Professional Documents
Culture Documents
The Psychopharmacology of Autism Spectrum Disorder and Rett Syndrome
The Psychopharmacology of Autism Spectrum Disorder and Rett Syndrome
Chapter 24
Abstract
Autism spectrum disorder (ASD) appears in early childhood and is characterized by persistent deficits in
communication and social interaction, as well as restricted interests, repetitive behaviors, and unusual sen-
sory issues. ASD can be idiopathic or syndromic, in the latter case representing one of the many manifes-
tations of a genetic disorder, such as Rett syndrome. Psychopharmacology cannot directly ameliorate core
autistic symptoms, but rather aims at treating comorbid disorders, such as ADHD, sleep disturbances, psy-
chomotor agitation and aggressiveness, seizures, and anxiety. Until the 1990s, it was mainly based on typ-
ical neuroleptics and tricyclic antidepressants, whereas second-generation antipsychotics later became
first-line drugs for these same indications. In general, selective serotonin reuptake inhibitors have not
proven effective in ASD patients. Psychostimulants are frequently prescribed, but display modest efficacy
and enhanced potential for side effects and noncompliance. Targeted patients benefit from mood stabilizers
and antiepileptic drugs. Experimental drug treatments for ASD include oxytocin and vasopressin, bume-
tanide, sulforaphane, nutritional supplements, memantine, and D-cycloserine. Work performed on syndro-
mic forms, represents an important source of information for experimental therapies of ASD and
knowledge of the unique mechanisms underlying autism in each individual patient may in the future
pave the path to personalized drug treatments.
*Correspondence to: Antonio M. Persico, Interdepartmental Program “Autism 0-90”, “Gaetano Martino” University Hospital, via
Consolare Valeria 1, I-98125 Messina, Italy. Tel: +39-090-221-3143, Fax: +39-090-293-0414, E-mail: apersico@unime.it
392 A.M. PERSICO ET AL.
interventions (Tonge et al., 2014; Wong et al., 2015). As mechanism, they enhance extracellular monoamine
yet, pharmacologic treatment is not directed to the core levels by reducing their uptake. TCAs, such as desipra-
symptoms of ASD, but rather to comorbid disorders mine, imipramine, nortriptyline, amitriptyline as well
and problem behaviors, such as irritability, anxiety, as buproprion, a dopamine reuptake blocker, have been
aggression, self-injurious behavior, ADHD, sleep disor- used to treat individuals with autism for comorbid anxi-
ders, and seizures. Until the end of the 1980s, typical neu- ety, obsessive–compulsive disorder (OCD), major
roleptics and tricyclic antidepressants (TCAs) were the depression, and attention deficit hyperactivity disorder
only two viable options to reduce behavioral symptoms (ADHD), but clinical response and side effect severity
in ASD. Currently, the array of drugs that have proven appear less favorable compared to the nonautistic popu-
their efficacy on disorders and symptoms comorbid with lation and warrant caution. Antagonizing multiple cho-
ASD has significantly risen, although the vast majority linergic, histaminergic, and adrenergic receptors, in
must be prescribed “off-label” (Persico et al., 2015). addition to monoamine transporters, TCAs frequently
For most medications, limited data are available regard- produce side effects, including dry mouth, constipation,
ing efficacy in autistic individuals, especially children urinary retention, blurred vision, sinus tachycardia,
and adolescents, owing to small sample sizes, diagnostic impaired cognition, sedation, impaired motor coordina-
heterogeneity, and methodological difficulties in orga- tion, weight gain, balance problems, orthostatic hypoten-
nizing multicenter trials (Persico et al., 2015). sion, cardiac toxicity, central nervous system (CNS)
This chapter provides an overview of psychopharma- depression, and seizures. Nonetheless, despite producing
cologic agents either approved or currently in use for more frequent and severe adverse effects than, for exam-
ASD and Rett syndrome. We then summarize innovative ple, fluoxetine or fluvoxamine, in healthy patients with-
pharmacologic agents undergoing clinical trials, aimed at out seizures and with normal cardiac function, TCAs are
translating our ever-increasing knowledge of the patho- generally safe and well tolerated.
physiology of these disorders into pharmacologic tools The predominantly 5-HT reuptake blocker most stud-
able to ameliorate the core symptoms of ASD. ied in ASD is clomipramine. Superior to placebo and to
desipramine (a predominantly NE reuptake blocker), clo-
mipramine was able to improve autistic symptoms,
TYPICAL NEUROLEPTICS AND
anger, and repetitive behaviors in 57% of ASD subjects
TRICYCLIC ANTIDEPRESSANTS
enrolled in a double-blind RCT (Gordon et al., 1993).
Neuroleptics have been the most widely studied class of However, progressively increasing doses of clomipra-
psychopharmacologic agents in autism over the past mine, up to a maximum of 250 or 5.0 mg/kg/day for
35 years. Starting in the 1960s, children and adolescents 5 weeks, showed no therapeutic advantage and resulted
with autism were enrolled in randomized controlled in severe side effects in an open trial on eight ASD chil-
trials (RCTs) of first-generation antipsychotics, espe- dren aged 3.5–8.7 years (Sanchez et al., 1996). An RCT,
cially haloperidol, thioridazine, and trifluoperizine. comparing the efficacy of clomipramine, haloperidol,
These agents were found to decrease hyperactivity, and placebo in 31 subjects with ASD, found no signifi-
stereotypic behaviors, social withdrawal, aggressive- cant difference between clomipramine and placebo on
ness, and temper outbursts. For example, the first stereotypy, irritability, and hyperactivity recorded using
RCT, enrolling 45 children aged 2–7 years with DSM- the Aberrant Behavior Checklist (ABC) (Remington
III autistic disorder, showed significant positive effects et al., 2001). Moreover, many participants receiving clo-
of haloperidol (0.25–4.0 mg/day) on global functioning mipramine discontinued treatment due to side effects or
and multiple behavioral problems, as measured with lack of efficacy. Nortriptyline, a predominantly NE reup-
Children’s Psychiatric Rating Scale (CPRS) and Clinical take blocker, may reduce hyperactivity, aggressiveness,
Global Impression (CGI) scores, but no improvement in and ritualized behavior (Kurtis et al., 1966), and be better
learning abilities was recorded (Anderson et al., 1989). tolerated than imipramine (Campbell et al., 1971).
