NEW YORK (GenomeWeb) – Q BioMed, a firm that acquires and develops what it calls "undervalued biomedical assets," is developing a molecularly guided therapeutic for pediatric minimally verbal autism (PMVA) and will for orphan drug designation with the US Food and Drug Administration and the European Medicines Agency (EMA) this month.
The New York-based company said it has discovered two microRNA biomarkers associated with expressed genes and pathways responsible for speech development, and it aims to complete a validation study on the biomarkers with undisclosed academic partners by the end of the year. Q BioMed is using the biomarkers to help identify clinical trial patients for its drug, internally named QBM-001, and hopes they can eventually be used as diagnostic and prognostic indicators.
A subset of autism spectrum disorder (ASD), PMVA occurs in toddlers that have minimal or delayed verbal communication. Clinicians diagnose the condition in patients because of its link to deficits in social perception, social judgement, and social interaction skills. Children with non-verbal or minimal verbal disorders are described as clumsy, uncoordinated, and having low scores in motor performance tests.
"This is definitely a neglected patient population, which is important to diagnose as early as possible," Q BioMed CEO Denis Corin explained. "RNA testing … will help us look at the pathology that's affecting the kids, [and] in the long term ensure the right kids are diagnosed and properly treated."
According to Robert Derham, Q BioMed's VP of orphan products, researchers have tried analyzing different biomarkers for autism in blood, urine, and other liquid samples. While several genetic complications have been linked to non-verbal autism, researchers currently cannot explain the root cause of PMVA.
Derham explained that Q BioMed began its biomarker discovery study in 2016, which was designed to stratify patient populations into different autism subsets. Collaborating with several academic centers including the State University of New York Upstate Medical University, Derham's team recruited about 500 young individuals who had varying levels of autistic behavior. The age of participants the team selected for the study ranged from 18 months to 7 years.
Acknowledging that normal sample collection in autistic children can be difficult, Derham and his team wanted to use a method that was minimally invasive and barely noticeable to patients.
"When you want to get a blood or other sample, it's like trying to pull the fingernails off someone," Derham said. "Instead, we realized that working with salivary glands and saliva cells yielded the amount of sample we wanted."
Clinicians used a swab brush tool to collect cells from patients' salivary glands, placed the samples in collection tubes with ribonuclease-deactivating elements, and sent the tubes to SUNY Upstate's lab for analysis.
Researchers then ran RT-PCR on the sample's miRNA to measure the DNA for over 1900 potential biomarkers. Derham's team then applied a proprietary algorithm developed by an undisclosed commercial partner to produce a heat map to visualize the miRNA sequence variation among the patients.
Derham explained that his team originally identified four miRNA biomarkers that affected specific biochemical pathways that the researchers were targeting, as well as pathways that correlated with manifested phenotypes occurring in several of the patients' daily routines and behaviors.
However, Derham noted that two of the biomarkers had better profiles for potential diagnostic and prognostic purposes.
"The biomarkers gave us insight to other problems and pathways that aren't distinct and [that] we hadn't clued in on before," Derham said. "It also allowed us to potentially develop follow-up therapies that could be more effective for the children."
Derham argued that salivary gland samples provide data that is much more stable over time than data collected from blood samples, which often have cell levels that can fluctuate daily.
"Even the autism community will admit that these children have been ignored, due to the challenge of working with them, since they're extremely sensitive to environmental change and have high aggression levels," Derham explained. "By performing a quick swab, it is easier for clinics, doctors, parents, and caregivers to help identify what type of autism the child might have."
The researchers dealt with issues such as analyzing miRNA from patients who had a non-verbal IQ of 60 or below. Derham explained that by using one identified biomarker, the group found a small amount of overlap between patients with IQs of 60 and below and those with an IQ of 60-85.
However, the team saw that the overlap disappeared when it looked at participants with an IQ of 50 or below. When parsing the data after the study, Derham said that the researchers found that lowering the threshold left too few children in a proposed "50 IQ and below" group for the data to be useful; at the same time, the threshold produced a group that did not overlap at all with other groups in the study and projected better predictability of above 90 percent for a validation study.
