Perkins School for the Blind Transition Center

An Autism Spectrum News Interview with Alexander Kolevzon, MD, Clinical Director of the Seaver Autism Center for Research and Treatment at Mount Sinai School of Medicine

We are indeed fortunate to have an opportunity to speak with Dr. Alexander Kolevzon, Clinical Director of the Seaver Autism Center for Research and Treatment at Mount Sinai School of Medicine, located in New York City.

Dr. Kolevzon eloquently provides us with an overview of our current understanding of what causes autism, and outlines the medical and psychiatric conditions currently understood to be associated with ASD. In addition, Dr. Kolevzon discusses the vital research and clinical trials being conducted at the Seaver Autism Center, in collaboration with other international research centers. He explains how this will help us to understand more about the genetic basis of autism spectrum disorders and how this knowledge will then be used to develop novel treatments. His comments conclude with a message to families that they should remain hopeful that our understanding of the relationship between genetics and autism is rapidly improving, and is leading to increased early detection of an autism spectrum disorder in individuals as early as 12-18 months.


Can you give us a general overview of our current understanding of the cause(s) of Autism?


Our understanding of autism has evolved dramatically since Kanner’s original description of the disorder in 1943. Twin studies carried out since the 1970s and more recent genetic analyses have come to demonstrate autism as primarily a genetic disorder. For example, research has shown that if one identical twin has autism, then the likelihood that the other identical twin will have an autism spectrum disorder (ASD) is approximately 90%. On the other hand, when a non-identical twin has autism, the chance that the other twin will have autism is no more than 5-10% – even though they share the same uterine, household, and family environment. This is a very convincing argument for a strong genetic contribution of autism.

Soon after the early twin studies, theories of common susceptibility variants arose where estimates of between 10 and 100 genes were thought to contribute to autism, all interacting with some degree of weak effect. These susceptibility genes are thought to be common variations in our genetic code resulting from single nucleotide polymorphisms (ie, SNPs) that in and of themselves may not produce clinically significant changes but when several come together and interact, can produce autism. While there may still be a role for common variants in the etiology of autism, these variants remain elusive and studies identifying them have not been consistent. Much larger sample sizes will be required to reliably detect common variations with weak effect.

Today, most scientists in the field of autism recognize that rare variations in the genetic code account for a significant proportion of cases. Such rare variants can arise from copy number variations (CNVs) which are changes in the number of copies of a gene or portions of it and can be caused by deletions, duplications, inversions, and translocations of pieces of the genetic code. High-resolution genetic techniques like array comparative genomic hybridization (aCGH) are now able to explore large swaths of the human genome and search for differences in copy number between a given individual and what is typically expected in healthy controls. The current yield in detecting causes of autism when these kinds of techniques are utilized is approximately 15-20% and rising. These CNVs are often occurring spontaneously in the child, but may also be inherited in a minority of cases, so it is also critical to perform genetic analyses in parents.

This kind of approach to understanding the cause of autism is extremely significant for families for several reasons. First, it can provide a medical diagnosis for autism and in each case, depending on the cause, these genetic aberrations may be associated with other medical comorbidity that will be important to treat and monitor. Second, it can provide families with estimates of recurrence risk should they be considering having more children. In other words, if the mutation or deletion in the genetic code occurred spontaneously in the affected child, then the recurrence risk can be about the same as for the general population. However, if the parents have similar changes in their genetic code, it may confer increased risk for a sibling to have autism. Finally, every identified cause of autism provides the opportunity to develop model systems (e.g., mouse models) by knocking out the causal gene and the model systems can be used to understand the neurobiology of the disorder and most importantly, to develop targeted novel therapeutics.

It is important to understand that autism is a behavioral diagnosis made based on observations of language, social cognition, and repetitive behaviors. The more we learn about the genetic causes of autism, the more we come to understand its etiology as residing in genes that regulate processes of neuronal migration, differentiation, synaptic plasticity, and the formation of neuronal networks. There are many, many single gene defects and known genetic syndromes that account for a growing percentage of cases of autism. The American College of Medical Genetics now recommends aCGH as a first line test for children with ASD (Genetics in Medicine, 2010; 12 (11): 742-745). Comprehensive clinical evaluation and genetic testing is required in children with autism to clarify underlying medical diagnoses and determine appropriate monitoring.


