Reduction of increased repetitive self-grooming in ASD mouse model by metabotropic 5 glutamate receptor antagonism; randomized controlled trial of early start denver model


▪ Reduction of increased repetitive self-grooming in ASD mouse model by metabotropic 5 glutamate receptor antagonism [Silverman, Tolu, Barkan, & Crawley, 2009]

Animal models of ASD are often developed for the purpose of determining the pathophysiology of ASD. Several models have been created and are in the process of being refined in which a human variant is “knocked-in” to the mouse. Another approach is to identify inbred mouse strains that may have behaviors consistent with behaviors typically found in ASD. Previously, the BTBR inbred mouse strain had been demonstrated to have decreased exploration of novel mice compared to novel objects in a three-chamber test of sociability. In this paradigm, a mouse is placed in a center chamber and is free to remain in the center chamber or explore a novel mouse in one chamber or a novel object in the opposite chamber. As replicated in the paper being reviewed, the BTBR mouse did not spend more time in the chamber with the novel mouse compared to the novel object, as was the case for the control C57BL/6J mouse in this and previous papers. In addition, this paper replicated previous findings of increased self-grooming in the BTBR mouse compared to the C57BL/6J mouse.

The objective of the paper under discussion was to test a pharmacological intervention, metabotropic glutamate receptor 5 (mGluR5) antagonism, by the compound 2-methyl-6-phenylethylnyl-pyridine (MPEP). Over a decade of study of the mechanism of fragile X syndrome had led to the hypothesis that mGluR5 antagonism was a possible treatment for signaling abnormalities affecting synaptic plasticity considered to be key to learning. MPEP had previously been demonstrated to ameliorate some components of the Fmr1 mouse knock-in phenotype, including synaptic plasticity, spine morphology, hyperactivity, seizures, prepulse inhibition deficits and hyperactivity. In addition, MPEP had improved fruitfly courtship behavior in the fruitfly model of fragile X syndrome. The manuscript under review does an excellent job of reviewing the relevant literature leading to the choice of MPEP administration to the BTBR mice. Review of what is known about pharmacological treatment of ASD led to the choice of risperidone as a contrasting pharmacological challenge.

The investigators studied a range of doses of MPEP and risperidone in three behavioral assays in BTBR and C57BL/6J mice: (1) self-grooming; (2) open field locomotion; (3) three-chamber sociability. MPEP significantly reduced elevated self-grooming of the BTBR mice with no effect on the lower level of self-grooming of the C57BL/6J mice. MPEP did not reduce overall locomotion, showing that the reduction in self-grooming was not due to a non-specific sedative effect. In contrast, risperidone reduced self-grooming behavior, overall locomotion, and time spent with either the novel mouse or novel object (spending more time in the chamber into which it was initially placed). The effect of risperidone was consistent with sedation, as expected from administration of a major tranquilizer.

Trials of mGluR5 antagonists are currently underway in fragile X syndrome and mGluR5 is being considered for trials in ASD [Berry-Kravis et al., 2009]. The somewhat unexpected finding of reduced self-grooming in the study by Silverman and colleagues being reviewed suggests the importance of including outcome measures for restricted and repetitive behaviors as well as social and learning measures in the clinical trials of mGluR5 antagonists. Given the specificity of drug effects from animal models through human clinical trials of ASD, it is suggested that clinical trials of ASD consider the possibility of specific effects on domains of function and dysfunction in ASD rather than expecting/measuring global responses of all ASD symptom domains. Finally, both animal studies such as this one, and clinical trials of individuals with ASD, will need to consider an interaction between pharmacological effects of increasing learning and models to teach more appropriate social behavior. In terms of the paper under review, it would be interesting to administer an mGluR5 antagonist and see if it affects training to increase species typical social behavior. In the case of clinical trials involving individuals with ASD, one would plan for a positive pharmacological response to facilitate, but not replace, social learning.

▪ Randomized controlled trial of Early Start Denver Model [Dawson et al., 2010]

There have been relatively few randomized controlled trials of early intensive intervention in ASD. As the authors point out, early screening by 18 months has been recommended by the American Academy of Pediatrics. Therefore, it is surprising that there were no randomized controlled trials of early intervention in ASD of toddlers less than 30 months old.

A detailed intervention manual and curriculum were previously developed for the Early Start Denver Model (ESDM). ESDM uses a developmentally informed curriculum that emphasizes interpersonal sharing in the context of positive affect, emphasizes adult sensitivity to the child with ASD, and focuses on both verbal and non-verbal communication development. Behavioral principles of ABA treatment are incorporated. In this study, ESDM treatment was guided by the treatment manual using a highly trained university-based clinical research team. Two 2 hr sessions per day for 5 days per week were provided for 24 months or until the 48-month-old birthday. In addition, parents were trained to use similar principles at home and reported spending an average of 16.3 hr/week using ESDM strategies and another 5.2 hr/week in other therapies (e.g. speech therapy).

