After Fatal Toxicity, Questions Over DPD Testing Before Chemo

Pam Harrison

March 13, 2019

Some cancer patients react very badly to commonly used chemotherapies — but there is a test that can identify those patients.

So why isn't this test being used?

This is the question currently being asked in a courtroom, where the families of four patients who died are now suing the health authorities because testing was not carried out before treatment was initiated. The argument is that such testing could have potentially prevented these deaths.

This scenario comes from France and concerns testing for deficiency of the dihydropyrimidine dehydrogenase (DPD) enzyme or for certain polymorphisms of the DPYD gene that encodes for that enzyme.

A partial or total absence of this enzyme allows 5-fluorouracil (5-FU) and the oral prodrug capecitabine (Xeloda, Genentech) to build up to toxic and sometimes life-threatening levels. Both of these chemotherapies are very widely used in the treatment of colorectal cancer, as well as in breast and other cancers.

French lawyers are arguing that DPD testing could have been done prior to treatment initiation to identify patients at risk for an overdose reaction — so why weren't these patients tested?

In fact, in 2018, the French Group on Clinical Pharmacology in Oncology recommended that physicians screen all patients for potential DPD deficiency to "red flag" those at potential risk for severe 5-FU toxicity.

While the litigation in France continues, another fatal case was highlighted by the media in the United Kingdom.

European experts have openly questioned whether DPD testing should be incorporated into oncology practice in order to safeguard patients from the possible lethal consequences of fluoropyrimidine treatment.

However, in pondering this issue, two experts in the United Kingdome came to opposing conclusions. In his Medscape video commentary, David Kerr, MD, professor of cancer medicine at the Oxford Cancer and Haematology Centre in Oxford, England, argues for germline testing for DPD deficiency, but in his video commentary, Karol Sikora, MD, professor of cancer medicine at the University of Buckingham argues against such testing.

With the US medical community not immune to litigation, Medscape Medical News wanted to find out what oncology experts in the United States thought of the idea of screening every patient for DPD deficiency prior to initiating 5-FU/capecitabine.

The three US experts approached were against routine DPD testing, and they explained why.

"The problem is that there are about 300,000 cases in the US each year where 5-FU or capecitabine is used up front," said Robert Diasio, MD, director, Mayo Clinic Cancer Center, Rochester, Minnesota. Currently, the genotype test in use today can detect four separate mutations in the DPYD gene known to be associated with DPD deficiency.

International guidelines advocate for the measurement of these four mutations, Diasio noted — and in France, Germany, and the Netherlands, the guidelines are apparently followed.

"The problem is, these mutations only account for about 5% of patients who are really DPD deficient," Diasio said.

In decades of his own research into the DPYD gene, Diasio and colleagues have identified between 70 to 80 other genetic variants in the gene that are also associated with decreased DPD activity and that can also predispose patients to severe 5-FU toxicity.

So by genotyping patients for these four mutations, "we could prevent some patients from getting a severe toxic reaction but if none of these mutations are present, that still doesn't mean you aren't going to get sick," Diasio said.

Patients can be tested for DPD deficiency using a functional enzymatic assay that Mayo Clinic researchers have also developed.

"However, this is a rather cumbersome assay and it's not very user friendly," Diasio admitted.

There also wouldn't be enough time or laboratories in the United States to begin to test all prospective 5-FU/capecitabine patients using their functional assay, he added.

"The beautiful thing about genotyping is that it is just so easy to do in terms of sequencing and it's become so inexpensive that it would be really very appealing to have a genotype test," Diasio said.

"So the genotypic test is absolutely the way to go and I do believe in it, but we're just not there yet," he concluded.

At-Risk Patients

Estimates of how many people are actually at risk for severe 5-FU/capecitabine toxicity vary widely, but it appears that deaths due to DPD deficiency occur in about 0.5% of patients who are treated with 5-FU, according to the National Institutes of Health.

Alan Venook, MD, the Madden Family distinguished professor of medical oncology and translational research, University of California, San Francisco, said he has almost never encountered cases such as these.

"I've been practicing oncology for a very long time and I've seen one case [of DPD deficiency] in my entire career," Venook observed.

"So the likelihood of being DPD deficient is very low, and you just can't be doing these tests in my opinion for something that is so infrequently seen as this is," he added. (The Food and Drug Administration for the moment agrees with Venook as it does not require physicians to prescreen patients for DPD deficiency prior to 5-FU/capecitabine treatment.)