The major limitation of these drugs is their side-effect Tianeptine, pharmacologically distinct from TCAs but
profile, ranging from relatively benign dry mouth and sharing at least some properties with this drug class,
sedation to severe parkinsonism, acute dystonia, tardive was able to reduce abnormal behaviors, such as inade-
dyskinesia, and neuroleptic malignant syndrome. Des- quate eye contact and speech abnormalities, significantly
pite these adverse effects, haloperidol and other neuro- more than placebo when rated by parents and teachers,
leptics remained the most prescribed drugs in ASD but clinician ratings did not show any significant dif-
until the 1990s. ference between tianeptine and placebo (Niederhofer
TCAs block several monoamine transporters, pri- et al., 2003). Adverse effects, including drowsiness
marily the serotonin (5-HT) and norepinephrine (NE) and reduced activity levels, were recorded in a sizable
transporters SERT and NET, respectively. Through this number of patients treated with tianeptine.
PSYCHOPHARMACOLOGY OF ASD AND RETT SYNDROME 393
ATYPICAL ANTIPSYCHOTICS (0.175 mg/day for >45 kg) (Kent et al., 2013a). The side
effect profile of risperidone in individuals with autism is
Second-generation antipsychotics (SGAs), also known generally favorable. Neurologic side effects are rela-
as “atypical antipsychotics” emerged in the 1980s. This tively unusual, although acute extrapyramidal symptoms
new group of neuroleptics shares some effects with first- still occur. The most common side effect of risperidone in
generation neuroleptics, but differs in neurochemical autistic patients is also a metabolic syndrome, with
profile, effectiveness, and side effects. A shift toward increased appetite, weight gain, and, less frequently,
the use of SGAs was largely driven by their lower risk sedation. Direct comparison of haloperidol and risperi-
of extrapyramidal side effects, although they can pro- done in 30 children with ASD, aged 8–18 years, found
mote a metabolic syndrome, characterized by weight risperidone to be more effective than haloperidol in
gain, elevated lipid and prolactin serum levels, and pos- the treatment of behavioral symptoms, impulsivity,
sibly type 2 diabetes mellitus. In general, they have a dif- language skills, and impaired social relations (Miral
ferent receptor blocking profile, being less effective on et al., 2008). Finally, risperidone modulates important
dopamine D2 as compared to D1 receptors and also astroglial functions, possibly exerting antioxidant and
blocking many 5-HT receptor subtypes. Atypical anti- neuroprotective effects in ASD (Quincozes-Santos
psychotics include risperidone, paliperidone, aripipra- et al., 2009).
zole, olanzapine, clozapine, quetiapine, ziprasidone, Aripiprazole acts as an agonist, partial agonist, or
asenapine, sertindole, and zotepine. The FDA approved antagonist at dopamine D2 receptors, depending on cell
risperidone in 2006 and aripirazole in 2009 for the treat- type and function examined (Shapiro et al., 2003). It also
ment of irritability in ASD, including tantrums, aggres- acts as an inverse agonist at 5-HT2B receptors, as a partial
sion, and self-injury, whereas the European Medicine agonist at 5-HT1A, 5-HT2A, 5-HT2C, D3, and D4 recep-
Agency has only approved risperidone. tors, as an antagonist at 5-HT7, a1-adrenergic, and
Risperidone, a potent antagonist at dopamine D2 and H1-histamine receptors (Shapiro et al., 2003). This pecu-
serotonin 5-HT2A receptors, has been one of the most liar mechanism of action allows aripiprazole to act simul-
extensively studied medications in ASD. Several open taneously as an antagonist in hyperdopaminergic and as
studies have supported the efficacy, safety, and tolerabil- an agonist in hypodopaminergic brain regions or condi-
ity of risperidone in the treatment of autistic behaviors, tions. Several open-label studies have shown the tolera-
irritability, aggression, and hyperactivity (Findling bility and safety of aripiprazole therapy, as well as its
et al., 1997; McDougle et al., 1997; Nicolson et al., effectiveness for the treatment of hyperactivity and irri-
1998; Masi et al., 2001; Nishimura et al., 2003; tability in autistic children and adolescents (Stigler
Gagliano et al., 2004; Kent et al., 2013a,b; Lamberti et al., 2009; Marcus et al., 2011; Ishitobi et al., 2013;
et al., 2016). Multiple RCTs have also produced defini- Maloney et al., 2014; Lamberti et al., 2016; Ichikawa
tive evidence of efficacy on irritability and suggestive et al., 2018). Numerous RCTs have also provided strong
evidence for repetitive behaviors. Long-term efficacy evidence of efficacy in reducing irritability, hyperactiv-
of risperidone was reported in the 2005 RUPPAN study, ity, and stereotypic behavior in autistic children and ado-
in which an initial open-label 4-month treatment period lescents aged 6–17 years, both acutely and following
at an established effective dose was followed by an long-term treatment (Marcus et al., 2009; Owen et al.,
8-week double-blind, placebo-controlled withdrawal 2009; Aman et al., 2010; Robb et al., 2011; Findling
period; over 60% of youth relapsed while on placebo et al., 2014; Ichikawa et al., 2017). In addition to these
as compared to 12.5% on risperidone (RUPPAN, studies, an 8-week open-label trial conducted on 10 pre-
2005a). However, symptom improvement with risperi- schoolers with ASD, aged 3–6 years, reported a signifi-
done came at the expense of weight gain, which averaged cant reduction in irritability, stereotypic behavior, and
5.1 kg (RUPPAN, 2005a). Several double-blind RCTs hyperactivity (Habibi et al., 2015). Direct comparison
found risperidone effective in treating motor stereoty- of aripiprazole to risperidone in a 2-month double-blind
pies, hyperactivity, impulsivity, agitation, and aggression RCT showed similar efficacy in treating irritability for
in monotherapy (McDougle et al., 1998; McCracken the two drugs, as measured with the ABC-I, but a faster
et al., 2002; Malone et al., 2002; Shea et al., 2004; onset of action for aripiprazole (Ghanizadeh et al., 2014).
