Amy Carpenter
December 2024—On the heels of the publication of a joint consensus recommendation on DPYD genotyping, setting out what variants to test for, two experts who have studied the use of DPYD genotyping shed light in CAP TODAY interviews on its importance and its real-world impact.
The consensus recommendation, issued by the Association for Molecular Pathology, American College of Medical Genetics and Genomics, the CAP, and other groups, is the latest in the AMP’s series that recommends a set of variants to include in clinical pharmacogenomics assays (Pratt VM, et al. J Mol Diagn. 2024;26[10]:851–863).
Testing for variants in the DPYD gene, the authors write, can help identify patients at risk of developing fluoropyrimidine toxicity who should receive reduced doses or avoid treatment with capecitabine and fluorouracil, as recommended by the guidelines of the Clinical Pharmacogenetics Implementation Consortium and Dutch Pharmacogenetics Working Group of the Royal Dutch Pharmacists Association.
The consensus recommendation answers one of the three main questions DPYD testing raises—what to test patients for, says Daniel L. Hertz, PharmD, PhD, assistant professor, Department of Clinical Pharmacy, University of Michigan College of Pharmacy. The other two questions, he says, are whether patients with cancer should be tested and how a patient with cancer who has a variant should be treated.
Patients who are DPYD variant carriers have a 70 percent risk of severe toxicity and a three percent risk of death, he says, if they receive the standard dose of fluoropyrimidine-based chemotherapy.
The authors of a prospective observational cohort study published this year found that patients with cancer who were tested pretreatment for five DPYD variants had reduced toxicities and hospitalizations compared with carriers who were not tested before initiating treatment, thus normalizing the risk to that of wild-type patients (Nguyen DG, et al. JCO Precis Oncol. 2024;8:e2300623). DPYD pretesting “should be considered standard practice,” the authors write.
The study was done at Atrium Health Levine Cancer, Charlotte, NC, which began to offer DPYD genotype testing in 2020.
The authors compared toxicity and hospitalization rates among pretreatment carriers, carriers who were not tested prior to starting treatment, and wild-type patients who received in-house DPYD testing across 14 cancer clinic locations. Of the 757 patients who received DPYD genotyping between March 2020 and May 2023 (implementation cohort), 442 patients were initiated on a fluoropyrimidine-based chemotherapy (outcomes cohort). The authors observed the outcomes cohort from March 2020 through December 2022 with three months of follow-up data. They found that of the 40 (5.3 percent) DPYD variant carriers for whom fluoropyrimidine chemotherapy had been initiated, 37 (93 percent) had the fluoropyrimidine dose reduced or discontinued after DPYD genotype results were known.
“We went from testing almost no one,” before the study, “to testing 65 to 70 percent of all the potential patients receiving a fluoropyrimidine-based chemotherapy,” says lead author Jai N. Patel, PharmD, associate professor of cancer biology, Wake Forest University School of Medicine, and Atrium Health Levine Cancer director of cancer pharmacology and pharmacogenomics. The goal is to reach 90 to 100 percent of such patients. “We now have greater buy-in from our providers because they see the data and that these patients exist,” Dr. Patel says.
DPYD genotyping was optional at the start, he says. Some physicians ordered the test and wanted the result pretreatment. Others did not order it until the first day of treatment or when toxicity was seen after starting treatment. “We were able to directly compare those two groups: variant carriers who started with a dose reduction and carriers who started at standard dose,” he says. In their study, the grade 3+ toxicity rate was not different in pretreatment carriers receiving fluoropyrimidine dose reductions (31 percent) compared with wild-type patients receiving standard dosing (30 percent). The grade 3+ toxicity rate in carriers who were not tested prior to starting treatment and received the initial full dose (64 percent) “was understandably lower than historical data (70%–75%) since nearly half had testing sent on the day of treatment start, and thus most received dose reductions in the second cycle, potentially avoiding severe toxicity,” the authors write. They found hospitalizations were reduced in carriers who received pretreatment genotype-guided dosing (25 percent) compared with carriers who were not tested prior to starting treatment (64 percent). Treatment delays due to toxicities were also highest in carriers who did not receive pretreatment testing.
The institution covered the cost of the optional testing in the first few years.
“We knew how bad the reimbursement landscape was and didn’t want that to be a barrier to at least offering the test,” Dr. Patel says. They started billing insurers for it in fall 2023. “We’ve noticed that probably 90 percent or so [of insurers] are authorizing the test despite the fact that many of these policies state that it’s not medically necessary.”
He and colleagues did an analysis of publicly available payer policies, not just for DPYD genetic testing but also for other pharmacogenetic tests. “We looked at about a dozen policies, focusing on the top payers, and pulled policies for eight commercial payers, three laboratory benefit managers, and the Centers for Medicare and Medicaid Services.”
Dr. Patel says they found “clear variability” among the payer policies in terms of whether DPYD testing is considered medically necessary. “Less than half of them state that DPYD testing is medically necessary, and more than half that it’s still investigational.”
