Off-Label Drug Use in Oncology Accounts for 18% of Spending

June 2013, Vol 4, No 5

Chicago, IL—An examination of the Surveillance, Epidemiology and End Results (SEER)-Medicare database from 1998 to 2008 revealed that a significant number of patients with cancer receive drugs that are neither indicated by the US Food and Drug Administration (FDA) for the specific condition nor endorsed by the National Comprehensive Cancer Network (NCCN) compendia—and 18% of the spending on cancer drugs is for off-label drug use.

“We found a lot of off-label use,” said Dawn L. Hershman, MD, MS, Associate Professor of Medicine and Epidemiology, Division of He­ma­tology/Oncology, Columbia Uni­versity Medical Center, New York, who presented these findings at the 2013 annual meeting of the American Society of Clinical Oncology.

The extent of off-label use is a policy concern, because the clinical benefits of such use for patients may not outweigh the costs or the adverse health outcomes, Dr Hershman said.

The study identified 42,634 patients with multiple myeloma or with metastatic breast, ovarian, lung, colon, or prostate cancer during that period in the SEER database. FDA drug approvals and their dates were identified; if the claim was within 1 year before the FDA approval date, it was categorized as NCCN compendia–supported. If the claim was made after the FDA approval date, it was categorized as FDA-approved. The researchers also determined which of the drugs were ever approved.

The Medicare reimbursement rate was calculated for the most common regimen and dosing, and was not adjusted for inflation.

The following numbers of unapproved anticancer agents used in the 7 common tumor types were: 20 in breast cancer, 8 in colon cancer, 8 in lung cancer, 18 in multiple myeloma, 28 in ovarian cancer, 33 in prostate cancer, and 13 in uterine cancer.

Overall, 55% of patients received an on-label drug only, meaning that 45% of patients received a drug that was neither FDA-indicated nor compendia-supported. A strong variability was evident by tumor type: patients with multiple myeloma were most likely (80%) to receive an off-label drug and patients with colon cancer were least likely (10%). This pattern was consistent across the 10-year period.

Of the drugs with no FDA indication for the specific tumor type, 70% were endorsed by compendia and 30% were not. The mean number of claims for drugs with no FDA indication or compendia support averaged 11 per patient overall.

“The bulk of unapproved drugs are, however, compendia approved, and this is reassuring, because we think the compendia are appropriate,” Dr Hershman noted. Approximately 50% of the compendia-supported drugs were eventually approved by the FDA for that indication, but those that were not FDA-approved were not deleted from the compendia, she added.

Table
Table: Cancer Drug Use and Associated Costs for 42,634 Patients, 1998-2008.
View larger version

Cost of Off-Label Use: $150 Million in This Cohort
A breakdown of the cost according to appropriate use revealed nearly $150 million in reimbursement for drugs that were neither FDA-indicated nor compendia-supported (Table). Most of the cost for off-label use was for the treatment of multiple myeloma and prostate cancer.

“So, of $850 million spent on drugs in this cohort, $150 million was for drugs only supported by compendia, and $150 [million] was for unapproved indications,” Dr Hershman concluded, adding that the cost of drugs has increased substantially in the past 5 years, and thus the actual costs would be much greater.

She cited data from another recent study (Conti RM, et al. J Clin Oncol. 2013;31:1134-1139) that detailed which patent-protected chemotherapy agents were the most likely to be used inappropriately. Conti and colleagues showed that off-label use (neither FDA-indicated nor NCCN compendia–endorsed) was greatest for rituximab (Rituxan) and gemcitabine (Gemzar; approximately 40%), and for bevacizumab (Avastin; approximately 25%). The least inappropriate use was for trastuzumab (Herceptin), for which essentially no inappropriate use was documented, and for pemetrexed (Alimta; <10%).
The estimated national spending on these common, patent-protected drugs was $12 billion, of which $2.5 billion was for off-label use and NCCN compendia–unsupported use, according to the study by Conti and colleagues.

While acknowledging limitations to the study that pertain to the use of billing codes, estimates of cost, and other factors, Dr Hershman commented, “We didn’t think we would find any unapproved use of drugs. We thought regulations were in place to stop that.”

The Good, the Bad, and the Ugly
Monika K. Krzyzanowska, MD, MPH, Director, Quality Program, Division of Medical Oncology and Hematology, Princess Margaret Cancer Centre, Toronto, Ontario, discussed the study, noting that some off-label use is not completely “inappropriate” but may exist in terms of a different administration schedule, or in a different context for the same disease. In some cases, of course, the drug is used for a disease in which it has not been well studied. Bevacizumab is a good example of this, Dr Krzyzanowska said.

She noted that “this matters,” because there can be “bad” consequences, such as a negative impact on trial accrual and undue financial costs, or “ugly” consequences, such as harm to the patient. Of course, some consequences are “good,” such as drug access for patients with rare diseases and the promotion of innovation, Dr Krzyzanowska said.

She said that Dr Hershman’s study confirms previous research showing that off-label prescribing is common in oncology and varies by disease, although a substantial proportion of off-label use is supported by the NCCN compendia.

Several questions remain to be answered, including:

  • Is compendium-compliant use of drugs appropriate?
  • What is the impact of off-label prescribing on patients (ie, benefit and toxicity)?
  • What is the motivation for off-label prescribing?
  • How can the cost of off-label therapy best be estimated?
  • What is the incidence of off-label use in oral chemotherapy?
  • How would molecular profiling impact off-label prescribing?

Clearly, Dr Krzyzanowska suggested, oncologists are not “choosing wisely” in all circumstances. “Greater scrutiny of off-label use is needed,” she pointed out, “especially for diseases or drugs with a high prevalence of this, for expensive drugs, and for situations where the benefit is likely small and the risk of toxicity substantial.”

Related Articles