FOOD WITH ORAL CANCER DRUGS?

Are oral oncology drugs being mislabeled? Maybe.
Most oral cancer drugs stipulate that they should be taken in a fasting state, but some actually have better bioavailability when taken with food. This is particularly true for the new prostate cancer drug abiraterone acetate (Zytiga), which has a greater interaction with food than any other marketed drug, yet it is labeled to be taken when fasting.
These points are highlighted in a comment published in the October 20 issue of the Journal of Clinical Oncology by Mark J. Ratain, MD, professor of medicine and director of the Center for Personalized Therapeutics at the University of Chicago, Illinois.
"My major concern is safety," Dr. Ratain told Medscape Medical News. "Very few patients would anticipate that the consequences of taking their medications with food...would be an effective overdose."
In the case of abiraterone acetate, the dose of the drug could be increased 10-fold by eating breakfast half an hour after taking the drug, he writes.
Dr. Ratain notes that food interactions are described in the labeling of many drugs, but are notably absent in the labeling of new oral cancer agents.
Most drugs can be taken either with or without food, he explains, because the effects of food are not clinically significant to drug absorption. There are cases in which absorption is decreased by food; generally, the labeling on those drugs instructs that they be taken on an empty stomach. Similarly, labeling on drugs for which absorption is enhanced by food typically instructs that they be ingested with meals.
"However, a recent study of oral agents approved by the US Food and Drug Administration [FDA] in the last 10 years demonstrated that oral oncology drugs have generally been labeled fasting, despite food effects as large as a 4-fold increase in bioavailability," Dr. Ratain notes.
Bias in Labeling?
This apparent bias for fasting labeling in cancer drugs is exemplified with abiraterone, he notes. Abiraterone, which was approved earlier this year in the United States for the treatment of metastatic prostate cancer, has a greater food interaction than any drug currently on the market. Depending on the fat content, food can increase drug exposure 5- to 10-fold, yet it is labeled to be taken in a fasting state, Dr. Ratain points out.
The effect of a high-fat meal indicates "that absolute bioavailability in the fasting state is no more than 10%," says Dr. Ratain, and is likely due to the low absorption rate under fasting conditions; 77% of the drug is excreted as abiraterone (or its prodrug acetate) in the stool.
"In addition, there is marked interindividual pharmacokinetic variability, with a coefficient of variation in the plasma concentration area under the curve of 59% under labeled conditions," he notes.
It is unclear why this faulty labeling is occurring, or why the FDA has not been more stringent about proper labeling for oral oncology drugs, Dr. Ratain said. He told Medscape Medical Newsthat the FDA has not responded to his inquiries and "essentially stonewalled" him because of the confidentiality of their discussions with sponsors.
"The blame could be assigned to sponsors, the Office of Oncology Drug Products, and/or the Office of Clinical Pharmacology and Biopharmaceutics," he said. "Given that there are multiple sponsors, all of whom who have a different approach to oncology..., the evidence points to an FDA origin."
Oral Drugs Relatively New
Alex Adjei, MD, PhD, who was approached by Medscape Medical News for independent comment, explained that until recently, oncology drugs have generally been given intravenously; for other diseases, oral drugs have been in common use for many years.
Perhaps that is why there has not been that much attention on the food-effect data, Dr. Adjei hypothesized. He is chair of the Department of Medicine at Roswell Park Cancer Institute, in Buffalo, New York.
"In clinical trials, we tend to give the drugs on an empty stomach," he said. "So that was how the drug was tested and that was how it was approved. We don't go back and change the label."
Dr. Adjei pointed out that unlike intravenous drugs, there is considerable variation in the way oral drugs are absorbed. "We try to simplify it by giving them on an empty stomach," he said.
Food-Effect Data Ignored
In a study published last year (Clin Cancer Res. 2010;16:4446-4451), Dr. Ratain and his colleague, Soonmo Peter Kang, MD, also from the University of Chicago, evaluated the effect of food on 99 oral agents that received FDA approval from January 2000 to May 2009. They compared food-labeling patterns between oncology and nononcology drugs using Fisher's exact test.
Food had a significant effect on drug bioavailability in 34 of the 99 drugs evaluated. In cases where food markedly enhanced bioavailability, 8 of 9 nononcology drugs were labeled to be taken with food to take advantage of the food–drug interaction. However, all oncology agents were labeled to be taken in "fasted" states (P = .01).