Troost et al., 2005; McDougle et al., 2005; Luby et al., The most frequent side effects include increased appetite
2006; Nagaraj et al., 2006; Pandina et al., 2007; Aman and weight gain, sedation, and psychomotor agitation,
et al., 2008; Kent et al., 2013a; Ghanizadeh et al., but side effects overall occurred less frequently com-
2014; Aman et al., 2015). Interestingly, in a double-blind pared to other SGAs, especially serious EKG and cardiac
RCT of 96 children, 5–17 years old, higher doses of ris- adverse events (De Hert et al., 2011; Ho et al., 2012).
peridone (1.75 mg/day for >45 kg) contributed to greater Children and adolescents tend to have more pronounced
change in the ABC-I as compared to lower doses side effects from aripiprazole when antipsychotic-naïve
394 A.M. PERSICO ET AL.
than if previously treated with another SGA, but drug adults who failed to respond to other first-line treatments
discontinuation rates are similar (10.8% vs 8.3%, (Zuddas et al., 1996; Chen et al., 2001; Lambrey et al.,
respectively) (Mankoski et al., 2013). 2010; Sahoo et al., 2017). It is used as second-line
Olanzapine use in ASD has been assessed only in one treatment because of potentially severe or even life-
small RCT (Hollander et al., 2006a), three prospective threatening side effects (especially agranulocytosis,
open-label trials (Potenza et al., 1999; Kemner et al., seizures, sedation, and cardiomyopathy), which require
2002; Fido and Al-Saad, 2008), and one open-label com- medical follow-up and regular blood tests for as long
parison with haloperidol (Malone et al., 2001). Hollander as the patient continues to take the medication. Quetia-
et al. (2006a) conducted an 8-week RCT on 11 autistic pine has produced contrasting results. Two small
children, aged 6–11 years, randomized to receive olanza- open-label studies found quetiapine not effective and
pine (mean dose 10 mg/d) or placebo. Significant poorly tolerated in children and adolescents with ASD
improvements were recorded in the primary outcome (Martin et al., 1999; Findling et al., 2004). However, a
measure (CGI-I), but not in secondary outcome measures retrospective study of 20 patients (aged 5–28 years)
(CY-BOCS and Overt Aggression Scale-Modified). who received a daily dose of quetiapine between
Common adverse effects included sedation and weight 25 and 600 mg over a period of 4–180 weeks showed
gain; rhinitis, “glazed eyes,” constipation, and insomnia modest improvements in maladaptive behavior, espe-
were also more frequent in olanzapine-treated children cially in CGI-S scores (Corson et al., 2004). Another ret-
compared to placebo. In a 6-week open-label study, rospective study suggests that 40%–60% of autistic
5/6 and 3/6 olanzapine- and haloperidol-treated ASD patients may benefit from quetiapine treatment
children, respectively, were rated as “responders” using (Hardan et al., 2005). Finally, an open-label trial of
the CGI-I (Malone et al., 2001). Collectively, these 11 adolescents aged 13–17 years found quetiapine well
results and those from open-trials indicate that olanza- tolerated and effective in treating aggressive behavior
pine may be beneficial, especially for irritability, to chil- and sleep disturbances, but less so in autistic behaviors
dren and adolescents who do not respond well to (Golubchik et al., 2011). Ziprasidone has an interesting
risperidone or aripiprazole. pharmacologic profile, characterized by: (a) strong
Lurasidone is a novel SGA with potent 5-HT2A, 5-HT2A receptor antagonism, (b) moderate antagonism
5-HT7, D2, and noradrenergic a1 receptor antagonism, at 5-HT1D, dopamine D2, noradrenergic a1, and hista-
as well as partial agonism at the 5-HT1A receptor mine H1 receptors, (c) antidepressant and anxiolytic
(Ishibashi et al., 2010). In addition to approved indica- effects through 5-HT1A receptor antagonism and
tions for adult schizophrenia and bipolar disorder, this norepinephrine reuptake inhibition. Ziprasidone
pharmacologic profile may also be associated with an (20–160 mg/day) can have beneficial effects for ASD
antidepressant effect and lesser side effects involving children and adolescents, without producing significant
weight gain and changes in metabolism. One recent mul- weight gain, metabolic changes, prolactin increase, or
ticenter double-blind RCT study enrolled 150 ASD other severe adverse effect (McDougle et al., 2002;
patients, aged 6–17 years, randomized to receive either Goforth and Rao, 2003; Duggal, 2007; Malone et al.,
lurasidone 20 mg/day (N ¼ 50), 60 mg/day (N ¼ 49), or 2007; Dominick et al., 2015). Side effects include QTc
placebo (N ¼ 51) for 6 weeks. Lurasidone was not signif- increases and occasional acute dystonic reactions
icantly superior to placebo in ABC Irritability or (Malone et al., 2007). The efficacy of ziprasidone on irri-
CY-BOCS scores at any dose, but produced significantly tability in youth with ASD may be somewhat lower com-
greater improvement at 20 mg/day over placebo using the pared to risperidone and aripiprazole, but it is well
CGI-I. Common side effects included nausea and drows- tolerated (Dominick et al., 2015). Paliperidone is the
iness, with only minor changes observed in weight and extended-release major active metabolite of risperidone.
