Dr. Richard Altman considered the chart of his patient, David Clay, and also his personal experience as Clay’s primary care physician at UCHealth Lone Tree Medical Center. Over the past few months, Clay had been experiencing more nighttime anxiety. Altman considered the citalopram (Celexa) dose he had previously prescribed and decided there was room to boost it.
As Altman keyed in the prescription order, a message popped up in UCHealth’s electronic health record. The message informed Altman that Clay has a version of a gene called CYP2C19. That polymorphism – in automotive terms, same model, but with an automatic transmission versus a stick shift – meant that Clay’s liver probably metabolized citalopram much faster than an average patient. That, in turn, meant that he would need a higher-than-average dose for the drug to have its normal effect. Altman was going to up the dose anyway, but the popup crystallized for both doctor and patient why prescribing a higher dose made sense.
This was a single alert, related to a single patient, in a single UCHealth clinic room one day in April 2020. It was also an early return on a major long-term investment by UCHealth and the University of Colorado School of Medicine aimed at harnessing genetic information to develop new treatments and improve day-to-day patient care.
That investment, announced in 2014, created what is today the Colorado Center for Personalized Medicine (CCPM) and its crown-jewel biobank that holds 85,000 biospecimens and counting.
Every UCHealth patient is offered the option to contribute through My Health Connection, and many already have. So far, 185,000 UCHealth patients have consented to contribute to the biobank. The 100,000-patient gap between consents and donors exists because biobank specimens are collected as part of an in-clinic blood draw the patient needs anyway. But some never do have blood draws in clinic – ophthalmology patients, for example. With others, those blood draws are rare. The CCPM is ramping up a campaign to collect saliva (an alternative source of genetic material) to close that gap.
Clay was among those who signed up, and a few years back, a routine blood draw yielded another sample for the Biobank’s ultracold freezers.
“I think the doc just asked if I would do it, because, at some point, it may help somebody,” he recalled.
The big idea behind personalized medicine is to feed artificial intelligence a buffet of genetic information from huge numbers of patients, have that AI tease out connections between genes and diseases, and then use those insights to develop therapies tailored to people with particular genetic characteristics. That work is underway, and it will take time.
In the meantime, the CCPM is focusing on pharmacogenetics, which has to do with how our genes influence how we metabolize drugs. The CCPM started with CYP2C19.
One might think that a gene that encodes a liver enzyme responsible for breaking down roughly 10% of all human pharmaceuticals would have earned a name more compelling than CYP2C19, but so it is. Among the drugs it metabolizes include proton-pump inhibitors for stomach ulcers, antidepressants, anticonvulsants, sleep aids, antimalarials, and drugs that inhibit viral replication. A different version of the gene can either slow down or speed up metabolism.
In Clay’s case, he consented and provided a biospecimen. That specimen was analyzed by a million-dollar Illumina gene sequencer. Citalopram was among the 10 CYP2C19-influenced drugs that the CCPM chose as initial targets for its feedback loop back to patients. The CCPM worked with UCHealth’s Epic team to create an alert that popped up at just the right time so as not to distract or even annoy physicians who have faced a growing number of alerts with the transition to electronic health records. Those alerts then popped up for patients with polymorphisms that either sped up drug breakdown (as in Clay’s case) or slowed it down (meaning a standard dose would become a larger dose than intended).
Those alerts haven’t happened often: 69 times in the past 18 months or so. But it’s a start, and Altman, for one, believes it’s a positive sign for the future of medicine. “Precision-medicine pharmacogenetics” may sound like buzzword gobbledygook, but it can help answer pressing exam-room questions, Altman says.
“’Why is my patient not responding to treatment?’ Or, ‘Why is my patient having such a serious reaction when I gave him a small dose?’” said Altman. “It’s about having insight into a huge new factor that influences care, like gender, or age, or smoking status.
In Clay’s case, that popup provided peace of mind as to the reason for needing the higher dose. The medication wasn’t failing; he was just processing it faster than most. And medication doses matter, Altman says, whether it’s about improving cholesterol numbers, lowering blood pressure, or improving mood.
“From the doctor’s perspective, you want to give the right dose – not more, not less – to get what the patient needs out of it,” he said.
The CCPM isn’t stopping with CYP2C19. Next up in its pharmacogenetics initiative is a gene called SLCO1B1. Certain polymorphisms of this gene can cause muscle pain or worse in patients on certain statins – somewhere between 7% and 29% of them can experience what’s known as statin-associated muscle symptoms (SAMS). An electronic-health-record popup related to statins could save patients a lot of pain.
The Center is also looking to add variations of the gene DPYD to its feedback loop. Certain flavors of that gene can slow down the breakdown of common chemotherapy drugs, resulting in higher risk of toxicity among cancer patients in treatment. Those investigations and others will soon benefit from a new generation of microarray chips that will boost sequencing throughput by a factor of seven.
The most important throughput for the biobank, though, is that of UCHealth patients who agree to contribute to the advance of medical science. Many patients want to contribute to the science as a way to contribute to the collective good and show appreciation for the care they have received.
Some do benefit, of course – David Clay among them.
“In my case, I did it because it might help somebody – and it came back and helped me,” he said.
To learn more or to sign up for the biobank at the Colorado Center for Personalized Medicine Biobank, visit https://www.uchealth.org/services/biobank/.