Day 2 of the Autism Brain Research meeting began far too early for me. Due to a combination of jetlag and blogging Day 1, I only managed a couple of hours sleep. So both my attention and my note-taking followed a U-shaped trajectory: coffee-fuelled and enthusiastic in the morning; present in body but not mind during the middle sessions; and then a second wind at the end, inspired by David Amaral's rousing closing speech. More on that later.
Cognition and development trajectories:
The morning session was home turf - cognition!
Tony Charman kicked off, highlighting again the heterogeneity within autism and pointing out that, even in biological subtypes of autism such as Fragile X, there is huge variability in clinical presentation. So the heterogeneity problem isn't solved by biology alone. He noted that, during the 1990s, many researchers (particularly in the UK) were optimistic that there might be a common cognitive explanation for autism, even if there were multiple genetic and neurobiological pathways leading to the cognitive deficit. It seems fairly clear now that this isn't the case. Nevertheless, cognitive research is still vital, acting as a potential signpost to different etiological mechanisms and potentially explaining the behaviours that characterise autism and suggesting useful intervention strategies. However, cognitive autism research is typically hampered by a combination of small sample sizes, lexclusion of individuals with low IQ, and narrow focus on only one cognitive domain at a time.
Charman then presented data from three recent studies from the SNAP cohort project, which involved a sample of over 70 adolescents with autism who each completed a battery of 58 cognitive tests, allowing potential subgroups to be pulled out (see the Jones et al. paper on auditory discrimination that I blogged about last week). He concluded by suggesting that a combination of approaches to cognition would be needed, akin to the two different approaches to autism genetics (candidate gene versus genome-wide screening) that were discussed yesterday. There wasn't much to disagree with here, although personally I think we can have the best of both worlds if we're thoughtful enough in the design of our studies.
Tony Charman kicked off, highlighting again the heterogeneity within autism and pointing out that, even in biological subtypes of autism such as Fragile X, there is huge variability in clinical presentation. So the heterogeneity problem isn't solved by biology alone. He noted that, during the 1990s, many researchers (particularly in the UK) were optimistic that there might be a common cognitive explanation for autism, even if there were multiple genetic and neurobiological pathways leading to the cognitive deficit. It seems fairly clear now that this isn't the case. Nevertheless, cognitive research is still vital, acting as a potential signpost to different etiological mechanisms and potentially explaining the behaviours that characterise autism and suggesting useful intervention strategies. However, cognitive autism research is typically hampered by a combination of small sample sizes, lexclusion of individuals with low IQ, and narrow focus on only one cognitive domain at a time.
Charman then presented data from three recent studies from the SNAP cohort project, which involved a sample of over 70 adolescents with autism who each completed a battery of 58 cognitive tests, allowing potential subgroups to be pulled out (see the Jones et al. paper on auditory discrimination that I blogged about last week). He concluded by suggesting that a combination of approaches to cognition would be needed, akin to the two different approaches to autism genetics (candidate gene versus genome-wide screening) that were discussed yesterday. There wasn't much to disagree with here, although personally I think we can have the best of both worlds if we're thoughtful enough in the design of our studies.
Kasia Chawarska then presented work using eyetracking techniques to investigate face-processing in young children with autism. One such study found that autistic children were quicker to disengage their attention from a picture of a face in order to saccade towards a new object. Another investigated recognition memory by measuring whether the child had a preference for looking at novel versus previously viewed faces. Interestingly the autistic children showing the poorest recognition looked less at the eyes and more at the mouth. Dr Chawarska argued that there may be different subtypes of autism related to face processing skills. This certainly fits with work my PhD student, Ellie Wilson, did with older children, and again highlights the importance of considering individual variation.
To round off the session, Cathy Lord spoke about longitudinal studies using the ADOS repeatedly between 10 and 40 months. Using some fancy statistics (which I didn't quite follow) she was able to split the group up according to their trajectories. What was most noticeable about her data, however, was how incredibly noisy it was, with children's ADOS scores varying wildly from session to session. Because she had multiple data points, this probably wasn't too much of a concern for the study. But the lack of stability worries me a lot given that we are all supposed to be using the ADOS if we want to get our autism research published. Lord also presented similar data from older children, again showing diverging trajectories in different subgroups*.