Can you give us a general overview of our current understanding of the medical conditions associated with Autism?


Some of what used to be considered comorbidity can now be considered signs and symptoms of underlying medical conditions causing the autism. Several examples illustrate the issue of comorbidity and the significance of detecting causes of autism. The most well-known example is Fragile X Syndrome. If you examine 100 individuals with Fragile X Syndrome, approximately 25-40% will have an ASD. And if you look at 100 individuals with autism, 1-2% will have Fragile X Syndrome. Silencing of the Fragile X mental retardation 1 gene (FMR1) is the cause of the Syndrome and the cause of the autism in these cases. Silencing of this gene is thought to lead to excess glutamate dependent protein synthesis and results in significant impairment in learning and memory likely responsible for the intellectual disability characteristic of Fragile X Syndrome. Fragile X Syndrome is also associated with a number of medical complications, including seizures, cardiac problems (e.g., mitral valve prolapse), hernias, joint problems, and scoliosis among others. People with Fragile X Syndrome also have increased rates of attention deficit, hyperactivity, and anxiety. Rett’s Disorder is another important example and a medical condition that causes autism where the precise gene, MECP2, has been identified on the X chromosome. It is critical to identify these children early as severe neurological problems may develop, including loss of muscle tone, feeding problems, apraxia, and breathing difficulties (e.g., apnea). Another example is illustrated by work in the Seaver Autism Center by our Director, Dr. Buxbaum, and colleagues where mutations in the PTEN gene were found in a child with autism and intellectual disability. Mutations and deletions in the PTEN gene also cause Cowden Syndrome which is associated with increased risk of certain types of cancers and noncancerous tumor-like growths called hamartomas. Phelan McDermid Syndrome is yet another important example of a medical condition causing autism. Phelan McDermid Syndrome is caused by mutations and deletions in the SHANK3 gene located on chromosome 22 and responsible for approximately 1-2% of cases of autism. It is associated with neonatal hypotonia, motor skills deficits and gait disturbance, and absent or severely delayed speech. Affected individuals are also at greater risk of seizures, renal abnormalities, lymphedema, gastroesophageal reflux, and arachnoid cysts, all of which require careful monitoring.

Other comorbidities common to ASD include intellectual disability, sleep disorders, gastrointestinal symptoms (e.g., bloating, diarrhea, and constipation), and fine/gross motor deficits. Less common but still significant are the presence of metabolic disorders, several of which have been associated with autistic features. Phenylketonuria, for example, has been associated with autism when uncorrected. And while rare, there is growing interest in the role of mitochondrial dysfunction in autism and in inborn errors of cholesterol synthesis (i.e., Smith-Lemli-Opitz Syndrome).


Can you give us a general overview of the psychiatric conditions associated with Autism?


One of the reasons ASD is so challenging to treat is because there is no uniform approach that works for a majority of children. Treatment needs to be tailored to individual needs and this is particularly true for medication treatment, where the issue of comorbidity plays a significant role. In addition to the clinical triad of symptoms that characterizes ASD (i.e., language, social, and behavioral impairments), there is a wide range of associated features that add to the complexity. Anxiety occurs as part and parcel of the repetitive behavior domain and is often manifested by symptoms consistent with obsessive compulsive disorder. People with autism can become preoccupied with rigid adherence to routines and engage in frequent ritualistic behaviors to soothe their anxiety but also to stimulate at times. When these routines are interrupted, some patients may erupt with impulsive aggression, which is probably one of the most common reasons families seek pharmacological treatment. Attention deficit, hyperactivity, and impulsivity are other common associated features and a frequent target for medication intervention. Although the current diagnostic criteria prohibit the concurrent diagnosis of Attention Deficit/Hyperactivity Disorder (ADHD) and ASD, ADHD symptoms can occur in a large percentage of kids with ASD and this exclusion is due to change in the next iteration of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V).