Venook also points out that 5-FU is now given by continuous infusion, and this appreciably lessens the risk of serious toxicity because peak drug levels are not as high as they are when the drug is given as a bolus. (More on accidental overdose below.)

Even in a litigious society like the United States, "the incidence of truly life-threatening 5-FU toxicity is so small, it's just not worth screening," Venook reiterated.

"And no, we don't worry about litigation because it's just so rare," he added.

Another expert suggested that there may be a place for genotype testing for patients needing 5-FU/capecitabine but perhaps done in a broader context.

"The rational for doing this testing is of course that if you found an abnormality, it would significantly alter your treatment plans," pointed out John Marshall, MD, director Otto J. Ruesch Center for the cure for gastrointestinal cancer, Georgetown University, Washington, DC.

"But the fact is that [DPD deficiency] is a very rare event so the culture of incorporating this testing is just not here," he added.

Even if there were a more comprehensive genotype test for DPD deficiency, it's questionable if many physicians would order it.

For example, physicians are encouraged to prescreen patients about to receive irinotecan (Camptosar, Pfizer) for reduced UGT1A1 activity. Reduced UGT1A1 deficiency increases the risk of patients developing severe neutropenia following initiation of irinotecan.

While a laboratory test is available to determine the UGT1A1 status of patients, "no one does that test either or very few people do, even when there is a label telling them that they should," Marshall said.

What Marshall does foresee is a time when broad molecular profiling is done on all tumors to identify rare events — not just "bits and pieces" for individual-decision-making, but broad pharmacogenomic testing so that physicians would know which drugs they need to avoid and which they should favor as they move forward through various treatment algorithms.

"We have a very busy cancer center and we hand out lots and lots of 5-FU and probably every 2 to 3 years, there is a patient where this does happen," Marshall said.

"So we do warn people, it's on the consent form that patients have to sign but DPD deficiency testing is not considered standard practice in the United States," he commented.

Accidental Overdose

As noted by oncologist Wen Wee Ma, MD, also at the Mayo Clinic in Rochester, Minnesota, not only is the current test for known variants of the DPYD gene less than perfect, but DPYD mutations are actually present in only about half of patients who develop severe 5-FU toxicity.

Other risk factors for excess toxicity include mutations in other genes, impaired renal function, and older age, he pointed out.

Moreover, potentially catastrophic side effects of 5-FU, though again extremely rare, can occur in patients who are not DPD deficient.

There is also now an antidote, uridine triacetate (Vistogard, Wellstat Therapeutics), that can treat excess 5-FU toxicity.

However, unless the timing of the toxicity coincides with the pharmacology of the drug, the antidote doesn't work, Ma explained.

Ma and colleagues tested the antidote in a clinical trial presented at the 2016 Gastrointestinal Cancers Symposium, which was reported by Medscape Medical News at that time.

That study enrolled 135 patients who were at elevated risk for 5-FU toxicity — 111 because of overdose or accidental capecitabine ingestion and 24 because of DPD deficiency and/or rapid onset of severe toxicities. They were treated with uridine triacetate under emergency or expanded access protocols, and the drug was provided by the manufacturer.

Almost all of the patients (130/135, 96%) recovered within 30 days of being treated with the antidote. One-third recovered well enough to resume chemotherapy in 30 days.

Mortality was 4%, which is dramatically decreased from the 95% mortality that is seen historically for patients who have had an overexposure or who are susceptible to severe side effects of 5-FU, Ma noted.

However, Ma emphasized that the antidote must be used quickly. Unless patients are treated within 96 hours of emergent symptoms, the drug doesn't work, he stressed.

No Testing in US

So for now, it would seem that in the United States, there is no routine DPD testing, and the experts approached did not foresee that there would or should be such testing.

While oncologists might want to remain vigilant when they first initiate treatment with 5-FU/capecitabine, and look out for signs of toxicity, to demand everybody be prescreened for toxicity risk is simply too much work for too little gain, or so the experts that talked to Medscape Medical News have suggested.

Diasio, Venook, and Ma have disclosed no relevant financial relationships. Marshall reports that he has served as a speaker or as a member of a speaker's bureau for Genentech, Amgen, Bayer, Celgene, Merck, and Caris Life Sciences D.

For more Medscape Oncology news, join us on  Facebook  and  Twitter.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....