This article is more than 5 years old. Information may no longer be current.
Be aware of CMS coverage restrictions for off-label uses for anticancer drugs
Off-label uses are often standard of care but critics have questioned expense and reliability of compendia.
Click Here to Manage Email Alerts
Late in 2008, the CMS changed its reimbursement rules and allowed coverage of off label uses for anticancer drugs.
Because of this change to the national coverage determination, physicians could be compensated for prescribing drugs for uses not approved by the FDA, if that specific indication was covered in one of four medical compendia and none of the other compendia recommended against the indication.
On one level, CMS was merely acknowledging reality; off-label uses make up an estimated 50% of anticancer prescriptions, and the National Comprehensive Cancer Network estimated in 2005 that off-label uses comprised up to 75% of all anticancer prescriptions.
According to the NCI, a 1997 survey of 200 oncologists conducted by the American Enterprise Institute and the ACS found that 60% of respondents had prescribed drugs off-label. “Frequently, the standard of care for a particular type or stage of cancer involves the off-label use of one or more drugs,” NCI said.
|
Photo by Eileen Noonan Images |
However, some experts who spoke to HemOnc Today remain concerned that the new rules may encourage physicians to choose to administer more expensive off-label drugs. Moreover, a study published in Annals of Internal Medicine in March 2009 showed a fair amount of disagreement among the compendia, leading the researchers to conclude that the compendia needed more methodological rigor when evaluating off-label uses.
“We do put patients at harm by using a drug that hasn’t been vetted in a particular setting,” said Douglas Yee, MD, director of the Masonic Cancer Center at the University of Minnesota and member of the HemOnc Today Editorial Board. “Chemotherapy drugs have certain toxicities. We always know, 100% of the time, that we’re exposing a patient to toxicity when we prescribe a drug, either on-label or off-label. The difference is that when you go off-label, you have less assurance that the drug is going to be of benefit. By opening up CMS coverage entirely, we do expose patients to some risk they otherwise might not have been exposed to.”
Supporters of the change said CMS merely gave physicians the flexibility they need to treat patients effectively while ensuring doctors would be paid for their efforts.
“If one were limited to the armamentarium of drugs that have an FDA indication, one would be very limited in our standard drugs with the highest level evidence in treating neoplastic disease. We’d be left with nothing,” said Samuel M. Silver, MD, director of the Cancer Center Network for the University of Michigan Health System and a
HemOnc Today Editorial Board member. “For, say, the treatment of curable diseases like Hodgkin’s, many of the drugs we use in the standard regimen are not FDA indicated for curative, front-line therapy. We would be left with drugs that haven’t been used in those combinations for decades because they’re toxic and much less effective than the drugs we have now.”
Off-label support
The Omnibus Budget Reconciliation Act passed in 1993 mandated Medicare coverage for off-label uses of anticancer drugs if those uses were supported by three designated compendia: the AMA’s Drug Evaluations, American Hospital Formulary Service Drug Information and the United States Pharmacopeia Drug Information. Publication has since stopped on both Drug Evaluations and Drug Information, and the NCCN’s Drugs and Biologics Compendium, Thomson Micromedex DrugDex and Clinical Pharmacology have been added to the list of approved compendia.
Additionally, if any of the recognized compendia include a “Not Recommended” listing for an off-label use, that use is not eligible for coverage even if that use is listed as accepted in another compendium. Uses that CMS deems not medically appropriate will not be covered even if support exists in one of the compendia, although policy differs depending on the CMS contractor.
|
|
In 2005, ASCO expressed support for a CMS decision allowing coverage for off-label uses of drugs approved for the treatment of colorectal cancer in specific clinical trials, including oxaliplatin (Eloxatin, Sanofi-Aventis), irinotecan (Camptosar, Pfizer), cetuximab (Erbitux, ImClone Systems) and bevacizumab (Avastin, Genentech).
A year later, the organization published a policy statement, titled “Reimbursement for Cancer Treatment: Coverage for Off-Label Drug Indications,” calling on HHS to include all qualified compendia for use in the Medicare program.
CMS also accepts scientific evidence supporting off-label indications from 26 peer-reviewed journals, including Blood, the Journal of Clinical Oncology and Clinical Cancer Research. The agency expanded the number of journals it would accept from 15 to 26 in part due to lobbying from ASCO.
