Why is it so hard to find an effective drug treatment for autism?

In the UK, 700,000 people are on the autism spectrum, which makes it a very common neurodevelopmental disorder. In the last decade, a lot of progress has been made to address the needs of patients, and families now have more support available, but challenges remain.

Specialist education and behavioural programmes can really help many autistic people live full and happy lives. In some cases drugs can be prescribed to treat some of the symptoms or conditions associated with Autism Spectrum Disorder (ASD). There is no "cure" to make the condition disappear. The very idea of a cure can even be upsetting – autistic people and their relatives may see autism as a part of who they are, and not as something to be removed.

But even with good support, living with ASD can be challenging and this is why scientists have been investigating drug treatment options for years in order to come up with long-term alternatives for patients and their families.

Why is it so hard to find such a treatment? IBTimes UK takes a look at the obstacles researchers in this field face.

Mysterious causes

If you don't know what to treat, coming up with therapeutic options is going to be a challenge. This is perhaps the core problem with autism – the exact causes are still unclear and this hampers research efforts to find an effective, long-term treatment.

ASD is a complex condition, which emerges as a result of a combination of genetic and environmental factors. It is in fact believed to have strong, complex and heterogeneous genetic underpinnings with the phenotypic expression of these genetic components highly variable, from one individual to the next.

There has nevertheless been recent progress on an important front: the discovery of several monogenic forms of autism has provided researchers with novel tools to investigate drug treatment.

These are rare conditions, each accounting for less than 1% of the general ASD population, stemming from genetic anomalies not seen in large pools of control chromosomes.

"Geneticists have recently identified monogenic forms of autism. Although they only affect a very small number of patients, it has conceptually provided us with a new entry point to understand causes of the disorder and to find treatments that target these causes – rather than the symptoms. It has been a major breakthrough and since then there has been a lot more interest in drug development for ASD on the part of the industry," Dr Eva Loth, Sackler Lecturer in Translational Neurodevelopment at Kings College London, told IBTimes UK.

Finding monogenic forms of autism has also allowed researchers to use new methodological tools to test drugs – for instance they have been able to work more efficiently with animal models that present the rare genetic anomalies identified in patients.

Biological pathways rather than genes

The problem is that targeting these genes may only help these patients, with a rare form of autism, but not the 99% of other people with ASD.

For them, scientists are investigating a new approach. They believe different genes may converge on a smaller number of common molecular pathways, so they are trying to find treatments that target these pathways rather than specific genes.

"The hope is that we can come up with a treatment to target molecular pathways rather than the genes. Take synapse development and function in autistic patients for example – depending on the patient, different genes may be involved in causing problems, but they are all laying on the same pathway. So the idea would be to target this pathway to help the greatest number of patients possible", Loth says.

But scientists still find it hard to get positive results in clinical trials with this strategy. Research into fragile X syndrome, which shares phenotypic similarities with certain forms of autism, has shown that even using the best animal models could be problematic.

"Scientists investigating fragile X syndrome had a very interesting strategy, they created animal models, tested molecular abnormalities and behaviours. They had promising results at preclinical trials and phase one clinical trials, but for some reason later trials failed. The reason may be that something gets lost in the translation – what works for animals doesn't necessarily work for humans. There are some things that you cannot model entirely, for example there are limits to reproducing abnormal social behaviours in mice," Loth explains.

"Therefore, our colleagues have started working with patient-derived pluripotent stem cells, which may overcome some of the problems with translatability, as the neurons are created from the patient him or herself."

A road map for the future

The heterogeneity of the disorder, both at the genomic and phenotypic levels requires a new way to think about autistic people and to study them. An interesting idea is to divide patients into subgroups to study smaller numbers of them at a time, but with more characteristics in common.

"ASD is very heterogeneous. It could be that single genes have a different impact depending on the other different genes that a person has – his or her so-called genomic background. This means there are potentially many different subgroups of autism based on many different high-risk genes and genomic backgrounds. It is possible that some previous clinical trials failed because the treatment was only effective in, say 10 or 20 or 30% of the people. So we hope that in the future we will be able to identify different biological subtypes of autism, to predict which treatment works best for each," Loth says.

The road map for the future is clear: scientists will need to find these subgroups, investigating drug treatments that target biological pathways. For this, conducting larger-scale studies will be crucial.

This includes the EU-AIMS Longitudinal European Autism Project, a multicentre study carried out in seven European centres and co-led by Dr Loth and Prof Jan Buitelaar at Radboud University Medical Centre in Nijmegen.

Over 400 children, adolescents and adults with ASD and 300 individuals with typical development have been recruited to the study. Each volunteer is seen on two days and assessed in terms of their clinical profile, cognitive strengths and weaknesses, brain structure and function, and genomics.

"We need much larger studies with much larger samples because otherwise it will be complicated to create subgroups and study them. There also needs to be more data sharing between scientists to check that our findings are robust and can be reproduced across different labs," Loth concludes.