metabolic parameters (Loebel et al., 2016). A case report Only two case reports and one 8-week open-label study
showed improvement in irritability, perseveration, and suggest that it may be effective on irritability in ASD
aggression in a 13-year-old autistic boy at an initial dose children and adolescents, at a mean final dosage of
of 10 mg/day, increased to 30 mg/day, without significant 7.1 mg/day (range 3–12 mg/day) (Stigler et al., 2010,
adverse effects (Millard et al., 2014). 2012; Kowalski et al., 2011). The drug is generally well
No RCT has been published to this date for clozapine, tolerated and side effects appear superimposable to those
quetiapine, ziprasidone, or newer SGAs, including pali- of risperidone, most frequently including weight gain
peridone, iloperidone, and asenapine. However, open tri- and increased prolactin levels (Stigler et al., 2012).
als report effectiveness, and these drugs are often used in There are no studies published to date on the use of
clinical practice. Clozapine has been reported to be effec- some newer SGAs, namely asenapine and iloperidone,
tive in improving hyperactivity and aggression in in ASD children and adolescents. These drugs are both
treatment-resistant autistic children, adolescents, and D2 and 5-HT2A receptors antagonists, approved by the
PSYCHOPHARMACOLOGY OF ASD AND RETT SYNDROME 395
FDA in 2009 for the treatment of adults affected by hyperactivity, impulsivity, and sleep disturbance, are
schizophrenia. common in ASD children and adolescents, frequently
In summary, antipsychotics can be very useful for leading to noncompliance or to treatment interruption.
children and adolescents with autism to treat irritability, Fluvoxamine has been proven effective in treatment
psychomotor agitation, temper outbursts, or aggression of repetitive, maladaptive, and aggressive behaviors in
toward self or others. Prominent hyperactivity and some adults, but not in children and adolescents with
impulsivity may also respond better to neuroleptic treat- ASD, who tend to experience adverse effects very fre-
ment than to typical first-line psychostimulants, in low- quently (McDougle et al., 1996, 2000; Sugie et al.,
functioning autistic individuals or in patients who 2005). Also, citalopram is not more effective than pla-
develop greater agitation when administered psychosti- cebo on repetitive behaviors in children and adolescents
mulants. Once problem behaviors are reduced, behav- with ASD, and is more likely to be associated with
ioral and educational interventions with autistic adverse effects, including increased energy level, impul-
children and adolescents often develop greater efficacy. siveness, decreased concentration, hyperactivity, stereo-
SGAs are equally or more effective than typical neuro- typy, diarrhea, insomnia, dry skin or itching, and
leptics and safer in terms of side effect profile. Typical prolonged seizure (King et al., 2009). There are no RCTs
neuroleptics remain a viable option when ASD patients of paroxetine, sertraline, escitalopram. A case report pro-
are nonresponsive to at least two SGAs, noncompliant vides anecdotal evidence of efficacy in the treatment of
to several SGAs due to important side effects, or when repetitive behaviors, anxiety, and temper tantrums for
SGAs, including clozapine, are contraindicated due to paroxetine in a 7-year-old nonresponder to fluoxetine
documented medical reasons. and fluvoxamine (Posey et al., 1999). One open-trial pro-
vides some evidence of efficacy for paroxetine on self-
injurious behavior in intellectual disability (Davanzo
SELECTIVE SEROTONIN REUPTAKE
et al., 1998). A retrospective study of 37 adults with intel-
INHIBITORS
lectual/neurologic disabilities and autistic traits reported
To date, three selective serotonin reuptake inhibitors improvement in only 13 cases and a high incidence
(SSRIs) have received regulatory approval in Europe of side effects, including vomiting, psychomotor agita-
for use in the pediatric population: fluoxetine for moder- tion, drowsiness, hypomanic behavior, aggressiveness,
ate to severe major depressive episodes starting at age 8, insomnia, seizures, and skin rushes (Branford et al.,
and sertraline and fluvoxamine for OCD after 6 and 1998). The effectiveness of sertraline in ASD children
8 years of age, respectively (Persico et al., 2015). These and adults is mainly linked to anxiety disorder. One case
drugs raised initial hopes of clinical usefulness in ASD, series of nine autistic children, aged 6–12 years, docu-
based on their well-known efficacy on obsessive– ments improvement in anxiety, irritability, and agitation
compulsive symptoms and on 5-HT blood levels, found with clinical response to low doses (25–50 mg/day)
elevated in approximately 25% of autistic individuals appearing generally in 2–8 weeks and, frequently, wan-
(Gabriele et al., 2014). Contrary to this expectation, ing after a few months (Steingard et al., 1997). Hence,
results with SSRIs in ASD have been generally disap- short-term sertraline treatment can reduce behavioral
pointing. Overall, RCTs involving fluoxetine, fluvoxa- reactions associated with situational transitions or envi-
mine, fenfluramine, and citalopram provided modest ronmental changes in children with ASD, though bene-
and conflicting evidence of efficacy on repetitive behav- ficial effects may be temporary. An open-label study
iors and compulsions (Hollander et al., 2005, 2012; showed significant improvement in repetitive and
Williams et al., 2013; Mouti et al., 2014), but fluoxetine aggressive symptoms, but not in social deficits, in
has been shown to increase regional blood flow in the 24 of 42 (57%) adults with ASD (McDougle et al.
anterior cingulate gyrus and orbitofrontal cortex 1998). The most common adverse effects were weight
(Buchsbaum et al., 2001). The antianxiety effects of flu- gain, anxiety, and psychomotor agitation. There is only
oxetine may also explain the occasional observation of one open-label study on 28 ASD children treated with
some improvement in (a) core ASD symptoms, including escitalopram, reporting significant improvement in
an increase in eye contact, social initiation and respon- CGI-I ratings and in ABC scores for irritability, lethargy,
siveness, and repertoire of interests; (b) irritability and stereotypy, hyperactivity, and inappropriate speech.
problem behaviors, including decreased tantrums and Dose-related adverse effects, notably irritability and
aggression toward self and others; (c) attention deficit hyperactivity, occurred at doses above 10 mg/day in
(initiating, shifting, and sustaining attention), leading 78% of cases (Owley et al., 2005).