When the medical policy states it is investigational, it is sometimes denied up front, he says, requiring peer-to-peer or appeal, and often it is reversed. But the delayed process means therapy might begin before approval is received. “And it’s pretty much useless at that point,” Dr. Patel says of the test’s benefit. Financial navigators submit a letter of medical necessity for every test. “I think that has helped because we’ve talked to other institutions that are not doing that, and they don’t see as high of an authorization rate.”
“It’s not foolproof,” however, he adds. “It’s not that 100 percent are being authorized.” Often the reasoning is that National Comprehensive Cancer Network guidelines have not endorsed the testing, he explains. “If and when the day comes that NCCN turns the corner and recommends testing, I think we’ll see almost unanimous coverage and adoption among medical oncologists who are not currently doing this.”
Some oncologists are concerned about reducing the dose very early in treatment as it could lead to underdosing, he acknowledges, “because there is a small handful of patients with DPD deficiency who theoretically could tolerate a full dose.” Published data show that dose reductions in those with DPD deficiency do not compromise survival among patients in the metastatic setting, he says, “though there are not as much data for those in the curative setting.”
Among academic centers, there is strong momentum, he says. “But probably over 90 percent of patients with cancer get treated in the community setting, so we have to do more education,” including of insurers. When coverage is denied and toxicity could likely have been avoided with a test, “I’ve gone back to the insurance company and told them. Those situations exist, and I’m sure they exist at several other places as well.” For community practices, education will have to include discussion of the available assays and how to integrate testing into the workflow.
Ideally, whether to test would be a shared decision between the patient and physician, Dr. Patel says, adding, “It should be the patient’s decision ultimately.”
Talking to the patient is what the Food and Drug Administration recommends, says Dr. Hertz of the University of Michigan.
“They have updated the package insert to say that clinicians should warn their patients about the risk for severe toxicity and discuss with them the opportunity for testing. Prior to that, the FDA didn’t say anything about talking to patients. They just said if you knew a patient was a carrier, then they were at risk.” It’s not the boxed warning he and colleagues wanted, but it “was an important step in the right direction,” he says.
“I think one of the big issues [for the FDA] is there are no FDA-cleared DPYD tests—everything is done with laboratory-developed tests.”
Another of the FDA’s reservations is something Dr. Hertz agrees with: There are some patients in whom testing would not be appropriate. One would be a patient who would choose full dosing even if they were found to carry a variant. Another would be a patient with certain disease characteristics who has to start treatment immediately and cannot wait five to seven days for a test result.
“It is reasonable to put in a drug label or in a guideline that you should test patients as long as it’s medically necessary or document in the medical record why testing was either opted out by the patient or deemed medically inappropriate,” Dr. Hertz says. “That will accomplish 98 percent of what we want, in that patients will be offered DPYD testing and a reasonable decision will be made.”
One barrier for the FDA could be the question of how variant carriers should be dosed. Here, CPIC has done the most work, Dr. Hertz says. The CPIC guidelines recommend a 50 percent up-front fluoropyrimidine dose reduction in patients harboring one no-function variant or one or two decreased-function variants (DPYD intermediate metabolizers) and avoiding fluoropyrimidine in patients carrying two no-function or one no-function and one decreased-function variant (DPYD poor metabolizers).
“I don’t know much about the politics or history, but my understanding is that the FDA has been reluctant to endorse CPIC guidelines and dosing,” Dr. Hertz says. “They could have looked at the evidence the same way everyone else has and said, ‘We think a 25 percent reduction in those variant carriers—those heterozygous patients—is a reasonable starting place, or a 50 percent reduction.’ Or they could have said there are guidelines to look at. But they stayed silent.”
Dr. Hertz credits the FDA pharmacogenetics team for being responsive, attending meetings, talking to him and others. “They have a certain idea of what kinds of studies should be done to generate personalized dosing for patients,” he says, and it comes down to this: “They haven’t seen the quality of evidence they usually use to generate dosing recommendations.”
In a 2023 article, Drs. Hertz and Patel and others call on the FDA to follow Europe’s lead in recommending DPYD testing and genotype-based dose adjustment to ensure patients with cancer receive safe and effective fluoropyrimidine chemotherapy (Hertz DL, et al. Clin Pharmacol Ther. 2023;114[4]:768–779).
Dr. Hertz says there is reason enough to recommend DPYD testing based on the clinical benefit demonstrated in prospective clinical studies. “And there is no strong evidence that patients have reduced treatment efficacy. No one has demonstrated that in a high-quality clinical study.”
Cost is always a consideration, and “the best studies show that testing is essentially cost neutral,” he says. And where there might be a cost concern, “It makes all the sense in the world” to include DPYD on other tumor genetic panels. “The marginal cost of adding that should be nothing from the tester’s side—there would be benefit with little to no downside.”
