My article, featured, and free, in The Carlat Child Psychiatry Report

Getting meds just right is challenging in autism. Sometimes we succeed. For example, a woman with minimal verbal ability is extremely aggressive. She is on a number of medications including valproate 1000 mg extended release, paroxetine 40 mg, ziprasidone 40 mg twice a day, alprazolam 1 mg twice a day, and a host of supplements. She settles down considerably when her compliance-based behavioral intervention is replaced with a developmental approach: The new team listens to her, and they build on her ideas to engage her in a collaborative relationship. The team eases her off multiple medications, leaving her on a low dose of risperidone, and she does well.

It doesn’t always work out so well. In another case, a young boy requires constant supervision, often by more than one person, to keep him from running into streets and climbing onto high banisters. Multiple medications haven’t worked, nor have any other treatment approaches. He has been tried on doses of aripiprazole up to 30 mg with no clear benefit; 1200 mg of oxcarbazepine, which gave him loss of motor tone; and 60 mg fluoxetine, which has only made him more impulsive and active and caused him to care even less about things or people. We keep trying, but he may require a restrictive placement.

Choosing meds: A practical approach
As these vignettes illustrate, no two people with ASD are alike, meaning that our medication choices have to be individualized even more so than in most other psychiatric syndromes. Only two medications, risperidone and aripiprazole, are FDA approved for ASD, and these approvals are limited to one specific set of symptoms—irritability/aggression. We must choose from a range of off-label meds to treat the wide variety of issues.

When I evaluate a patient with autism, here’s my approach.

1. Identifying specific target symptoms 
In my experience, there are 12 categories of symptoms that might respond to a variety of treatments, including medications (see Autism Treatment Planning Chart on p. 3). In real patients, it’s not always easy to distinguish these symptom clusters.

  • I talk to as many informants as possible, including parents, teachers, and therapists, and in collaboration with parents we try to prioritize the symptoms.
  • While I refer patients for a general physical exam, I do vitals and a screening neurological exam. I also generally get a CBC, chemistries, fasting lipids and glucose, a genetic CMA (chromosome microarray assay) screen, look at metabolic and nutritional status, and consult others as indicated. I often get an EKG for neuroleptics, and I consider a 24-hour EEG.
  • Before prescribing anything, I make sure that all of my patients have appointments for occupational therapy assessment, speech and language assessment with attention to nonverbal communication, and developmental testing.
  • I gather as much specific symptom data as possible, such as frequency, severity, and duration of difficult moments, as well as the circumstances surrounding them. I often work together with parents to design an individualized template for tracking symptoms.
  • I have parents keep detailed sleep logs that include bed time, activities before bed, food, medication, and time to sleep and awakenings at night along with the outcomes of those awakenings (eg, child comes to parent’s bed, child easily redirected back to bed, etc).

2. Choosing the right medications: A case-based discussion
The most common candidate medications for ASD are alpha agonists, stimulants, SSRIs, neuroleptics, and antiepileptics. Because there are so many symptoms and so many possible treatments, I often use the Autism Treatment Planning Chart to help guide my decision making.

Here’s how I treated a 10-year-old boy with ASD who presented with symptoms of anxiety, inattention, overactivity, depression, perseverative thinking, sensory sensitivity, and tantrums.

Alpha agonists. I began by prescribing an alpha agonist. These medications are often my first choice, because they reduce the ‘fight-flight’ response by shifting autonomic function to give parasympathetic tone more of an edge over sympathetic tone. They can also help ease anxiety and tics.

I started with guanfacine 0.5 mg at night, and gradually increased it as needed. I try to avoid going beyond 2 mg twice a day and I adjust the dose based on sedation or, rarely, dizziness. I am less impressed by its cousin clonidine, which is far more potent, more sedating, but also short acting, so when used for sleep induction it can lead to rebound wakefulness. On guanfacine, my patient was more able to interact, but otherwise he was still quite symptomatic.