altogether to better social adaptation and inclusion In summary, SSRIs are definitely more effective in
(Hollander et al., 2012). However, improvement is neurotypical individuals suffering from OCD and
often clinically marginal and side effects, especially depression than in autistic individuals affected by
396 A.M. PERSICO ET AL.
obsessive and anxious–depressive symptoms. Second, reducing inappropriate sexual behavior associated with
they appear relatively more effective and better tolerated ASD (Nguyen and Murphy, 2001; Albertini et al.,
in adult autistic individuals, as compared to ASD chil- 2006; Coskun and Mukaddes, 2008; Coskun et al.,
dren and adolescents. Nonetheless, a trial with an SSRI 2009). The efficacy of trazodone, a heterocyclic antide-
should be considered whenever compulsive behavior pressant, has been the object of only a few case reports
and anxiety are posing major hurdles to developmental, involving ASD children. One case report showed a
educational, and social progress. A valid trial should last reduction of aggression at a dose of 150 mg/day in
10–12 weeks but, contrary to nonautistic patients, shorter divided doses (Gedye, 1991); another case study
trials (i.e., 1–3 weeks) may suffice and prevent subse- reported beneficial effects of trazodone on sleeping in
quent behavioral activation. Low SSRI doses should an 8-year-old autistic boy also treated with fluvoxamine
be prescribed with frequent clinical monitoring of target (Parker and Hartman, 2002); priapism was described as a
symptoms and potential side effects. dose-dependent side effect in a 13-year-old autistic boy
(Kem et al., 2002).
SEROTONIN AND NOREPINEPHRINE
ANTIANXIETY MEDICATIONS
REUPTAKE INHIBITORS AND OTHER
ANTIDEPRESSANT MEDICATIONS Antianxiety agents such as benzodiazepines have been
relatively understudied in ASD. They are clinically use-
Among serotonin and norepinephrine reuptake inhibitors
ful at times in the treatment of anxiety, particularly as
(SNRIs), more data are available for venlafaxine, which
add-on to an SSRI. However, behavioral side effects such
inhibits the reuptake of serotonin, norepinephrine, and,
as disinhibition, crying, and irritability limit their use.
to a lesser extent, dopamine (Ellingrod et al., 1994). Sig-
Instead, two RCTs have been performed on buspirone,
nificant improvement in repetitive behaviors, restricted
a 5-HT1A receptor agonist. An 8-week RCT involving
interests, social deficits, communication and language,
40 ASD children and adolescents showed that low-dose
inattention, and hyperactivity were reported in an open,
buspirone is more effective than placebo as add-on
retrospective study involving 10 autistic children,
to risperidone for treating irritability (Ghanizadeh
adolescents, and young adults (Hollander et al., 2000).
and Ayoobzadehshirazi, 2015). At a mean dose of
Clinical response was observed at low doses
6.7 mg/day, 13/16 (81.2%) buspirone-treated patients
(6.25–50 mg/day, mean daily dose 24.37 mg) and the
showed a 30% decline in irritability scores compared
drug was well tolerated. Another open trial reported
to 7/18 (38.9%) patients in the placebo group. Contrary
beneficial effects of low-dose venlafaxine on self-injury
to buspirone alone, which tends to decrease appetite and
and hyperactivity in two 17-year-old autistic boys
body weight, the most common adverse effect adding
and in a 23-year-old autistic woman (Carminati et al.,
buspirone onto risperidone was increased appetite, as
2006). A recent RCT found no convincing evidence of
well as drowsiness and fatigue. Another recent multicen-
efficacy, contrasting six patients receiving venlafaxine
ter RCT enrolled 166 ASD preschoolers aged 2–6 years
(18.75 mg/day), along with their usual treatment (zuclo-
randomized to 24 weeks of 2.5/5.0 mg of buspirone twice
penthixol and/or clonazepam), with seven patients who
daily vs placebo (Chugani et al., 2016). No difference in
received placebo plus usual care (Carminati et al.,
ADOS Composite Total Score among the 3 treatment
2016). At the end of the study, global improvement
groups was recorded at 24 weeks, but the ADOS
was observed in 33% and 71% of venlafaxine- and
Restricted and Repetitive Behavior score showed a
placebo-treated patients, respectively, with irritability
time-by-treatment effect: 2.5-mg buspirone produced
improving equally in the entire sample. Only marginal
significant improvement, whereas placebo and 5.0-mg
evidence of lower cumulative doses of clonazepam and
buspirone yielded no change. Adverse events did not dif-
zuclopenthixol in the venlafaxine-treated group was
fer significantly among treatment groups. This study thus
obtained. Mirtazapine, a TCA, acting as an antagonist
supports low-dose buspirone treatment (2.5 mg/day) as
at a2 adrenergic and at multiple serotonin receptors,
an adjunct therapy to target restrictive and repetitive
improved aggressiveness, self-injury, irritability, hyper-
behaviors in conjunction with behavioral interventions.
activity, anxiety, depression, and insomnia in 9 of
26 (34.6%) ASD subjects, according to one open-label
ANTIEPILEPTICS AND MOOD
study (Posey et al., 2001). Mirtazapine did not improve
STABILIZERS
core social and communication deficits. Adverse effects
were minimal and included increased appetite, irritabil- Antiepileptic drugs (AEDs) have been extensively
ity, and transient sedation. Other case reports suggest employed in ASD as mood stabilizers and to attenuate
efficacy for this drug, at a dose ranging from 5 to problem behaviors, but only a limited number of open-
30 mg/day, not only on aggressiveness but also in label studies and RCTs have been performed in autistic
PSYCHOPHARMACOLOGY OF ASD AND RETT SYNDROME 397
children and adolescents (Hirota et al., 2014; significantly lower (Wang et al., 2017). EEG normaliza-
Canitano, 2015). tion was significantly more frequent in the treatment
Valproic acid was found to attenuate irritability in an group (Wang et al., 2017).