Concerns with compendia
Amy P. Abernethy, MD, associate professor of medicine at Duke University School of Medicine and co-director of both the Duke Quality Cancer Care Initiative and the Duke Integrative Oncology Program, served as co-author of the study published in Annals of Internal Medicine mentioned previously. In this study, funded by the US Agency for Healthcare Research and Quality, Abernethy and colleagues reviewed each compendium’s stated methods as well as, literature related to off-label indications of 14 cancer drugs in 2006 and updated literature related to a single off-label indication from 2006 to 2008, and determined the breadth of content and citations.
As an example, the researchers in 2006 identified 43 published results from phase 1 to phase 3 studies and 15 conference abstracts evaluating gemcitabine (Gemzar, Eli Lilly) to treat bladder cancer. The various compendia cited as many as seven and as few as zero of those published results.
Abernethy and colleagues found another 25 published reports upon performing another review in 2008. The researchers said only DrugDex increased its citations, going from three to 11, whereas the other compendia “had little or no change.”
“Current compendia lack transparency, cite little current evidence, and lack systematic methods to review or update evidence,” they concluded.
In an editorial accompanying the Abernethy study, Harold C. Sox, MD, then editor of Annals of Internal Medicine, said the compendia were “inconsistent, incomplete and out-of-date.”
“Abernethy and colleagues’ review strongly suggests that even an enhanced compendia-based system is too weak to deal with the methodological challenges of the evidence base for off-label indications,” he wrote. “Compendia play a useful role, but Congress has assigned them a task that requires considerably greater rigor than they are accustomed to using for their main purpose.”
Abernethy told HemOnc Today that because compendia frequently contradict each other and may contain out-of-date information, they should not be considered as authoritative sources. She said that since the article was published, the various compendia had made improvements in managing conflicts of interest and updating references as new information became available. However, she said the compendia have an impossible task.
“The challenge the compendia have, which is the challenge of a physician taking care of patients, let alone any third-party payer trying to make reimbursement decisions, is that the evidence is moving so fast and the process for systematic evidence review in this setting when we’re trying to make decisions in the context of evolving technology — we don’t have a good process for that,” she said. “My sense is that we’ve learned through this process that, overall, their intentions are pretty good. But the main thing our subsequent work showed is that the task itself, in the current way in which we approach evidence synthesis like this either in the physician’s head or for the compendia as an authoritative resource, is a fundamentally undoable task.”
NCCN CEO William T. McGivney, PhD, defends Drugs and Biologics Compendium. Information included in the compendia is based on the results of thousands of studies and the opinions of the world’s leading physicians who meet continually to discuss the latest developments in the field, he said.
|
|
“We have more doctors on our breast cancer panel than all the other compendia for all cancers,” he said. “We have the world’s thought leaders, the doctors who’ve established the literature. They know the literature backwards and forwards. If there is any group to evaluate the evidence, it should be an organization like NCCN.
“We specifically cite the seminal study that supports the indication, the dose and the schedule of administration. If you go to our breast cancer guidelines, there are 121 pages with 375 references. I don’t know how many more references you need to support your recommendation.”
Other concerns
Yee has had success using off-label regimens to treat patients, but he said there is a threat the practice could increase costs for Medicare and Medicaid, and give physicians a reason to prescribe a more expensive off-label drug.
“The disadvantage is that costs go out of control,” he said. “Many oncology practices have chemo charges as part of their revenue stream. Certainly, if you can continue giving the right chemotherapy, that may be beneficial to your practice. I don’t necessarily think that’s bad, especially if we know that the outcomes clearly benefits the patient. However, we can be in the position of prescribing expensive drugs without clear data that the drug has benefit for our patients.”
Yee’s other concern is that there is not enough follow-up reporting to show efficacy. As he noted, scientists often see ideas that work in theory fail when put into practice.
Unlike clinical trials in which data are recorded, reported and peer-reviewed before publication, outcomes from off-label use are not collected in any systemic way. Not all patients are eligible for trials and ASCO estimated in 2008 that only 2% to 7% of eligible patients participate, so recording these outcomes would strengthen the science supporting a given off-label use, Yee said.
“Off-label coverage opens the door for uses of drugs that physicians think are a good idea, but don’t have any data for,” he said. “In an ideal world, if a physician does use an off-label drug, then that practitioner would be required to report the results in the context of knowing whether a drug works or not. The regulatory bodies could review those data and make firmer statements about whether the drug has benefit. If we go to a lot of off-label use, I believe there has to be some regulatory/reporting mechanism. This type of research falls under the category of comparative effectiveness, and I think most oncologists would be willing to do that. I don’t think oncologists like to be told that we can or can’t do something, but I think oncologists are happy to report the results of a particular agent in their practice.”
