small randomized, double-blind, parallel-group study of divalproex in ASD showed significant benefit of divalproex in contrast to placebo, but divalproex is also associated with multiple adverse effects (Hollander et al. 2006).
Given the significant impairment associated with compulsive behavior in children with ASD, novel treatments are needed. One medication with significant promise is riluzole, which is approved for the treatment of amyotrophic lateral sclerosis and acts to decrease glutamatergic neurotransmission. Riluzole has recently been reported to be helpful for OCD in an open-label trial of adults with treatment-refractory OCD (Coric et al. 2005). A case series showed longer-term benefit in a subset of patients (Pittenger et al. 2008). Recently, Grant and colleagues (2007) reported an open-label trial showing significant response in 4 of 6 children with treatment-refractory OCD. Riluzole is associated with a risk of elevated liver transaminases but has otherwise been associated with few side effects (Grant et al. 2007; Pittenger et al. 2008). I present here a case in which riluzole showed some benefit but a possible drug–drug interaction that limited its use.
Case Report
A 13-year-old girl was referred to the Treatment Resistant Autism Consultation Clinic for compulsive behavior that showed only partial response to prior treatments. Her past diagnoses included autistic disorder, mild intellectual disability (mental retardation), OCD, motor tic disorder, and bipolar disorder not otherwise specified (BD-NOS). Her compulsive behaviors at the time included repeating words, demanding a particular vocal response from her parents, and seeking specific food items repeatedly without clear relief from receiving the items. Her medication regimen at evaluation included divalproex extended release (ER) 750 mg twice daily, risperidone 2.5 mg every evening (qHS), fluoxetine 25 mg daily, melatonin 3 mg qHS, and lorazepam 1 mg as needed. Her mood instability, which followed a seasonal pattern with increased sleep and worsened irritability in the winter and decreased sleep in the summer, had improved markedly on the divalproex in a dose-dependent manner. Recent valproic acid blood levels had ranged from 100 to 120 μg/mL.
Some improvements in compulsive behavior had been noted with increased doses of divalproex, as well as with risperidone and fluoxetine, but she continued to have marked compulsive behaviors that led to multiple episodes of screaming and self-directed aggression each day, occupying more than an hour of tantrums each day and resulting in frequent dismissal from the classroom. Importantly, her compulsive behaviors had shifted from ordering of objects, closing of doors, repetitive flushing, and skin picking before risperidone to food-centered compulsions after risperidone reached a dose above 1 mg. Previous medication trials had included sertraline, paroxetine, citalopram, clonidine, oxcarbazepine, and methylphenidate. The oxcarbazepine and methylphenidate, as well as higher doses of each of the SRIs, had led to worsened irritability. Due to their remote location, the family has minimal access to behavioral treatment approaches and has not been able to access residential treatment due to financial constraints.
After a discussion of risk and benefit with various approaches and verbal consent for off-label treatment, we initiated riluzole 50 mg qHS for 1 week, followed by 50 mg twice daily for 6 weeks, with monitoring of complete blood count and liver function tests every 2 weeks. She showed significant benefit at the 100-mg daily dose, with decreased compulsivity and screaming. We were able to taper and then stop the risperidone after she achieved this dose for 2 weeks. Unfortunately, after being on this riluzole dose for a month, she showed a mild elevation of liver enzymes, with serum glutamic oxaloacetic transaminase (SGOT) and serum glutamic pyruvic transaminase (SGPT) levels in the 40–50s. After 6 weeks on this dose, she became listless, preferring to spend most of her time in bed, and had moderately decreased food and fluid intake. This continued for 2 days, at which point a physical exam and repeat laboratory testing was done, revealing SGOT and SGPT levels over 100 but no increase in bilirubin. The riluzole was stopped. She remained listless over the next week, however, and continued to spend most of her time in bed. Repeated examination continued to show no physical signs of acute illness, and repeat blood work revealed a valproic acid level of 160, although liver transaminases were now under 50. Divalproex ER was held for 1 day, followed by resumption at 500 mg twice daily. A repeated level remained elevated, and her dose was further decreased to 250 mg in the morning and 500 mg at night.
Over the course of the next 3 months, risperidone had to be restarted for worsening tics and compulsivity. After a brief improvement, her food seeking worsened. She was then cross-tapered to aripiprazole with marked worsening over the course of 1 month. She was switched to molindone with significant improvement. Her valproic acid level decreased on a stable dose of divalproex ER 750 mg daily, from 100 μg/mL to 80 μg/mL to 60 μg/mL. Her divalproex ER dose has now been increased to 500 mg in the morning and 750 mg at night.
Discussion
It is difficult to draw conclusions from a single case, but the relationship between riluzole treatment and elevated valproic acid levels and subsequent return of these levels to normal is suggestive. The other possibility would be an interaction between divalproex and risperidone, but this seems unlikely because the decreased valproic acid level continued after switching from risperidone to molindone. The moderate increase in liver transaminases seen during treatment with riluzole would not by itself be expected to yield a general effect on drug metabolism, especially in the absence of elevations in bilirubin.
Riluzole is metabolized in part by the P450 enzyme CYP1A2 (van Kan et al. 2005) and has been shown to inhibit CYP1A2 activity (Sanderink et al. 1997); however, this enzyme plays a minor role in valproate metabolism (Kiang et al. 2006). Valproate has a diverse pattern of metabolism, including other P450 enzymes, such as CYP2CP and CYP2A6. Other medications have been shown to lead to increased valproic acid levels, however, including felbamate, sertraline, isoniazid, and cimetidine (PatsalosPerucca 2003). Other possible mechanisms of interaction could include drug absorption or drug transport (Patsalos and Perucca 2003; Kiang et al. 2006). The mechanism of interaction between riluzole and valproic acid levels in this case is unclear, but clinicians should consider monitoring levels of anticonvulsants and be wary of drug–drug interactions when using riluzole.
Footnotes
Acknowledgments
I would like to thank Drs. Susan McGrew and Dr. Ed Cook for helpful clinical insights in the course of this treatment.
Disclosure
Dr. Veenstra-VanderWeele receives research support from Autism Speaks, the American Academy of Child and Adolescent Psychiatry, the Agency for Health Care Research and Quality, Seaside Therapeutics, and Roche Pharmaceuticals.