initial open trial including 14 young ASD patients (with Topiramate is an interesting drug, because of its
and without seizures) (Hollander et al., 2001). Three sub- weight-containing and mood-stabilizing effects, but the
sequent RCTs enrolling 12–30 ASD subjects yielded few studies performed to date have yielded conflicting
contrasting results (Hellings et al., 2005; Hollander results. An 18-month open-trial involving 10 ASD chil-
et al., 2006b, 2010). Hellings et al. (2005) found no sig- dren and adolescents undergoing weight gain with SGA
nificant difference in ABC irritability scores, with high treatment reported modest variation in weight excess and
interindividual variability and large placebo effect. On frequent adverse behavioral effects, dampening enthusi-
the contrary, using more severe symptomatic inclusion asm for topiramate add-on use (Canitano, 2005). Simi-
criteria to reduce interindividual variability, Hollander larly, limited efficacy and frequent side effects were
et al. (2010) showed a significant advantage of sodium reported in another case series (Mazzone and Ruta,
divalproex over placebo. Furthermore, Hollander et al. 2006). However, an 8-week RCT involving 40 autistic
(2006b) reported significant improvement in repetitive children, 4–12 years old, randomly allocated to receive
behaviors in a small RCT involving 13 autistic children. either topiramate + risperidone (Group A) or placebo
Valproate was also found to attenuate the irritability asso- + risperidone (Group B), showed beneficial effects for
ciated with fluoxetine treatment in ASD significantly topiramate on irritability, stereotypic behaviors, hyperac-
better than placebo (Anagnostou et al., 2006). tivity, and/or noncompliance (Rezaei et al., 2010). Lamo-
Levetiracetam binds to the synaptic vesicle glycopro- trigine, an AED mainly active in reducing voltage-gated
tein SV2A, inhibiting presynaptic calcium channels and Na+ currents, showed no efficacy on irritability and
reducing neurotransmitter release. This pharmacologic social behavior in an RCT involving 27 autistic children
action limits nerve signal conduction across synapses aged 3–11 years old (Belsito et al., 2001). The use of
(Lynch et al., 2004; Vogl et al., 2012). The use of levetir- oxcarbazepine for irritability and agitation in ASD has
acetam in ASD is controversial due to several studies been addressed in only 1 case series, involving
reporting major behavioral side effects. Levetiracetam 30 patients, aged 5–21 years, taking on average 2 other
was initially found to improve attention, hyperactivity, psychotropic medications, including SGAs, SSRIs,
emotional lability, and aggressiveness in an open trial alpha agonists, and divalproex (Douglas et al., 2013).
involving six drug-naive ASD boys (Rugino and Fourteen patients (47%) achieved a CGI-I score of
Samsock, 2002). Wasserman et al. (2006) compared “much improved,” while another 10 patients (33%) were
levetiracetam to placebo in a sample of 20 ASD children, “improved”; 7 patients (23%) had a clinically significant
aged 5–17 years, and found no evidence of efficacy on adverse effect leading to drug discontinuation. Finally, a
global functioning, irritability, repetitive behaviors, recent case report points toward possible beneficial
impulsivity, and hyperactivity. The most relevant side effects of low-dose phenytoin (5 mg/day for over
effect was aggression. Levetiracetam was also found 18 months) on verbal and nonverbal communication in
ineffective for controlling stereotyped movements in a 19-year-old autistic young man treated with stimulant
patients with ASD (Rajapakse and Pringsheim, 2010). medication since early childhood (Bird, 2015).
Importantly, one case report documents an autistic Lithium is a “classical” mood stabilizer also used in
regression, characterized by stereotypic behaviors and autistic individuals with comorbid recurrent affective
severe social and communicative dysfunction after disorder and/or mood instability, especially in the pres-
administration of levetiracetam in a 6-year-old girl with ence of mania, euphoria, or elevated mood. There are
cerebral palsy, whose main symptoms were mild intellec- currently no RCTs of lithium in children and adolescents
tual disability and focal epilepsy. This regression sub- with ASD. Two older case reports document the efficacy
sided after drug discontinuation, suggesting that of lithium therapy, the first one at blood levels
levetiracetam may occasionally provoke ASD-like >1.0 mEq/L on two 4-year-old autistic children with
behavioral adverse effects (Camacho et al., 2012). hyperactivity resistant to stimulant drugs and with cycli-
Finally, 70 children with ASD and subclinical epilepti- cal mood swings (Kerbeshian et al., 1987); the second
form discharges (SEDs) identified by EEG were ran- case report describes a beneficial response to lithium in
domly divided into two equal groups treated either two individuals with ASD whose symptoms included
with levetiracetam (60 mg/kg/day) and educational train- agitation, self–injurious behavior, aggression, and
ing (treatment group) or with educational training only insomnia (Steingard and Biederman, 1987). In a more
(control). At the 6-month follow-up, PEP-3 scores were recent retrospective chart review of 30 children and ado-
significantly higher for the treatment group compared lescents with ASD treated with lithium because of
to controls, whereas CARS and ABC scores were comorbid bipolar disorder, 43% were rated as improved