It appears that the Office of the Inspector General is trying to explore the cost-efficiency of off-label uses. In its fiscal 2010 Work Plan, the OIG said it will review off-label uses beginning in fiscal 2011 to see if they produced results better than those achieved with FDA-approved regimens in an effort to trim the $2.7 billion the government spent on chemotherapy drugs in 2007.
|
|
“We will determine whether patients with particular indications were prescribed anticancer drugs approved by [the] FDA for such indications before resorting to anticancer drugs not approved for those indications and, if so, whether there were improvements in the patients’ medical conditions prior to use of off-label drugs,” the plan says. “If the beneficiaries’ medical conditions improved prior to use of off-label drugs, we will determine how much Medicare could have saved had anticancer drugs continued to be used within indicated usage.”
Abernethy said the investigation was a reasonable look into medical practice.
“If you controlled Medicare, you’d want to be asking this question. If you’re a physician and you don’t want anyone to pay that much attention to what you’re doing, you wouldn’t want anyone asking this question,” she said. “But in a world where at the end of the day you should feel proud of what you’re doing because it’s the right thing for the patient, this question is totally fine.”
Silver, however, said the OIG was overstepping its bounds.
“They’re trying to see how much money could be saved if a physician didn’t use an off-label drug when a patient failed on an on-label drug,” he said. “But because of the history and expense of FDA labeling, sometimes the most appropriate drugs are off-label to begin with, and it would be inappropriate to use the on-label drug. This is ill-conceived.”
Potential solutions
In addition to Yee’s more rigorous reporting requirements, Abernethy suggested a way to better disseminate information on cutting-edge research. She said it is impossible for the compendia to keep up with the science, if only because publishing on paper takes so long, but there are better, faster ways to deliver updated information.
In an article in the Journal of Clinical Oncology published online ahead of print in June, she lays out a plan of “rapid learning health care.” The Institute of Medicine introduced rapid learning health care in 2006 with the idea of moving beyond traditional approaches to faster ways of finding, synthesizing and disseminating evidence to point of care.
Rapid learning health care is based on six principles:
- Collecting data in a planned and strategic manner;
- Analyzing captured data;
- Generating evidence through retrospective analysis of existing data as well as data from prospective studies;
- Implementing new insights into subsequent clinical care;
- Evaluating outcomes of changes in clinical practice;
- Generating new hypotheses for investigation.
“The care of a particular individual is informed by all the people who look like her and who have come before her,” Abernethy said. “Then her care is reinvested into that data system in a very fast, real way so that as we run into the next similar-looking patient, we’re able to rapidly query that system and make the next set of decisions.”
Abernethy said physicians and researchers are just now figuring out how to implement such a system as electronic health records become more widespread, reliable and secure.
“The idea is to get us to a trustworthy system of linked data that is adequately queryable so that we can find these answers,” she said.
For his part, McGivney said the current system relying on compendia and peer-reviewed journals does an excellent job keeping track of new developments in scientific research and off-label uses. In March 2009, the US Agency for Healthcare Research and Quality said that the compendia needed more references, but another AHRQ report released the following October said that there was not enough good evidence out there to support even the current references.
“They can’t have it both ways,” McGivney said. “We track the literature assiduously. Whenever there is any change in the literature of an FDA position, our doctors meet and address that issue specifically. You can’t get any more comprehensive. You can’t get any more evidenced-based. You can’t have any more expert judgment integrating the decision-making process than that.
“The perfect is the enemy of the good, and while sometimes it’s nice to write general articles about what the ‘perfect’ might be, every day patients are diagnosed with cancer. Every day, those patients need treatment recommendations. That’s what we do. We address the needs of patients based on our expert judgment and based on what the scientific evidences assert should be done for those patients.” – by Jason Harris
Should compendia be considered authoritative sources?
The compendia may differ in their indications, but they represent the best way we have to evaluate off-label uses.
There are dozens and dozens of individual cancers. Not every drug is going to get an on-label indication for every area that it works in. There has to be some standard or order, thus are born the compendia which take the public literature and assess the validity of a drug’s use in areas beyond those indicated on the label.
The various compendia differ in their degree of rigor, in the criteria they use to determine whether a drug is indicated for its off-label use. Because there are thousands of papers published in the oncology literature every year, the speed at which the various compendia keep up with things — or not — differs.
|
|
ASCO does not publish a compendium nor do we have plans to. I can only imagine the degree of effort required to put a compendium together. It is not surprising that all the compendia don’t reach the same conclusion. The question for those of us in the community and insurers is, what do we do with that?