"The food-labeling pattern of recently approved oral oncology drug products is inconsistent with fundamental principles of oral drug delivery," they write. "The labels of 3 agents (erlotinib, nilotinib, and lapatinib) minimized bioavailability through food restrictions, which is in contrast to the labeling principles used for all other classes of oral agents."
In an editorial that accompanied the study by Drs. Ratain and Kang (Clin Cancer Res. 2010;16:4305-4307), Rajul K. Jain, MD, from the University of Texas M.D. Anderson Cancer Center, Houston, and colleagues explained that the effects of food are complex, and that multiple factors determine final labeling recommendations. "When analyzing the label given to any product or product class, all of the potential influencing factors should be considered," they said.
They also noted that "sponsors are encouraged to conduct food-effect studies early in a product's development so the results can be implemented into larger clinical trials."
In his comment, Dr. Ratain refers back to that editorial, and points out that the agency has "expressed concern about the potential for high intraindividual variability in cancer patients resulting from variability in oral intake secondary to disease or concurrent medications." This appears to indicate a preference for labeling drugs to be used when fasting, even if their bioavailability would be substantially increased when taken with food.
He explains that there isn't any evidence that this concern applies to all oral oncology drugs, and that there is no evidence that intraindividual variability in diet is greater for cancer patients than for those with other life-threatening conditions who also take daily oral drugs.
"If the US Food and Drug Administration were truly concerned about intraindividual variability in diet, why has this never come up in the labeling of nononcology drugs?" he asks.
He notes that the FDA was aware of the food-effect data from early studies of abiraterone and could have required at least 1 randomized study to be conducted to review this, so that it could have been labeled to be taken with food. The FDA has the authority to require such studies as a postmarketing requirement, but chose not to.
With the risk for poor adherence to food-effect labeling, there is a risk for overdose or underdose of a given drug, with the more serious being overdose, Dr. Ratain says.
In addition to this safety issue, there are economic implications, Dr. Ratain notes.
The standard dose of abiraterone for men without liver problems is 1000 mg once daily. But because the bioavailability of the drug is greatly increased when taken with food, it would be interesting to have studies of abiraterone at a dose of 250 mg once daily, administered with a standard breakfast. If this approach is shown to be safe and effective, it could save patients and their payers approximately $3750 per month at current prices.
"It also would send a message to the pharmaceutical industry that flawed drug development has financial consequences," he notes, "along with potential risks of product liability because of defective labeling."
He concludes: "After all, given the high attrition rate for oncology drugs, do we really need to flush our few successful products down the toilet?"
Impracticalities of Dosing and Diet
Lawrence J. Lesko, PhD, FCP, former director of the Office of Clinical Pharmacology at the FDA's Center for Drug Evaluation and Research, and a coauthor of the editorial by Dr. Jain, pointed out that the principal of labeling is consistent safety.
"The most consistent way to take drugs is on an empty stomach," said Dr. Lesko, who is now at the Department of Pharmaceutics at the College of Pharmacy at the University of Florida in Gainesville. "It would be impractical to expect that patients are going to adhere to a specific diet when taking these drugs. And it is often unclear what 'food' may mean to the patient. Is it a glass of juice, breakfast cereal, a cheeseburger?"
The goal is to minimize the variability, he told Medscape Medical News. "Food studies are generally conducted in healthy volunteers and not cancer patients. Labeling can't be done on a theoretical basis; it has to reflect what was done in the study."
Dr. Lesko noted that another problem with tying an oral drug dose to food is that many cancer patients have problems with appetite, nausea and vomiting, and gastrointestinal motility. "Their appetite can change on any given day," he said. "Some days, they may not be able to tolerate a high-fat meal, for example, which raises the bioavailability of abiraterone. It would just be impractical to do this on a day-to-day basis."
The other issue that Dr. Ratain focuses on in his comment is the comparison of oncology and nononcology drugs, Dr. Lesko pointed out. "But you really can't compare them."
For most nononcologic agents, food doesn't matter, so it isn't that important how the product is labeled, he said.
But that is not the case with oral oncologic agents. "Every drug in oncology has a narrow therapeutic index, and taking the drug with food could affect the therapeutic range," he said. "The labels are intended to minimize variability and mimic approval data."