Behavioural interventions:
The second session focused on behavioural interventions. There was some interesting stuff, although I can only remember one presenter even mentioning anything vaguely brain-related. So it wasn't clear how any of the research linked up with the rest of the meeting.
Fred Frankel reported on a program to enhance children's friendship-making skills, but whilst he was able to show improvements on a number of social measures, there was no control group, so it wasn't clear whether the program was causing the improvement.
Aubyn Stahmer echoed the general theme of heterogeneity in autism and discussed different ways of optimising and adapting interventions programs for different children as they develop.
Judith Reaven presented data from a randomized control trial of "Face Your Fears" - a CBT group therapy treatment for anxieties. This involved gradual exposure to anxiety-provoking stimuli, self-rating of anxiety before and after exposure, and talking through the situation. There was evidence for efficacy in young kids and preliminary support for a modified version for teenagers.
Finally, Sally Rogers provided an overview of four studies published this year looking at various interventions that, in different ways, involved training parents to manage their interactions with their autistic child. There was some evidence for changes in the children's attention to other people, although no improvements on standardised measures of language beyond the (fairly large) improvements also shown in the control groups. In fact, the improvements in language shown by the untreated control children was perhaps the most remarkable finding, highlighting the need for properly controlled studies before claims of treatment efficacy can be made.
Pharmacological interventions:
This session also provided mixed evidence for successful interventions. The general theme seemed to be that non-core symptoms can be treated pharmacologically, but with potentially serious side effects. And there's no evidence at this point in time for interventions targeting the diagnostic features of autism.
By this point in the afternoon, I was really struggling to stay with it, so I need to put a massive disclaimer here: what follows may well not be a fair representation of the three presentations. In any case, this is a research blog, and none of the information here should be taken as clinical advice.
Christopher McDougle reported on the use of antipsychotics as a treatment for 'irritability'. I haven't heard this term used clinically before, but apparently the FDA define it as "symptoms of aggression towards others, deliberate self-injury, tantrums, and quickly changing moods". Traditional antipsychotics such as haloperidol that block dopamine pathways have proven successful at reducing 'irritability' but have severe side-effects in dystonia and dyskinesia. Second generation antipsychotics that affect serotonin pathways are also successful compared with placebo but lead to severe weight gain in most cases. These side effects are obviously a particular concern if the drugs are being given to children.
Bryan King talked about the overlap between autism and ADHD, pointing out that, while an autism / PDD diagnosis precludes a diagnosis of ADHD, the patterns of ADHD symptoms shown by a large subgroup of kids with autism are almost identical to non-autistic ADHD kids. Drug trials of methylphenidate have shown some success, but slightly lower response rates than for ADHD without autism. King also reported a study of citalopram (an SSRI) for repetitive behaviours in autism, which showed no benefits beyond placebo.
Craig Erickson closed the session by highlighting the challenges for developing pharmacological autism interventions: autism presents a moving developmental target; there's little agreement on appropriate outcomes measures; placebo responses can be large; and (again) there is huge heterogeneity in autism. I think the same could be said of all autism interventions. Erickson then gave a run-down of various drug treatments that have either been shown not to work or show promise but lack compelling evidence as yet. In this latter category he included ongoing work on oxytocin; therapies combining pharmacological interventions with social skills training; and drug therapies derived from animal models of Fragile X syndrome.
Future directions:
David Amaral gave the final talk entitled "Autism research: The promise and the pitfalls". The promise part was really a summary of the preceding presentations, and the pitfalls were highlighting some of the themes emerging from the meeting: the inadequate supply of post-mortem tissue; the need to recognise heterogeneity and the developmental nature of autism; and the "plethora of misleading information" available out there in the real world. Amaral also urged researchers to keep an open mind about possible causes of autism, not to take anything off the table, and to be prepared for surprising results.
