An Asian woman sitting at a table and smiling at a woman healthcare professional sitting next to her

Pharmacogenomics (PGx) has been widely used in many medical disciplines, including oncology, where it helps determine if specific treatments may be toxic to patients or if they will work at all.

While PGx has been found useful in certain clinical management, several factors need to be addressed before PGx can become the standard of care for every cancer patient in this country.

We spoke with Peter H. O'Donnell, MD, associate professor of medicine at the University of Chicago, about how PGx has already changed oncology, and how it will transform therapeutic practice in the future.

What are the most exciting areas where PGx is becoming standard of care for cancer medication management and improving patient outcomes?

There is a global debate right now around preemptive PGx testing for variants of the DPYD gene that can guide treatment with fluoropyrimidines, used in the treatment of solid tumors such as colorectal, pancreatic, gastric, breast, and head and neck cancers. There is widespread recognition that patients who carry variants in DPYD may be at higher toxicity risk, and that upfront dose-reduction in such individuals decreases toxicity, including potentially fatal toxicities. The attention around this has been a positive step in the march towards more personalized care.

However, I must admit that I am biased about this topic. I believe that PGx is important for safety, protecting patients, and using medications in a way that takes advantage of all the tools that we have available to us in 2023. That includes genomic information about the patient.

But I can see the other side of this debate, especially in this field where the bar is high for adoption. I’m an oncologist, and all oncologists want to ensure they’re doing everything they can to keep their patient alive and ensure survival. The hesitance around DPYD testing is well-intentioned because we want to make sure that we’re not rushing into this and sacrificing cure.

We also need to make sure we’re doing DPYD testing in a way that acknowledges equity and diversity. It’s clear that not all test panels for DPYD are the same, and that could create disparities and lead to patients receiving results that are falsely reassuring. Those shortcomings need to be considered.

How should this testing better acknowledge patient diversity?

I’m of the opinion that there are five clinically actionable variants in DPYD that we should be considering. If a test panel, for example, only includes two of those five variants, and a patient doesn’t have either of those two, then a report will say that the patient is a normal metabolizer when, in fact, it may be that we’re just missing another variant they could be carrying.

How is this an equity issue? Of those five variants, there’s some debate around the fifth variant, often referred to as the African variant, which was discovered in people of African descent. Our institution believes that it’s important to include this variant in DPYD testing. Otherwise, we could be creating disparities if we’re only testing for variants identified in people of European descent.

What else in this field is exciting to you when it comes to PGx and cancer care?

Right now, DPYD is the hottest topic, but there are other important applications of PGx in this field, too. We also currently support testing for the UGT1A1 gene for irinotecan, another drug used for solid tumors. Variation in that gene has been known for a long time, but testing for it has had spotty adoption.

I also think there’s a huge opportunity in cancer care to think about the role of PGx for pain medication use. Such a huge part of cancer care is making sure we’re dealing with a patient’s pain. The use of PGx for managing pain is something our institution is studying right now.

Finally, cytotoxic chemotherapy, while still a mainstay for many kinds of cancers, is considered one of the older types of therapies now. Immunotherapies have, of course, revolutionized the care of cancer and the practice of oncology. I would say that PGx is in the nascent stage around immunotherapies, with a lot of promise.

What are the primary barriers to adoption of PGx in cancer care?

There is a high bar in oncology for the adoption of PGx. It’s usually used for some kind of upfront dose reduction or avoidance of therapy. For some cancers, there is no other alternative treatment. What would an oncologist do if their patient has high risk of toxicity, and yet this is the only therapy that exists to treat a patient’s disease?

The second is one that’s particularly relevant to clinical laboratorians: turnaround time. Patients often need to start their cancer therapy quickly. It’s not uncommon for a patient to start within a week after the clinical consultation. How many labs can turn around a PGx test in that span of time?

Finally, we face a knowledge gap for providers. It’s not a criticism of providers, but PGx is something that most practicing physicians—let alone practicing oncologists—were not taught in medical school. They are learning it on the fly. They’re following their associations’ guidelines around this, and professional associations have been slow to embrace it.

In addition, it is important for institutions that are using PGx to provide clinical decision support at the point of care. They can’t just put a ‘star allele’ result in a patient’s chart and expect a clinician to know what to do with that information.

And yet, while oncologists—just like all providers—may feel underprepared or under-educated in the germline PGx space, I would argue that, at the same time, oncologists are among the best prepared for considering such information compared to other clinicians. That’s because oncologists are already used to interpreting genomic information—on the somatic side. They have a lot of experience thinking about it. Oncologists could potentially be “leading-edge” adopters, or champions, of similarly considering germline PGx information during prescribing.

What are the drivers for PGx adoption?

First, academic institutions are often the leaders in these efforts, which I think will filter down to community practices. Second, patients and patient advocates are demanding PGx as part of safe and effective prescribing. That’s really important. We must listen to the patients.

Third, improvements in technology are making this something where it’s hard to find a reason not to do it. It’s become very affordable. The ability to genotype is old technology at this point, so we don’t encounter the same kinds of technological barriers we did 15 years ago. Those have been solved.

Going back to barriers, though, one still left is, how do we make this accessible wherever the patient shows up? We’re seeing a lot of siloing of this information within an institution that performs this testing. If, however, a patient shows up at an ER down the street, how do we design the system so that patients have ownership of their information, and it travels with them?

Do you see opportunities in the expansion of molecular testing to accelerate PGx?

Every oncologist runs next-generation somatic tests on their patients. If it’s become the standard of care to test the somatic genome, why wouldn’t we just complement that with the germline genome? We could bundle test orders and have the somatic and germline done with one test. In 2023, technology is certainly there to do that.

Where can oncologists and clinical laboratorians collaborate?

I see an opportunity in the development of clinical decision support. Most clinical laboratories are used to generating molecular reports that are too long, too dense, and unreadable for the average clinician.

So, we need to make them more interpretable, and we need to provide solid decision support—or interpretations—that accompany those raw lab results.

For example, we could design and write these reports in a way that helps with clinical decisions. Or even better, put the information within the clinical workflows, rather than in a ‘report.’ That allows the information to be pushed to the clinician—at the moment when the prescription is being considered¬≠—rather than sought within a difficult-to-find report among a larger array of results. That would promote adoption.

Where do you think PGx will be in 20 years?

My dream is that for every patient who walks into a doctor’s office or hospital, we’re scanning a QR code, or using some other technologic or electronic means of utilizing that information, which has already been obtained, maybe at birth, and making that accessible at the point of care. That way, no prescription is written without germline PGx information being considered. I believe we can one day get there. 

Jen A. Miller is a freelance journalist who lives in Audubon, New Jersey. +Twitter: @byJenAMiller