Another limitation the team encountered included dealing with patients that were still teething. Samples from these individuals are often contaminated with blood, potentially leading to testing errors. Therefore, the researchers excluded patients younger than 18 months, which is typically when children have stopped teething.
Derham said that the discovery study was not powered for predictability and diagnostic capabilities; instead, the group is addressing this in the validation study. He noted that Q BioMed will publish the current study's clinical data once it is completed and the company receives US patents related to the biomarkers' diagnostic and therapeutic capabilities.
Q BioMed now aims to complete its validation study with undisclosed partners by the end of the year. According to Derham, the team is working with 10 academic centers as partners, consultants, and potential study sites.
The researchers will need to use samples from a new group of autistic children for the study, as the discovery cohort subjects do not fall into the required age range for testing.
Q BioMed is also requesting a pre-Investigational new drug (IND) meeting with the FDA and EMA while performing the ongoing validation study, anticipating that it will file the IND submissions by December. The firm then plans to launch a yearlong clinical study in early 2020 to test QBM-001 on a minimally verbal sub-population of patients.
According to Derham, his team is also considering evaluating an additional biomarker and will also compare PMVA patients to other subsets of ASD patients.
Other companies and research groups are also looking to establish autism-linked biomarkers to guide downstream therapy. Moleculera is currently developing a panel of autoimmune biomarkers that it believes may be useful to identify autism patients likely to respond to intravenous immunoglobin therapy. Oxford BioDynamics said in November that it had partnered with Italian research institute Casa Sollievo della Sofferenza to identify blood-based biomarkers of ASD.
Meanwhile, researchers at Stanford University, the New York Genome Center, and the University of California, Los Angeles collaborated as part of the Hartwell Autism Research and Technology Initiative (IHART) to develop an open repository of phenotypic, metabolomic, proteomics, and genomic data of individuals living with ASD.
While groups have targeted autism as a general disease, Derham emphasized that several genetic and behavior differences exist between ASD subpopulations. Therefore, he argued that treatment types that work well for certain patients may fail in other subpopulations.
Derham said that Q BioMed previously found common metabolic pathways that were disrupted because of genetic mutations in PMVA or non-verbal patients, as well as during the discovery study. The company is therefore focusing on populations that it believes will respond to QBM-001.
"If more companies spent time developing a companion diagnostic like us, we'd be able to weed out the smaller percent of patients that don't function well on the downstream therapeutic," Derham explained. "While you'd never be able to treat said patients when you reach the market, it would make clinical trials faster, smaller, and have a higher success rate."
Joseph Buxbaum, a psychiatry professor at Mount Sinai's Icahn School of Medicine, argued that Q BioMed's claims of biomarker usage in diagnostic and therapeutic applications raise multiple concerns.
"The world of psychiatry and neurology is littered with biomarkers that did not pan out and a very healthy degree of skepticism is warranted for any claim for novel biomarkers," Buxbaum, who is unaffiliated with the study, explained. "Unfortunately, there is simply insufficient data … to evaluate whether this biomarker will prove useful."
Buxbaum highlighted that that Q BioMed's statement about assessing over 1,900 biomarkers in biospecimens from about 500 patients in the study will eventually lead to patient discrimination. He also argued that because researchers know that autism is highly heterogeneous, "no compelling evidence exists for different biological bases for more severe or less severe manifestations of autism." He believes that the issue then raises the question of whether changes in a given biomarker reflect the cause or consequences of the disorder.
"It will be interesting to see how these biomarkers pan out in independent samples, comprised of a more representative mix of the population," Buxbaum said.
Derham agreed that researchers have struggled to identify links between exact biomarkers and ASD because there have been over a thousand genes implicated in the disease. At the same time, he argued that most researchers in the field believe autism biomarkers exist, highlighted by large biomarker discovery projects such as the Autism Biomarker Consortium.
"Autism research is in the same stage of infancy cancer was thought to be in the late 70's, as we're only just parsing out patients into more subsets, such as fragile X Syndrome, PTEN, and other syndromes that fall under the spectrum that are mutation-related diseases," Derham explained. "There is very little research with PMVA, which has made a biomarker quest a little more challenging, but very rewarding from the feedback and encouragement of the autism community."