Dr. Emily Simonoff from Kings College London published an epidemiological study looking at the 3 month point prevalence of psychiatric comorbidity among 112, 10-14 year old children with ASD (Journal of the American Academy of Child and Adolescent Psychiatry, 2008;47(8):921-929). The results suggest that 70% of participants had at least one psychiatric disorder in addition to ASD and about 40% had two or more. The most common diagnoses were social anxiety disorder, ADHD, and oppositional defiant disorder. Other common comorbidity included panic disorder, obsessive compulsive disorder, and tic disorder. While psychosis and paranoia are often discussed in the context of ASD, the rates of true psychotic symptoms are probably not greater than in the general population. Patients with autism may become obsessively preoccupied with perceived injustice or wrongdoing; though it may appear paranoid, they are more likely to be misperceiving social cues or obsessively preoccupied as opposed to experiencing paranoid delusions.

At the moment, currently available medications are not reliably effective to treat core symptoms of ASD. While some medications have shown evidence of efficacy for the repetitive behavior domain (e.g., selective serotonin reuptake inhibitors), more recent studies indicate that the percentage of patients likely to respond is around 40%, and not a greater percentage than respond to placebo. The medication treatment of associated features, like ADHD symptoms or aggression, is more reliable. Stimulants (e.g. methylphenidate/ritalin) and non-simulants (e.g., atomoxetine/strattera) effectively treat ADHD symptoms in some patients with ASD although less reliably than in typically developing children with ADHD. And second generation antipsychotics like risperidone (risperdal) and aripiprazole (abilify) are effective to treat irritability and aggression associated with ASD. Of course, medication must be used with extreme caution and certainly the more powerful antipsychotics require careful monitoring of weight and metabolic functioning, among other potential side effects.

The more we learn about the genetic causes of autism, the better we understand the neurobiology and the greater the potential for developing novel therapeutics targeted at the specific cause of the autism. Taking this sort of approach, the treatment of both core and associated features of autism will likely reach new depths in the coming years.


Are there standard methods used to diagnose the various types of co-morbid conditions associated with autism?


Tools used to diagnose comorbid conditions in the general population can also be used in working with individuals with ASD. Using clinician, parent, or teacher ratings can be particularly helpful for tracking progress after an intervention is implemented. However, standardized tools are less helpful in refining diagnoses in complex disorders like ASD. At this time, refined diagnostic conclusions are most reliably made by clinicians with experience working with and treating comorbidity in ASD.


As seizure disorders associated with autism are a big concern for parents, can you comment on the warning signs, complications, and behavioral regression that are sometimes seen with seizure disorders?


Seizures are estimated to occur in approximately 30% of individuals with autism and a significantly larger percentage has abnormal activity patterns on electroencephalography (EEG) without frank seizures. There is also frequently an association between behavioral symptoms of autism (e.g., agitation) and seizures. Likewise, the effective treatment of seizures in autism is often associated with behavioral improvement.

There are typically two peaks for the onset of seizure disorders in autism. The most common is in pre- and early adolescence between the ages of 10-13. The other peak is in early infancy, but when they occur this early, it is often associated with a comorbid medical condition like tuberous sclerosis, for example. Seizures may appear without warning signs or known triggers and can of course be very scary for families. But the onset of seizures does not typically have any association with the severity of other symptoms of autism. In general, behavioral regression associated with seizures is quite rare.

Estimates of regression in autism vary and were originally thought to be about 30%. However, as high-risk siblings of children with autism are now being followed, the longitudinal course of the syndrome is becoming better understood and rates of true regression appear much lower. There is also significant controversy about the role of epilepsy and EEG abnormalities and regression in autism; some studies have reported higher rates of epilepsy in patients with regression and others have found no relationship at all.

Treatment for seizures in autism is the same as treatment for seizures in the general population. Anticonvulsants, like divalproex sodium (depakote) or carbamazepine (tegretol) are commonly used. Children with autism do tend to be exquisitely sensitive to medication in general, and the anticonvulsants are no exception. Whenever possible, treatment should be initiated at very low doses and increased slowly. The advantage to most anticonvulsants is that medication levels in the blood can be carefully monitored, although of course drawing blood can be highly distressing for some children and their families. Nevertheless, in many cases seizures are managed quite effectively, including spontaneous remission without medication treatment.


Are any newly developing interventions predicted for the near future?


The Seaver Autism Center and several other centers around the US and abroad are working hard to understand more about the genetic basis of ASD and then using this knowledge to develop novel treatments. We have been actively recruiting patients for participation in the Autism Genome Project, and our Center has contributed hundreds of samples to this important initiative. Every affected patient and family is an opportunity to identify new causes of autism and to develop new treatments.