The investigators randomized 30 toddlers with ratio IQ greater than or equal to 35, who were between 18 and 30 months old with autistic disorder or PDD-NOS, to either ESDM or referral to early intervention providers and resources in the community. In addition to being a randomized, controlled trial, the authors employed the next essential procedure of using blinded raters for assessments. A final important component of the trial was that there were two primary outcomes defined at the beginning of the study, the first being the Mullens Scales of Early Learning (MSEL) Early Learning Composite standard score and the second being the Vineland Adaptive Behavior Scales (VABS) composite standard score. The MSEL was the only fully blinded outcome measure, since the VABS is a parental report measure and parents were not blind to treatment assignment. Secondary measures included subscale scores of the MSEL and Vineland, DSM-IV diagnosis, Autism Diagnostic Observation Schedule (ADOS) severity scores, and the Repetitive Behavior Scales-Revised (RBS-R) total score.

The primary and fully blinded outcome measure showed a striking improvement from mean standard MSEL Early Learning Composite standard score of 61 to 79 for the ESDM group, which was significantly greater than the improvement from 59 to 66 for the control group. The second primary outcome measure, the VABS adaptive behavior composite, showed a change from 70 to 69 in the ESDM group, which was significantly less than the decrease in VABS in the control group from 70 to 59.

Given the significant effect of primary outcome measures, the analysis of the subscale scores of MSEL and VABS provide information about what domains of function were affected by ESDM relative to controls. MSEL receptive and expressive language, but not MSEL visual reception, improved significantly more in the ESDM than control group. For the VABS, Communication, Daily Living Skills, and Motor skills showed less decline in the ESDM group compared to controls, but change in VABS Socialization score was not different between the groups. Consistent with the absence of difference in VABS socialization scores, there was not a significant difference in change in ADOS severity scores. RBS-R total scores also did not have a significant change after ESDM treatment relative to control intervention.

More children after ESDM intervention changed diagnosis from DSM-IV autistic disorder to PDD-NOS compared to controls. However, it is unclear which DSM-IV domains showed improvement. Given the significant improvement in communication rated by both MSEL and VABS, it would have been useful to know whether the categorical shifts in diagnosis were due to changes in the communication domain or also related to the other two domains. Otherwise, what emerges from the study is a pattern of improvement in communication, daily living, and motor skills without a change in socialization or restricted and repetitive behavior. Obviously, this does not decrease the importance of improving communication and adaptive behavior in ASD. However, it leads to the question about whether the ESDM has specificity for ASD treatment or whether this is a study of the importance of the principles (including intensity of treatment) of early intervention for improving function for others with developmental disorders without ASD.

The study is of interest in relationship to a two-hit hypothesis for ASD. The first hit has been considered to be an internal developmental neurobiological vulnerability that leads an infant with ASD to be less intuitively attuned to nonverbal and verbal communication. This leads to the paradoxical second hit of “environmental deprivation” in the context of an enriched environment. The ESDM used in the study incorporated principles of early intervention, which facilitate the learning of the child. It is surprising that this did not lead to more definite improvement in social function, but the intuitive social function either may be more fundamentally driven by the original developmental challenge, or may show improvement later, or may be improving but the measures may be insensitive to change. Although the ADOS severity scores did not change, it would have been interesting to report social items that might be expected to be targeted by the intervention, and whether they changed. Otherwise, an absence of response in restricted and repetitive behaviors may have masked an improvement in social function. In addition, although this was a large, if not the largest sample of early intervention, the power to detect change was limited in this study in which there were only 24 toddlers in each group.

There are two fundamental take home messages in this paper: (1) a methodologically rigorous study of early intervention (specifically ESDM) demonstrated statistically significant and clinically meaningful improvement in language as measured by the MSEL; (2) not withstanding questions about possibly modest improvements in social functioning and repetitive behavior, the paper highlights the small effect size of a model early intervention trial on improvement in social function and reduction in repetitive behavior. The first point is extremely important and has important implications for translation and dissemination of existing treatment, since communication is fundamental to further development. The second point is a reminder that much is left to be done in research to develop even more effective early and later intervention strategies, as well as to understand fundamental neurobiological processes, which may be reducing the capacity for benefitting from these interventions.

This leads to a convergence of the two papers being reviewed in this issue. It may be that combination of mGluR5 antagonists (or other agents that are yet to be developed to improve learning in ASD) and intensive teaching strategies such as ESDM may be more effective than either approach alone.