Stimulants. While in the ideal world, ASD patients would not be prescribed multiple medications, in my practice they often need combination treatment to address their symptoms. For this boy we added amphetamine mixed salts, extended release, at 5 mg in the morning. This seemed to help, but there was still a thorny depressed mood and negative mindset, perhaps a bit worse in the afternoons as he withdrew from the stimulant. To address this, we added an SSRI, fluoxetine.

SSRIs. While studies generally have not shown SSRIs to be helpful for perseveration, there are individuals who respond at times, perhaps especially those who have more classic looking obsessions and compulsions. Similarly, while some people worry about SSRIs and suicidal ideation, I find these meds sometimes critically helpful in treating depressive symptoms, particularly suicidal ideation, in children and adolescents with ASD. They can help with anxiety as well.

However, most patients show some degree of behavioral activation on SSRIs, so dosing needs to be extremely gradual, and here we started with 5 mg fluoxetine. Even at this low dose, he became somewhat activated, so we reduced to 2.5 mg with good results. There are a couple of other potential side effects to keep in mind. First, a recent study found that normal doses of SSRIs can increase seizure risk, and patients with ASD are already prone to seizure activity. Second, while not relevant for this young boy, sexual side effects can actually be beneficial for people with ASDs who are struggling to control their libido.

Neuroleptics. Because he still had a lot of irritability, we added a small dose of neuroleptic. In my experience, haloperidol, risperidone, aripiprazole, olanzapine, and quetiapine all may be effective in helping people with ASDs to have less reactivity and, occasionally, far more clarity and complexity of thought. I have had less luck with medications that do not increase appetite, such as ziprasidone and lurasidone. Since these meds can cause cardiac rhythm changes, I get an EKG in addition to monitoring fasting lipids and glucose. I screen for abnormal involuntary movements using the AIMS if possible at a minimum of every six months, and I educate families about the possibility of neuroleptic malignant syndrome. In this case, we used 2 mg of aripiprazole and ended up in a common clinical dilemma because the medication was incredibly helpful but it caused an insatiable appetite and weight gain.

Anti-epileptic drugs (AEDs). Because of the weight gain, we discontinued the aripiprazole and tried an AED. Beyond treating seizures, many AEDs are used to treat irritability, but each has its potential issues. Valproate can be very helpful but requires careful monitoring of levels and liver function. Lamotrigine has few side effects beyond the serious though rare Stevens-Johnson Syndrome. Oxcarbazepine can reduce motor tone, leading to floppier posture. Gabapentin is very benign and can be helpful for anxiety and for sensory hypersensitivity. Topiramate can reduce perseverative thinking and typically reduces appetite, but may interfere with word finding, which can be problematic for this population. In this case we tried topiramate 25 mg, gradually increasing to 50 mg twice a day, with reasonable results in reducing both irritability and some of the perseveration.

Table: Autism Treatment Planning Chart
Click to view full-size table as PDF.


3. General guidelines that may be helpful
Regardless of which meds you recommend, here are some things to keep in mind—based on long, humbling experience.

  • Whenever we medicate a child, the team often becomes overly focused on medication, failing to look at other system problems. For example, a child who is reacting to parental separation may need more time in therapy rather than a new pill.
  • People with ASDs may have more sensitivity to medication, and families frequently want to stop a medication early when anything seems amiss. So make sure the team is aware of possible side effects ahead of time.
  • Start lower and go slower than usual when treating this population, and avoid multiple changes and big dosage alterations that can be poorly tolerated and cause confusion about what is or is not working.
  • Predict that if things get better, there will still be tough times when people may throw up their hands saying “nothing works.” If you’ve carefully documented the initial target symptoms, it will help you remind caregivers that there has been overall progress.
  • When the results are not clear after a decent trial, I ease down on a medication in a gentle on-off-on trial to help assess whether a medication is helping.

CCPR Verdict: Meds for autism: A complex art and science.