398 A.M. PERSICO ET AL.
on CGI-Irritability scores and 71% of patients with two possibly more in children than in adults. However,
or more pretreatment mood disorder symptoms were response rates tend to be somewhat lower, and adverse
rated as improved (Siegel et al., 2014). Patients most effects more frequent, in autistic individuals compared
improved were those with mania and elevated mood. to typically developing children with ADHD
Mean lithium blood level was 0.70 mEq/L, and the mean (RUPPAN, 2005b). Interindividual variability in
duration of lithium treatment was 29.7 days. At least one response to MPH by autistic children may be influenced
adverse effect was reported by 47% of patients, most by multiple monoaminergic gene variants (McCracken
commonly vomiting (13%), tremor (10%), fatigue et al., 2014). Six RCTs of MPH in children with ASD
(10%), irritability (7%), or enuresis (7%) (Siegel have been performed to date. In the initial RUPPAN
et al., 2014). study (2005b), 66 of 72 children, aged 5–14 years, with
A very interesting case report describes the efficacy of a DSM-IV diagnosis of Autistic Disorder or Pervasive
lithium as a rescue therapy for regression and catatonia and Developmental Disorder-Not Otherwise Specified
autistic traits in a 21-year-old man and in a 17-year-old girl (PDD-NOS) were randomized to a double blind cross-
with Phelan–McDermid syndrome, due to chr 22q13.33 over study of MPH vs placebo. A significant improve-
deletion involving the SHANK3 gene (Serret et al., ment was recorded on the ABC hyperactivity subscale,
2015). The former patient, unresponsive to multidrug ther- with 49% of participants responding at all MPH dose
apies including antipsychotics and other psychotropic increments and maximum benefits at medium-high doses
drugs, responded to 1500 mg/day of lithium, yielding (0.25 and 0.5 mg/kg per dose, respectively). Improve-
blood levels of 0.8 mEq/L. Lithium successfully corrected ment in hyperactivity/impulsivity and inattention, with-
a severe regression characterized by increasing psycho- out significant worsening of stereotypic or oppositional
motor agitation, aggressiveness, impulsivity, urinary, and defiant behavior, was later confirmed in the same sample
bowel incontinence, promoting cognitive and social using other outcome measures, such as SNAP-IV and
improvement over the subsequent 3-month period. The CY-BOCS scores (Posey et al., 2007), with improvement
latter patient underwent a period of severe behavioral again most evident at 0.25–0.50 mg/kg doses. The most
regression and catatonia. One year of lithium treatment common adverse effects of MPH included irritability,
(1000mg/day, lithium blood level 0.7 mEq/L), in associa- appetite loss, insomnia, and emotional outbursts. Similar
tion with a low dose of clonazepam (0.2 mg/day), spurred results were obtained in other RCTs involving autistic
improvement in motor and verbal abilities allowing the children (Quintana et al., 1995; Handen et al., 2000;
patient to recover to similar levels of functioning as Jahromi et al., 2009). Dosage for ASD children ranges
before regression. In both cases lithium was well tolerated. from 7.5 to 50 mg/day, sometimes divided and often
Another case report is on an adult male with Phelan– dosed by weight (0.3–0.6 mg/kg/day); for preschool chil-
McDermid syndrome whose mood and behavior dren ranges between 5 and 20 mg/day in divided doses
responded to a combined treatment with lithium and olan- (Doyle and McDougle, 2012). A 4-week RCT conducted
zapine, following resistance to various combinations of by Pearson et al. (2013) in 24 ASD children, with mean
SGAs and AEDs (Egger et al., 2017). age 8.8 years and comorbid ADHD, demonstrated ben-
Future prospective studies of lithium are needed to eficial effects on hyperactivity, with the combination
assess whether lithium strictly benefits comorbid bipolar of extended-release MPH in the morning and
symptoms or improves some ASD-specific symptoms. immediate-release MPH in the afternoon. This study is
especially interesting, because this combination is fre-
quently applied in clinical practice. In a recently pub-
STIMULANTS AND a2 ADRENERGIC
lished RCT (Kim et al., 2017), 27 subjects, aged 5–17
AGONISTS
years with ASD and ADHD, were randomized to receive
The comorbidity between ASD and ADHD, now recog- either low (10 mg/day) or medium (30 mg/day) doses of
nized by DSM-5, is present in 30%–40% of autistic indi- an extended-release liquid formulation of MPH. Signif-
viduals (Berenguer-Forner et al., 2015). This high icant decreases in irritability and hyperactivity ABC
incidence and the negative impact of inattention, hyper- scores were recorded more frequently in the medium
activity, and impulsivity on treatment programs and on dose group (mean dose ¼ 20.28 mg/day). Also, rebound
social adjustment, have spurred interest in the use of psy- hyperactivity and aggression at the end of the day were
chostimulants in ASD patients with comorbid ADHD. significantly lower only in the medium-dose group.
Methylphenidate (MPH) acts as a dopamine and, to a In several studies, and also in clinical practice, intel-
lesser extent, as a norepinephrine reuptake blocker. It lectual disability and lower IQ tend to be associated with
is the psychostimulant most prescribed for children with lower response rates to MPH and with more side effects
ADHD. In general, MPH is known to be moderately (Aman et al., 2003). However, at least 1 RCT involving
efficacious in improving ADHD symptoms in ASD, 122 children and adolescents with severe ADHD and
PSYCHOPHARMACOLOGY OF ASD AND RETT SYNDROME 399
intellectual disability (including many with comorbid norepinephrine reuptake blocker, in reducing depressive
ASD) found MPH (0.5/1.0/1.5 mg/kg/day) significantly and inattentive/hyperactive symptoms (Golubchik et al.
superior to placebo and no correlation between IQ 2013). A significant, yet modest, decrease in the severity
or ASD symptoms and clinical response (Simonoff of depressive and ADHD symptoms assessed with
et al., 2013). CDRS and ADHD-RS, respectively, was recorded.
Atomoxetine is a selective dopamine and norepineph- A significant positive correlation was found between
rine reuptake inhibitor, approved for the treatment of changes in depression and ADHD severity scores. Most
ADHD in children, adolescents, and adults starting at patients reported minimal or tolerable side effects.
