NEW YORK (GenomeWeb) – Psychiatric pharmacogenomics firm CNSDose announced its entry into the commercial market during last week's Personalized Medicine World Conference, with plans to launch its depression medication dosing test this May.
The assay's architect, Ajeet Singh, a privately practicing psychiatrist and pharmacogenetics researcher, began his efforts to develop and bring CNSDose to market by demonstrating in a blinded, randomized trial that his proprietary polygene approach improved outcomes over the trial-and-error prescribing that characterizes standard psychiatric practice.
In that study, published last year in the journal Clinical Psychopharmacology and Neuroscience, researchers led by Singh compared genetically guided antidepressant prescribing using CNSDose to traditional unguided prescribing in 148 Caucasian patients with a principle DSM-5 diagnosis of major depressive disorder recruited in Victoria, Australia.
CNSDose incorporates multiple genes involved in both the metabolism of antidepressants in the liver and those that influence how readily the medications cross into the brain via the blood-brain-barrier.
In the trial, Singh and colleagues used a PCR and MALDI-TOF-based method to genotype the SNPs at the heart of the CNSDose prediction algorithm. They then generated reports on recommended dosing and shared them with the psychiatrists of half of the study participants.
The other half of the cohort also submitted a buccal swab for genetic testing but no results were returned. In this way, patients were blind to what treatment group they were in.
At the end of the 12-week study period, those whose doctors had access to the CNSDose report had more than double the symptom remission rate of the unguided group (72 percent versus 28 percent) based on assessments by an independent rater who was also blinded to which group the patients belonged.
The genetically-guided group also had reduced side effects and less need to take sick leave due to their depression, Singh and his coauthors reported.
As a company, CNSDose has since moved on to pursue a second replication study by independent academic researchers from the University of Melbourne, Deakin University, and the University of Toronto, with an expected publication date some time later this year, Singh told GenomeWeb.
The test joins several existing genetic assays to support more accurate drug dosing — some specific to psychiatric medications, and others covering a broad range of therapies and disease areas — in a market that has been somewhat capricious in its acceptance and endorsement.
PGx firm Genelex, for example, has had trouble convincing payors of its test's clinical utility and health economic value. PGx dosing in the cardiovascular space has also met hurdles in demonstrating clinical utility in the scientific literature that would support more widespread adoption.
At the same time, Assurex Health, another company focused on the psychiatric space, has been reimbursed by Medicare has for its GeneSight Psychotropic PGx test since it received a positive local coverage decision by contractor Palmetto GBA in 2014.
Assurex has also seen milestones in adoption by the US Department of Veterans Affairs, and among pharmacist groups. As of February this year, the company said it had tested a quarter million patients with a menu that has broadened over time to include guidance for dosing of depression, but also anxiety, bipolar disorder, pain medication, and ADHD treatments.
In some ways, psychiatry may be an easier field in which to prove the added value of pharmacogenomics than other disease areas, Singh told GenomeWeb.
Because current dosing strategies rely predominantly on trial and error, the burden of proof for a test in improving upon the status quo may not be as high, and also represents the possibility of immense benefit to patients, for whom any delay in achieving remission can mean significant suffering or other practical and personal deficits.
On the flip side, Singh said, the addition of genetics doesn't fundamentally alter patients' risk. "If you take a drug like Zoloft, which has a range of dosing from 20 mg to 450 mg, and at each dose you have to wait a month to see if it works, a whole year can go by. If you add in a genetic method, at worst it's just still the same trial and error," he said, "So we think we are likely to fall into a low-risk group in terms of, for example [the FDA's classification system.]"
In developing the CNSDose report, Singh has been relatively judicious in comparison to some other companies in the space, focusing narrowly on alterations in two sets of genes: those involved in metabolism of drugs in the liver, and others that play a role in transport across the blood brain barrier.
"We self describe as a second-generation pharmacogenomics report," Singh told GenomeWeb. "Whereas first-generation reports looked only at liver enzyme variants, critically we are looking at the blood-brain barrier as well … and ours is the only evidence-based product covering both of those that is ready to go to market."
In contrast, Assurex's GeneSight couples liver metabolism genes, as well as pharmacodynamics variants in serotonin transporter and receptor genes.
"We don't do any brain receptors, because my opinion is that the clinical utility of those is very thin," Singh said. "I think the brain is too complicated, I just try to understand the box it comes in, which means looking at pharmacokinetics where the science is much better."
CNSDose is also specific, at least in its initial version, to antidepressants — reporting on the 20 most commonly prescribed drugs, which make up about 95 percent of the market, Singh said.
Finally, CNSDose's testing model is also distinct. Rather than performing its own genetic testing of patient samples in house, the company plans to contract with outside labs that have that capacity, take in the resulting genetic data, conduct its own proprietary analysis, and then report back to ordering physicians.
"The idea is that we can provide clinical insight on DNA data. Labs globally can partner with us in exclusive or non-exclusive deals," Singh said. "They then take care of analytic validity and handshake with our HIPAA-compliant cloud. The trade secret is in the algorithm, and we spit out our report."
Currently, he said, the company has a single lab affiliate in Sydney. Successfully launching more widely later this year will depend on making agreements with a larger "ecosystem of labs."
In the report that CNSDose generates, a patient's genetic variants as measured by an affiliate lab are translated into a recommended low, medium or high dose for the 20 drugs the assay covers.
"We also stratify into different dose brands," Singh said at the PMWC meeting, "and recommend first the meds that appear to work well at an average dose." It makes sense to try those first, he added, before moving out to something that might require higher dosing.
During his presentation, Singh shared some data from the newer independent replication study the company hopes to publish this year. This study focused on prescribing of a single antidepressant, desvenlafaxine, rather than the full panel that CNSDose covers. According to Singh, this allowed researchers to calculate traditional sensitivity and specificity percentages, which he shared in his talk.
Based on the analysis of 119 subjects, he said, there was a strong correlation between what CNSDose predicted — either a low, medium, or high dose — with what was a clinically effective dose for patients. For the company's medium-dose recommendation, researchers calculated 92 percent sensitivity and 85 percent specificity. Interestingly, for the high-dose prediction, the specificity rose to 99 percent, something important for clinical practice where there may be hazards in recommending a higher dose for certain medications.
Currently in what Singh called a "beta launch," CNSDose plans to officially launch on May 8, concurrent with a presentation at the Royal Australian and New Zealand College of Psychiatrists Conference in Hong Kong.
While its test will initially be focused on antidepressants, Singh said the company's eventually plans to expand to other disease areas.
"What I am hoping we've got with our technology … if we have the hepatic and blood brain barrier keys, is that we should have utility in other medications [that act on the central nervous system]," Singh said. "Pain, anti-convulsives, dementia, ADD drugs … we are hopeful downstream we might actually have a platform technology with [what] we are trying to commercialize."