Scientists at the UC Davis MIND Institute used machine learning to identify several patterns of maternal autoantibodies highly associated with the diagnosis and severity of autism.

Their study (“Risk assessment analysis for maternal autoantibody-related autism (MAR-ASD): a subtype of autism”), published Molecular Psychiatry, specifically focused on maternal autoantibody-related autism spectrum disorder (MAR ASD), a condition accounting for around 20% of all autism cases.

“The incidence of autism spectrum disorder (ASD) has been rising, however ASD-risk biomarkers remain lacking. We previously identified the presence of maternal autoantibodies to fetal brain proteins specific to ASD, now termed maternal autoantibody-related (MAR) ASD. The current study aimed to create and validate a serological assay to identify ASD-specific maternal autoantibody patterns of reactivity against eight previously identified proteins (CRMP1, CRMP2, GDA, NSE, LDHA, LDHB, STIP1, and YBOX) that are highly expressed in developing brain, and determine the relationship of these reactivity patterns with ASD outcome severity,” the investigators wrote.

“We used plasma from mothers of children diagnosed with ASD (n = 450) and from typically developing children (TD, n = 342) to develop an ELISA test for each of the protein antigens. We then determined patterns of reactivity a highly significant association with ASD, and discovered several patterns that were ASD-specific (18% in the training set and 10% in the validation set vs. 0% TD). The three main patterns associated with MAR ASD are CRMP1 + GDA (ASD% = 4.2 vs. TD% = 0, OR 31.04, p = <0.0001), CRMP1 + CRMP2 (ASD% = 3.6 vs. TD% = 0, OR 26.08, p = 0.0005), and NSE + STIP1 (ASD% = 3.1 vs. TD% = 0, OR 22.82, p = 0.0001).

“Additionally, we found that maternal autoantibody reactivity to CRMP1 significantly increases the odds of a child having a higher Autism Diagnostic Observation Schedule (ADOS) severity score (OR 2.3; 95% CI: 1.358–3.987, p = 0.0021). This is the first report that uses machine learning subgroup discovery to identify with 100% accuracy MAR ASD-specific patterns as potential biomarkers of risk for a subset of up to 18% of ASD cases in this study population.”

“The implications from this study are tremendous,” said Judy Van de Water, PhD, a professor of rheumatology, allergy, and clinical immunology at UC Davis and the lead author of the study. “It’s the first time that machine learning has been used to identify with 100% accuracy MAR ASD-specific patterns as potential biomarkers of ASD risk.”

Previously, Van de Water found that a pregnant mother’s autoantibodies can react with her growing fetus’ brain and alter its development.

The research team obtained plasma samples from mothers enrolled in the CHARGE study. They analyzed the samples from 450 mothers of children with autism and 342 mothers of typically developing children, also from CHARGE, to detect reactivity to eight different proteins that are abundant in the fetal brain. They then used a machine learning algorithm to determine which autoantibody patterns were specifically associated with a diagnosis of ASD.

The researchers created and validated a test to identify ASD-specific maternal autoantibody patterns of reactivity against eight proteins highly expressed in the developing brain.

“The big deal about this particular study is that we created a new, very translatable test for future clinical use,” explained Van de Water, adding that this maternal blood test uses an ELISA platform, which is quick and accurate.

The machine learning program crunched roughly 10,000 patterns and identified three top patterns associated with MAR ASD: CRMP1+GDA, CRMP1+CRMP2, and NSE+STIP1.

“For example, if the mother has autoantibodies to CRIMP1 and GDA (the most common pattern), her odds of having a child with autism is 31 times greater than the general population, based on this current dataset. That’s huge,” said Van de Water. “There’s very little out there that is going to give you that type of risk assessment.”

Researchers also found that reactivity to CRMP1 in any of the top patterns significantly increases the odds of a child having more severe autism.

Van de Water noted that with these maternal biomarkers, there are possibilities for early diagnosis of MAR autism and more effective behavioral intervention. The study opens the door for more research on potential pre-conception testing, particularly useful for high-risk women older than 35 or who have already given birth to a child with autism, she said.

“We can envision that a woman could have a blood test for these antibodies prior to getting pregnant. If she had them, she’d know she would be at very high risk of having a child with autism. If not, she has a 43% lower chance of having a child with autism as MAR autism is ruled out,” according to Van de Water, who is currently researching the pathologic effects of maternal autoantibodies using animal models.

“We will also use these animal models to develop therapeutic strategies to block the maternal autoantibodies from the fetus,” she said.