Thursday, 5 June 2014

Corticosteroid therapy in regressive autism

I note that the recent paper by Frank Duffy and colleagues [1] (open-access here) seems to be receiving quite a bit of interest, and their suggestion that corticosteroid therapy might be something to look at more scientifically when it comes to some cases of regressive autism. The fact that the Duffy paper also got it's own commentary [2] made me think that this is something I should be discussing on this blog.

So, a few details about the Duffy paper bearing in mind it is open-access:

  • Well, first and foremost this was a chart review study - a retrospective case-control study - which in the authors own words aimed "to identify by retrospective data analysis, quantifiable evidence supportive of beneficial effects of adrenal corticosteroid (prednisolone) therapy on brain, language and behavioral function of children with a history of sudden autistic regression (R-ASD)". Regression, as regular readers might know, is something which crops up on this blog quite a bit (see here).
  • Researchers looked at 20 steroid treated participants with regressive autism (STAR) and 24 untreated participants diagnosed with an autism spectrum disorder (ASD) aged between 3 and 5 years old. This untreated group comprised of 7 children with a history of developmental regression and 17 without (collectively labelled NSA). The Developmental Neurophysiology Laboratory (DNL) held at Boston Children's Hospital was the source material for participants and their data, which: "maintains an archived comprehensive database of several thousand previously studied clinical patients and research subjects". Although STAR participants were all characterised by "having autism with a historically documented period of regression at onset" authors were specifically looking at "those patients, who were clinically treated with corticosteroids (STAR group) subsequent to an initial neurophysiological study that showed an absent or distorted FMAER [4 Hz frequency modulated evoked response], and who all had a second neurophysiological study after the treatment period was concluded, i.e. after at least six and not more than 36 months".
  • Results: well, there were quite a few of them but perhaps most important was the suggestion that following a steroid treatment period of between 4 and 14 months (mean = 9.1 months) "STAR group subjects’ language ratings were significantly improved and more STAR than NSA group subjects showed significant language improvement". Further: "the steroid treatment appears associated with a significant increase in the specific FMAER stimulation elicited 4 Hz response amplitude of the superior temporal gyri (STG) in both hemispheres of the study children with regressive autism". I note also that the authors also reported that: "The scaled DSM-IV symptom summary scores demonstrated a highly significant improvement after steroid treatment".
  • Authors were also keen to see whether the positive changes noted in the steroid treatment group might have just been spontaneous events part and parcel of the nature of some regression in autism (see a previous post touching on this issue). By comparing the regressive and non-regressive participants in the NSA control group they concluded: "spontaneous NSA group improvement of language in the regressive autism subpopulation was not observed". So at least some support for active treatment having an effect over and above just a natural tendency for improvement in language abilities.
  • The monitoring of side-effects should always be an important part of any intervention for autism or anything else. Duffy et al do talk about the side-effects from the use of steroid treatment and for example report that: "almost all steroid-treated subjects gained weight". Other adverse effects were also reported including two participants who "experienced behavioral regression after steroid taper". Just looking at some of the other side-effects noted from prednisolone use (see here) this is not something entered into lightly.

I've not done justice to all the details included in the Duffy paper particularly with their focus on "an innovative EEG/ERP biomarker strategy monitoring the frequency modulated auditory evoked response (FMAER)". The commentary from Golla & Sweeney [2] provides more details about the hows and whys of this method; something I would not pretend to even know about. Suffice to say that they list it as a "promising, objective diagnostic tool for detecting sensory processing abnormalities in the auditory cortex of functional significance, and a potential biomarker for identifying subgroups of patients for novel treatments and for tracking effects of novel therapies".

That being said, I'm sure quite a few people who were drawn to this paper were perhaps more eager to look at the intervention itself and the results reported potentially impacting on language and behavioural scores in their chosen cohort. I don't want to get too excited with the results at the moment because of the nature of the Duffy study methodology (various study limitations are listed by the authors in their paper) and the strong requirement for more "formal prospective randomized, double blind, placebo controlled, crossover study".

The question of how and why oral prednisolone "administered by the parents on a daily basis at 2mg/kg/day" might impact on the presentation of autism is also a question not yet answered. It would be easy to say that the more typical use of such a steroid as a way of preventing or reducing some of the compounds involved in the process of inflammation might be part and parcel of any effect. Certainly there is quite a bit of evidence of inflammatory processes and markers to be potentially related to cases of autism (see here) although I don't think Duffy et al actually looked at anything like this in their report. When used for specific ailments such as the inflammatory bowel diseases like Crohn's disease one might also assume that some of the work describing similar issues to be present in cases of autism might be relevant too (see here). The autoimmune aspect to steroid use is another possible angle in light of a growing interest in autoimmunity and autism (see here) but again, much more research is needed to tease out any association and importantly, any mechanisms linking physiological changes from treatment with behavioural changes.

I might also highlight again that prednisolone is a corticosteroid, a synthetic drug not a million miles away from steroid hormones such as cortisol, the primary stress hormone. Readers who might have read my recent post on elevated amniotic fluid steroid hormones being potentially associated with autism (see here) might remember the sentiments of the paper by Simon Baron-Cohen and colleagues [3]. I don't want to make too many connections where none may exist but certainly it strikes me as unusual that a drug designed to elevate steroids might show some advantage on a condition potentially [partially] brought about by elevated steroid levels. Does this for example, imply that 'bathing' in elevated levels of steroid hormones during the nine months that make us might have some effect on receptor programming [4] for example?

I'll be keeping my eyes open for any further work on this topic.


[1] Duffy FH. et al. Corticosteroid therapy in regressive autism: a retrospective study of effects on the Frequency Modulated Auditory Evoked Response (FMAER), language, and behavior. BMC Neurol. 2014 May 15;14(1):70.

[2] Golla S. & Sweeney JA. Corticosteroid therapy in regressive autism: Preliminary findings from a retrospective study. BMC Medicine 2014, 12:79

[3] Baron-Cohen S. et al. Elevated fetal steroidogenic activity in autism.  Molecular Psychiatry. 2014. June 3.

[4] Matthews SG. Antenatal glucocorticoids and programming of the developing CNS. Pediatric Research. 2000; 47: 291–300.

---------- Duffy FH, Shankardass A, McAnulty GB, Eksioglu YZ, Coulter D, Rotenberg A, & Als H (2014). Corticosteroid therapy in regressive autism: a retrospective study of effects on the Frequency Modulated Auditory Evoked Response (FMAER), language, and behavior. BMC neurology, 14 (1) PMID: 24885033