6 years of age. A 4-week RCT performed on 97 ASD Clonidine is a selective a2 adrenergic and
individuals with comorbid ADHD, aged 6–17, found imidazoline-1 receptor agonist, approved by the FDA
some evidence of efficacy for atomoxetine (1.2 mg/kg/ in 2010 for the treatment of ADHD in pediatric patients
day) on hyperactivity, stereotyped behaviors, inappro- starting at age 6. Two RCTs provided evidence of effi-
priate speech, and fear of change, but no significant cacy on impulsivity, hyperarousal, and self-stimulating
global improvement using the CGI; side effects included behavior in ASD children and adults (Fankhauser
nausea, decreased appetite, fatigue, and early morning et al., 1992; Jaselskis et al., 1992). This drug is also effec-
awakenings (Harfterkamp et al., 2012). A subsequent tive on hyperactivity, aggressiveness, mood instability,
open-label follow-up study conducted on 88 patients irritability, and on nighttime awakenings, improving
taking the drug for 28 additional weeks showed impro- sleep quality (Ming et al., 2008). In these studies, drug
vement in hyperactivity, impulsivity, and inattention dosage ranges from 0.1 to 0.2 mg/day and side effects
(Harfterkamp et al., 2013). Adverse events, including included drowsiness, sedation, hypertension, and
gastrointestinal symptoms, irritability, tinnitus, mood decreased activity. Guanfacine is a central a2 adrenergic
swings, and sedation, were generally mild and tended agonist. A retrospective study of 80 ASD children and
to diminish over time, especially nausea and fatigue. adolescents, aged 8–18 years, reported improvement in
Dosages generally ranged from 1.2 to 1.4 mg/kg/day. hyperactivity, inattention, insomnia, and tics (Posey
Another recent RCT in 128 ASD children (ages 5–14) et al., 2004a). An open trial found 12 out of 25 (48%)
lasting 6 weeks for dose adjustment based on clinical ASD children, either nonresponders or intolerant to
response (up to a maximum of 1.8 mg/kg/day), followed MPH, responded well to guanfacine (Scahill et al.,
by dose maintenance for an additional 4 weeks, con- 2006). The efficacy of this drug in improving hyperactiv-
firmed that: (a) the most frequent side effects are ity in ASD children was then confirmed by two RCT
decreased appetite and fatigue, (b) most side effects do studies (Handen et al., 2008; Scahill et al., 2015). Dos-
not raise concern and last 4 weeks of less, and ages ranged between 0.25 and 9 mg/day, often in divided
(c) unlike typically developing children, ASD children doses. The most frequent side effects included irritability,
did not display serious adverse events, including QTc drowsiness, decreased energy, sleep disturbance, consti-
changes or suicidal ideation (Tumuluru et al., 2017). pation, headache, and nocturnal enuresis (Handen et al.,
An interesting recent RCT demonstrates the effective- 2008; Scahill et al., 2015).
ness of atomoxetine over placebo, underscoring at the
same time the usefulness of parent training in the context
FUTURE PERSPECTIVES IN THE
of an effective multidisciplinary treatment approach. In
PHARMACOTHERAPY OF AUTISM
particular, the combination of atomoxetine + parent train-
SPECTRUM DISORDER
ing (ATX + PT) produced higher response rates with
lower atomoxetine doses compared to the doses pre- The lack of psychoactive drugs able to ameliorate the
scribed in the “atomoxetine alone” group. Meanwhile, “core” symptoms of ASD is currently felt as perhaps
ATX + PT and ATX alone were superior to placebo the main unmet need in pediatric neuropsychophar-
+ PT and to placebo alone, respectively, in improving macology (Persico et al., 2015). For this reason, an
ADHD symptoms (Handen et al., 2015). A subsequent impressive effort is being put forth by academia and
24-week extension study involving 117 children, aged pharmaceutical companies to move the field of autism
5–15, out of the 128 (91%) acute-trial participants, con- psychopharmacology forward. No other behavioral
firmed the efficacy of atomoxetine, providing further evi- disorder of childhood and adolescence is the target of
dence of the greater benefits obtained by combining as many ongoing clinical trials listed on the NIH web
pharmacologic and psychosocial treatments over either site https://clinicaltrials.gov/. Some experimental drugs
treatment alone, both in terms of treatment outcome derive from animal and human research on monogenic
and patient compliance (Smith et al., 2016). forms of syndromic autism, for example, Rett syndrome,
An open-label trial of 11 children with ASD explored fragile X, tuberous sclerosis, and Phelan–McDermid
the safety and efficacy of reboxetine, a selective syndrome, among many. Other drugs derive from
400 A.M. PERSICO ET AL.
cellular, animal, molecular, and human studies of idio- Both oxytocin and vasopressin have a short half-live
pathic ASD, whose pathogenetic underpinnings are of around 3 min in plasma and 20 min in CSF, where they
beginning to be relatively well understood. Drugs are inactivated by placental leucine aminopeptidase
derived from research on syndromic forms were initially (LNPEP) (Gazis, 1978). Therefore, only through the
spurring high expectations, because in mouse models nasal route is it possible to ensure that a substantial
they displayed a surprising degree of effectiveness when amount of the active ingredient reaches the CNS, given
administered not only prenatally, but also postnatally or its peptide nature and half-life; in fact, if this therapy was
even in adult animals. Unfortunately, these drugs have administered parenterally, only 1% of these neuropep-
not met their primary endpoints in human RCTs, once tides would reach the CNS unmodified.
again proving that animal and cellular models can indeed Based on these data, intranasal oxytocin and vaso-
contribute preliminary but never conclusive evidence of pressin, as well as the AVPR1a antagonist RG7314,
efficacy, given the exquisite complexity of the are under scrutiny. Intranasal oxytocin, certainly the most
human brain. studied compound in current experimental ASD pharma-
On the other hand, among drugs derived from patho- cology, yields insufficient evidence of efficacy on core
genetic studies of idiopathic ASD, some may display ASD symptoms, according to two recent meta-analyses
measurable efficacy on core ASD symptoms, although including 17 and 12 RCTs, respectively (Keech et al.,
with different levels of evidence. We hereby summarize 2017; Ooi et al., 2017). Given the genetic variability of
the most studied and/or the most promising pharma- the OXTR gene and anatomical differences in nasal cav-
cologic agents in the experimental pipeline for the ity influencing oxytocin absorption into the CNS, this
treatment of core autistic symptoms: interindividual variability in response to intranasal
oxytocin in ASD is not surprising.