To illustrate some of these points, Amaral presented two strands of research from his own lab. The first related to MRI studies of the amygdala. The available cross-sectional data suggests early overgrowth, such that the amygdala are at their final adult size by around eight years in autism, compared with 18 years or so in typical development. Looking at much younger kids showed that the amygdala are significantly larger than in controls by 3 years of age. However, there appeared to be three subgroups within the autism cohort: kids with rapid amygdala growth but normal total cerebral volume; kids with the opposite pattern (enlarged total cerebral volume but normal amygdala growth); and kids for whom both amygdala and total cerebral volume were normal.
Amaral also presented some fascinating work on autoimmune function in autism. In one study, his team looked for autoantibodies in the blood samples of kids with autism and then used them to stain monkey brains. They identified some unusual antibodies that targeted Golgi cells in the cerebellum. In another study, his group detected IgG antibodies in the blood of a small subgroup of mothers of autistic children. These were then injected into pregnant rhesus monkeys. The offspring showed no robust changes in social behaviour, but did show evidence of increased hyperactivity and whole body stereotypies.
Conclusions:
It's been a fascinating couple of days, although I was disappointed there wasn't more emphasis on brain imaging, electrophysiology, or links between brain and cognition. Partly because those are the areas I'm most interested in personally, but also because this research may help bridge the huge gulf that currently exists between the genetic / neurotransmitter level research and the intervention side of things. I'm also looking forward to hearing people talk about "ortism" again rather than "artism"!
The main issue that was hammered home again and again is the heterogeneity that exists within autism. It seems that real progress can only be made, both in terms of understanding the causes of autism and designing and optimising intervention strategies, once we address this problem. But things at last seem to be moving in that direction.
NB: Again, please note that this summary is based on the notes I made during the conference. I'm not an expert in most of these areas and there may be some inaccuracies, in which case I'll be happy to correct any that are pointed out to me. You can email me on jon.brock@mq.edu.au or just leave a comment at the bottom of this post. I'll try and get round to adding some links, but need to go souvenir shopping before I catch my plane!
Behavioural interventions:
The second session focused on behavioural interventions. There was some interesting stuff, although I can only remember one presenter even mentioning anything vaguely brain-related. So it wasn't clear how any of the research linked up with the rest of the meeting.
Fred Frankel reported on a program to enhance children's friendship-making skills, but whilst he was able to show improvements on a number of social measures, there was no control group, so it wasn't clear whether the program was causing the improvement.
Aubyn Stahmer echoed the general theme of heterogeneity in autism and discussed different ways of optimising and adapting interventions programs for different children as they develop.
Judith Reaven presented data from a randomized control trial of "Face Your Fears" - a CBT group therapy treatment for anxieties. This involved gradual exposure to anxiety-provoking stimuli, self-rating of anxiety before and after exposure, and talking through the situation. There was evidence for efficacy in young kids and preliminary support for a modified version for teenagers.
Finally, Sally Rogers provided an overview of four studies published this year looking at various interventions that, in different ways, involved training parents to manage their interactions with their autistic child. There was some evidence for changes in the children's attention to other people, although no improvements on standardised measures of language beyond the (fairly large) improvements also shown in the control groups. In fact, the improvements in language shown by the untreated control children was perhaps the most remarkable finding, highlighting the need for properly controlled studies before claims of treatment efficacy can be made.
Pharmacological interventions:
This session also provided mixed evidence for successful interventions. The general theme seemed to be that non-core symptoms can be treated pharmacologically, but with potentially serious side effects. And there's no evidence at this point in time for interventions targeting the diagnostic features of autism.
By this point in the afternoon, I was really struggling to stay with it, so I need to put a massive disclaimer here: what follows may well not be a fair representation of the three presentations. In any case, this is a research blog, and none of the information here should be taken as clinical advice.