The compendia are evaluated by the Centers for Medicare and Medicaid Services. To that extent, we have to let CMS use their criteria to determine that a given compendium meets the standards, or doesn’t, and the compendia that hurdle the bar stay and those that fall behind be discarded.
I wouldn’t call the task impossible. The task is difficult. Can computer algorithms comb the data and present the literature? Absolutely. The data is easily searchable. The next step, however, is that people with knowledge have to read the papers, analyze the data, evaluate whether there is some level of rigor to the work and decide whether the evidence merits inclusion in a compendium. It’s arduous. It takes a lot of effort and even if done perfectly, one might expect some level of difference between one compendium and another.
I doubt that more compendia would make much difference. It seems to me that a handful of compendia are more than enough. The literature is the literature. More compendia would mean that some agents would get additional plusses and some would get additional minuses, but it would not create this ideal situation that everyone would agree about a given indication.
Allen S. Lichter, MD, is CEO of the American Society of Clinical Oncology.
The compendia are not up to the task of evaluating off-label uses and should not be used to make coverage decisions.
Based on the research by Abernethy et al, the evidence is that the compendia are not up-to-date, frequently disagree and are a poor substitute for the processes we use for other types of coverage decisions, which are based on systematic reviews and open meetings where conflicts of interest are clear. Therefore, I don’t think they are a very suitable basis for what amounts to public policymaking.
The process used by CMS for its national coverage decisions should also be applied when deciding whether to pay for off-label uses: a systematic review of the evidence of the effectiveness and harms of a treatment, and some type of open process involving a panel of experts that includes input from specialists but maintains a significant role for other health professionals.
It’s true that the science moves fast in the cancer field, but it is also true that evidence tends to accrue relatively slowly. The big question is whether Medicare should be paying for uses of cancer drugs that aren’t supported by substantial evidence that the drugs are effective and that their harms are in reasonable balance with their benefit.
|
|
The compendia don’t have the resources to do what I would consider a proper job of evaluating the evidence and determining that benefits are so substantially in advance of harms that Medicare or insurance companies should pay for these uses. Basically, it is ultimately up to the editor to decide whether or not something appears in a compendium. According to the Abernethy article, the editors recognize that it is fully at one person’s discretion whether something goes in the compendium and therefore is paid for, or is left out of the compendium and not paid for. That is way too much power for one person, particularly if that person’s conflicts aren’t known to the public.
Compare that process to the process used by CMS for other coverage decisions: systematic reviews and in many cases an open public process. The scales are way down on the side of a more thoughtful process. It would be expensive, but look at the kind of drugs that are being subject to these decisions. These are expensive drugs, drugs that are associated with serious harms and, in many cases, have unproven benefits for the specific indications being paid for that are listed in the compendia.
Harold C. Sox, MD, is editor emeritus of Annals of Internal Medicine and emeritus professor of medicine at Dartmouth College.
For more information:
- Abelson R, Pollack A. Medicare widens drugs it accepts for cancer. The New York Times. Jan. 27, 2009. Available at: www.nytimes.com. Accessed: Sept. 27, 2010.
- Abernethy A. Ann Intern Med. 2009;150:336–343.
- Arnst C. Soaring cancer drug costs may cripple Medicare. Bloomberg Businessweek . Jan. 27, 2009. Available at: www.businessweek.com. Accessed: Oct. 5, 2010.
- ASCO. J Clin Oncol. 2006;doi:10.1200/JCO.2006.06.8940.
- Bach PB. N Engl J Med. 2009;360:626-633.
- Bristol N. Lancet. 2009;373:532.
- Centers for Medicare and Medicaid Services: Off-label use of colorectal cancer drugs allowed in select clinical trials. J Oncol Practice. Available at: jop.ascopubs.org/content/1/1/12.full. Accessed: Oct. 20, 2010.
- FY 2010 Office of Inspector General Work Plan. Available at: www.entnet.org/Practice/upload/Work_Plan_FY_2010.pdf. Accessed: Oct. 22, 2010.
- Gillick MR. Ann Intern Med. 2099;150:344-347.
- Hampton T. JAMA. 2007;doi:10.1001/jama.297.7.683.
- Largent EA. Arch Intern Med. 2009;169:1745-1747.
- Recent developments in Medicare coverage of off-label cancer therapies. J Oncol Practice. 2009;doi:10.1200/JOP.0913001.
- Ridley DC. Arch Intern Med. 2006;166:1021-1026.
- Sox HC. Ann Intern Med. 2009;150:353-354.
- Zon R. J Clin Oncol. 2008;doi:10.1200/JCO.2007.15.6398.