Our clinical team is working closely with basic scientists at the Seaver Center to develop treatments for Phelan McDermid Syndrome, a known cause of autism responsible for approximately 1-2 percent of cases. Drs. Joseph Buxbaum, Takeshi Sakurai, and Ozlem Bozdagi have created a mouse model with a deleted SHANK3 gene and found these mice to have significant deficits in the integrity of neuronal communication. These deficits negatively impact the process of what is called long term potentiation, a process that is known to underlie learning and memory. We are currently exploring the use of specific compounds to reverse the deficits in the mouse. We aim to bring such potential medications to clinical trials in children with Phelan McDermid Syndrome.

Using a similar strategy, Seaside Therapeutics discovered that a medication called arbcalofen reversed physiological and behavioral deficits in a mouse model of Fragile X Syndrome and preliminary trials in both Fragile X Syndrome and ASD were recently completed. Press releases from the company in September, 2010 reported positive effects in both trials and our Center has recently joined a national network in a multi-centered trial of arbaclofen to treat symptoms of social withdrawal in ASD. Arbacolofen is a GABA-B agonist that inhibits release of glutamate into the synapse and reduces postsynaptic glutamatergic neurotransmission. We are very pleased to be part of this important new direction of treatment research.

In addition, our center has been very invested in exploring the utility of oxytocin in autism. Dr. Jennifer Bartz is currently studying the effect of an oxytocin nasal spray on social cognition and emphathic accuracy in adults with autism and is also looking to better understand the neural circuitry underlying these deficits.

In terms of behavioral treatment, Drs. Latha Soorya and Ting Wang in our Center are working to help children improve social skills through a highly structured, manualized cognitive-behavior based curriculum designed to target non-verbal communication skills, emotion recognition, and perspective taking. Their team is also doing functional MRI before and after treatment to see which brain regions are involved in social cognition and whether brain activity patterns change with skill acquisition.


How can families best cope with the associated features of autism, such as: anxiety, aggression, and hyperactivity?


Families need to be prepared to deal with a wide variety of associated features of ASD. Treatment usually includes a combination of therapy and possibly medication options. Therapy may occur in group or individual settings and is typically focused on behavioral change using a cognitive behavioral or a behavioral model depending on the patient’s cognitive and functional abilities. With the help of a trained clinician, a functional behavior analysis can be useful to understand the antecedents to certain behaviors and associated symptoms and to develop targeted treatment plans to address the behavior. The specific intervention depends on the type of symptom that needs to be treated, but the impact of psychotherapy and behavioral treatment combined with changes in the structure of the environment can often have dramatic effects. At the same time, associated features like anxiety, aggression, and hyperactivity, also may occur in the context of developmental changes (e.g., puberty), and in these cases therapy is still important to pursue, but medication may also be considered.

Parents will benefit from education and training to help them learn strategies to best manage their child’s symptoms. Parents are usually the best advocate for their child and need to ensure that school and home-based interventions are implemented in the most consistent, structured, and predictable way possible across all settings where the child exists. Families can also benefit tremendously from the support of being with other families affected by autism. Autism Speaks ( is an excellent resource for community and support networks. In addition, support for siblings of affected individuals may be critical to explore, especially when associated features like aggression have a major impact on the entire family system.


What message would you like to leave with the readers of Autism Spectrum News about the hope for the future in our understanding and treatment of the conditions associated with autism?


This is an extremely exciting time in autism research. With respect to genetic advances, we are in a phase of exponential discovery where new genes that cause autism are being identified at a very rapid pace. Larger and larger samples of patients and families are becoming available for more and more advanced methods of genetic analyses. Each new gene identified teaches us more about the neurobiology of autism and creates important opportunities for model systems. Model systems can then be used to develop novel therapeutics based on targeted molecular approaches. At the same, early detection methods are improving so that the behavioral diagnosis can be reliably made as early as 12-18 months and array CGH is being recommended for all patients suspected of having an ASD. Finally, studies of behavioral interventions, like the Early Start Denver Model (Dawson et al, Pediatrics, 2010; 125(1):e17-23) have shown the robust impact of early intervention as increasingly obvious.


For more information about any of the studies mentioned or to find out more about the Seaver Autism Center for Research and Treatment at Mount Sinai School of Medicine, please call 212-241-0961 or go to our website at

Have a Comment?