“So we have a demonstrated clinical benefit with no countervailing demonstrated risk, either clinical or financial, so the benefits clearly outweigh the risks.”
To be fair to the FDA, Dr. Hertz says he and colleagues have been told that updating drug labels and package inserts is a “challenging, cumbersome process and should probably not be used as the primary way by which genetic testing is recommended. They think this is where oncology guidelines should step in and recommend this testing.”
In a patient found to be a variant carrier, “we should dose-reduce to start and dose-escalate based on tolerability,” Dr. Hertz says, given that about 30 percent of patients who carry deleterious variants in the DPYD gene will not experience toxicity if treated with a full dose of fluoropyrimidine chemotherapy.
DPD deficiency is present in up to seven percent of the population and more prevalent in some patient populations than others. “The FDA claimed in their response to our petition that DPYD testing would exacerbate health inequities because the variants found in non-Europeans are not known,” Dr. Hertz says of a 2020 citizen petition submitted to the FDA by a patient advocate and signed by pharmacogenetics experts. “It’s actually the opposite, since the rates of DPD deficiency are higher in Black populations, and a high frequency variant, p.Y186C, is included on most testing. So testing should reduce that disparity.”
At Atrium Health Levine Cancer and Wake Forest Baptist Comprehensive Cancer Center, the plan now is to integrate DPYD testing into the chemotherapy order set, Dr. Patel says.
“The provider has to opt out of the test. Until now, the provider has had to opt in and order the test,” with the aid of best-practice advisory alerts.
Opt-out DPYD testing is the practice, too, at Dana-Farber Cancer Institute, where a patient died in 2021 from complications after the first cycle of single-agent capecitabine prescribed for recurrent, indolent breast cancer. The death was consistent with fluoropyrimidine toxicity as a consequence of DPD deficiency, though it wasn’t proved, authors of an article published this year wrote. They describe the opt-out program put in place since then (Jacobson JO, et al. JCO Oncol Pract. 2024;20[8]:1115–1122). “It’s unfortunate that that’s what it’s taking these days, but here we are,” Dr. Patel says.
The program’s design at Dana-Farber relies on process standardization and closed-loop reporting, and the lynchpin is an automated EHR-based alert for physicians who order a chemotherapy regimen containing 5-FU or capecitabine. If not performed previously, the testing is recommended and a link to the test order is provided. Providers who decide against testing must decline the recommendation to close the alert.
“This approach, in contrast to mandating testing, was selected because preemptive DPYD testing has a strong, but incomplete evidence base for which some providers do not believe routine testing is warranted,” the authors write.
Of the 1,043 patients preemptively tested for DPYD variants, 43 patients (4.1 percent) had high-risk results, all categorized as intermediate metabolizers. Among 25 evaluable patients tested preemptively, and for which results were available, 24 were treated with the CPIC-recommended dose reduction. An ordering error prevented the intended dose reduction in one case. Six patients required hospitalization—three for fluoropyrimidine toxicity and three for unrelated reasons.
DPYD testing increased from a baseline of 26.5 percent pre-intervention to a sustained rate of 90 percent after formal implementation, which took place largely in 2023. The authors conclude that preemptive DPYD testing is feasible and high rates can be achieved using an opt-out, reminder-based program.
Drs. Patel and Hertz and coauthors in 2023 surveyed institutions and clinicians known to do pretreatment DPYD testing, and the survey findings from 24 U.S. sites that have implemented the testing or have a plan in place were published this summer (Tracksdorf T, et al. Support Care Cancer. 2024;32[8]:497). Among the survey’s findings are the following:
- Thirty-three percent of sites ordered DPYD testing for all fluoropyrimidine-treated patients. At the remaining sites, patients were tested depending on disease characteristics (for example, cancer type) and clinician preference (for example, familiarity with the testing).
- Almost 50 percent of sites depend on individual clinicians to remember to order testing without the assistance of electronic alerts or workflow reminders.
- DPYD testing was performed most often in commercial laboratories that tested for at least four of the five DPYD variants considered actionable.
- About 90 percent of sites reported receiving results within 10 days of ordering.
- To cover the cost of pretreatment DPYD testing, 67 percent of sites reported billing for insurance reimbursement. Twenty-nine percent indicated the tests were covered by institutional or research funds. One site said patients pay out of pocket.
- Academic institutions/hospitals made up 70 percent of the sites that conduct pretreatment DPYD testing.
The authors conclude that routine testing is feasible, despite robust, institutionwide systems not always being in place to maximize its efficiency or effectiveness. They write: “With increasing evidence and awareness of its clinical utility, it is anticipated that the FDA and relevant professional guidelines will recommend pre-treatment DPYD testing in the near future. As such, institutions should begin to develop approaches to facilitate rapid and efficient implementation of DPYD testing to ensure patient safety and improve treatment outcomes in patients with cancer.”
Amy Carpenter is CAP TODAY senior editor.