Christopher McDougle reported on the use of antipsychotics as a treatment for 'irritability'. I haven't heard this term used clinically before, but apparently the FDA define it as "symptoms of aggression towards others, deliberate self-injury, tantrums, and quickly changing moods". Traditional antipsychotics such as haloperidol that block dopamine pathways have proven successful at reducing 'irritability' but have severe side-effects in dystonia and dyskinesia. Second generation antipsychotics that affect serotonin pathways are also successful compared with placebo but lead to severe weight gain in most cases. These side effects are obviously a particular concern if the drugs are being given to children.
Bryan King talked about the overlap between autism and ADHD, pointing out that, while an autism / PDD diagnosis precludes a diagnosis of ADHD, the patterns of ADHD symptoms shown by a large subgroup of kids with autism are almost identical to non-autistic ADHD kids. Drug trials of methylphenidate have shown some success, but slightly lower response rates than for ADHD without autism. King also reported a study of citalopram (an SSRI) for repetitive behaviours in autism, which showed no benefits beyond placebo.
Craig Erickson closed the session by highlighting the challenges for developing pharmacological autism interventions: autism presents a moving developmental target; there's little agreement on appropriate outcomes measures; placebo responses can be large; and (again) there is huge heterogeneity in autism. I think the same could be said of all autism interventions. Erickson then gave a run-down of various drug treatments that have either been shown not to work or show promise but lack compelling evidence as yet. In this latter category he included ongoing work on oxytocin; therapies combining pharmacological interventions with social skills training; and drug therapies derived from animal models of Fragile X syndrome.
Future directions:
David Amaral gave the final talk entitled "Autism research: The promise and the pitfalls". The promise part was really a summary of the preceding presentations, and the pitfalls were highlighting some of the themes emerging from the meeting: the inadequate supply of post-mortem tissue; the need to recognise heterogeneity and the developmental nature of autism; and the "plethora of misleading information" available out there in the real world. Amaral also urged researchers to keep an open mind about possible causes of autism, not to take anything off the table, and to be prepared for surprising results.
To illustrate some of these points, Amaral presented two strands of research from his own lab. The first related to MRI studies of the amygdala. The available cross-sectional data suggests early overgrowth, such that the amygdala are at their final adult size by around eight years in autism, compared with 18 years or so in typical development. Looking at much younger kids showed that the amygdala are significantly larger than in controls by 3 years of age. However, there appeared to be three subgroups within the autism cohort: kids with rapid amygdala growth but normal total cerebral volume; kids with the opposite pattern (enlarged total cerebral volume but normal amygdala growth); and kids for whom both amygdala and total cerebral volume were normal.
Amaral also presented some fascinating work on autoimmune function in autism. In one study, his team looked for autoantibodies in the blood samples of kids with autism and then used them to stain monkey brains. They identified some unusual antibodies that targeted Golgi cells in the cerebellum. In another study, his group detected IgG antibodies in the blood of a small subgroup of mothers of autistic children. These were then injected into pregnant rhesus monkeys. The offspring showed no robust changes in social behaviour, but did show evidence of increased hyperactivity and whole body stereotypies.
Conclusions:
It's been a fascinating couple of days, although I was disappointed there wasn't more emphasis on brain imaging, electrophysiology, or links between brain and cognition. Partly because those are the areas I'm most interested in personally, but also because this research may help bridge the huge gulf that currently exists between the genetic / neurotransmitter level research and the intervention side of things. I'm also looking forward to hearing people talk about "ortism" again rather than "artism"!
The main issue that was hammered home again and again is the heterogeneity that exists within autism. It seems that real progress can only be made, both in terms of understanding the causes of autism and designing and optimising intervention strategies, once we address this problem. But things at last seem to be moving in that direction.
NB: Again, please note that this summary is based on the notes I made during the conference. I'm not an expert in most of these areas and there may be some inaccuracies, in which case I'll be happy to correct any that are pointed out to me. You can email me on jon.brock@mq.edu.au or just leave a comment at the bottom of this post. I'll try and get round to adding some links, but need to go souvenir shopping before I catch my plane!