This article is more than 5 years old. Information may no longer be current.
Researchers: Off-label cancer treatment recommendations often based on weak evidence
Click Here to Manage Email Alerts
National Comprehensive Cancer Network guidelines frequently recommends that newer branded drugs be used in settings beyond their FDA-approved indications, according to results of a retrospective, observational study published in The BMJ.
“The strength of the evidence cited by National Comprehensive Cancer Network supporting such recommendations is weak. Our findings raise concern that the National Comprehensive Cancer Network justifies the coverage of costly, toxic cancer drugs based on weak evidence,” Vinay K. Prasad, MD, MPH, hematologist-oncologist and assistant professor of medicine at Oregon Health and Sciences University and a HemOnc Today Next Gen Innovator, and colleagues wrote.
Prasad and colleagues sought to determine the differences between recommendations by National Comprehensive Cancer Network (NCCN) guidelines and FDA approvals of 47 molecular therapies the FDA approved between 2011 and 2015 for adult hematologic or solid cancers.
During the period analyzed, FDA approved these therapies for 69 indications; however, NCCN recommended those same therapies for 113 indications.
Of note, FDA did grant additional approval for six (14%) of the additional NCCN-recommended uses during a 21-month follow-up period.
“Yet, many recommendations may continue to lack empirical support,” Prasad and colleagues wrote. “When it comes to costly cancer drugs with serious toxicities, it is important to recognize that merely because some drugs were later validated does not mean that off-label recommendations for all drugs is, on average, a good thing. It remains an open question.”
HemOnc Today spoke with Prasad about the study, the potential implications of the findings, and the directions he believes subsequent research should take.
HemOnc Today also invited Robert W. Carlson, MD, CEO of NCCN, to address the conclusions Prasad and colleagues reached in their study. Carlson’s response appears in the Perspective that accompanies this article.
Question: What prompted you to conduct this research?
Prasad: The role of NCCN guidelines in cancer care is ubiquitous. First, they are involved in making coverage decisions. CMS by law has to pay for many recommendations that are made by NCCN, so Medicare has to pay for those drugs and then cannot negotiate the price. Second, anyone who practices clinical oncology in the United States knows these NCCN recommendations are widely cited. Every week, when I am in clinic and I see my fellows, they open up NCCN guidelines and say, ‘NCCN says we should do X, Y or Z.’ They know that I am going to respond by asking them, ‘Why do they say to do X, Y or Z?’ I then force them to dig into each question further. For our study published in The BMJ, we decided to ask the question of what the evidence was each time NCCN recommended something beyond what the FDA approved it for.
Q: How did you conduct the study?
Prasad: On a single day in March 2016, we froze the entire FDA drug approval list, and we gathered the entire NCCN 4,000-page document and froze it, as well. We compared these two documents and, in the month that followed, we painstakingly compared every time a drug was mentioned and how many times it was recommended by NCCN.
Q: What did you find?
Prasad: The first thing we found was that FDA had approved these new drugs for nearly 70 indications. Then we saw that every single indication the FDA had approved the drug for was in the NCCN document. It missed nothing. In addition, NCCN recommended another 40 or so indications for which these therapies were not FDA approved. We then asked what the evidence was for those 40 or so additional purposes. In most cases, NCCN offered no evidence. In some cases, the evidence was very weak and came from uncontrolled clinical studies.
Q: Did your findings surprise you?
Prasad: We were surprised that NCCN was recommending these drugs so often beyond the approved uses, and we were surprised that the evidence cited was poor, or that evidence was not cited at all. In the United States, we have a system in which officials are stating they are going to pay for all of these cancer drugs that are recommended by experts. Those experts are going to be recommending them beyond what they have been approved for and they are not going to tell you the evidence that they are citing to recommend them, or there is no good evidence, and you have to pay for them once they recommend it. The other thing that serves as sort of a backdrop for this is that many of the experts on the panel have financial ties to drug makers. When putting all of this together, we have to ask ourselves if this is the optimal system. The whole picture is what is concerning. There is the potential for bias here.
Q: What are the next steps?
Prasad: It is important to draw attention to the issue, and I encourage others to take a dive into NCCN documents. What this really calls for is that CMS and also private insurers that use NCCN documents when making coverage decisions should probably perform an audit of that choice and ask themselves if it is the wisest way to do it. In the case of the federal government, it seems pretty reasonable that they convene a set of nonconflicted experts to make off-label recommendations that may be a better way to maximize patient outcomes, as well as be a steward of federal resources. CMS should audit if this is the best use of taxpayer money. CMS has a sworn oath, which is to do what is best in the interest of their beneficiaries. In paying for some of the off-label uses of these drugs, it is essentially making a choice, which is that instead of broad campaigns to improve antihypertensive use, instead it is covering some of these $100,000 drugs with side effects for off-label use with no data cited about how they benefit patients. This runs up against its duty, and it should reconsider if it cannot be a better steward of its own beneficiaries.
Q: Is there anything else that you would like to mention?
Prasad: The current system of drug regulation and drug payment in cancer has turned out to be very bizarre. The more we look at it, the more we cannot help but feel that the purpose is to make it easy to get coverage for very costly drugs with very limited and uncertain evidence base. It does not seem to be a bug of the system. It seems to be a feature of the system. This study shows a clear example of where we see this happening, and I think it should make all of us who believe in rational drug policy take a closer look. – by Jennifer Southall
Reference:
Wagner J, et al. BMJ. 2018;doi:10.1136/bmj.k668.
For more information:
Vinay K. Prasad, MD, MPH, can be reached at Oregon Health and Science University, 3181 S.W. Sam Jackson Park Road, Portland, OR 97239; email: prasad@ohsu.edu.
Disclosure: Prasad reports no relevant financial disclosures.
Perspective
Back to Top
Robert W. Carlson, MD
The NCCN Clinical Practice Guidelines in Oncology are based on the strongest scientific evidence available. More than 24,500 scientific publications cited in the guidelines discussion sections provide the evidence NCCN uses to make recommendations. These citations are available to anyone for free at www.nccn.org.
Oncologists know that many cancer treatment regimens are effective against cancer beyond FDA-approved indications. In situations — particularly with less common cancer types — where the highest levels of evidence simply do not exist, NCCN’s recommendations are based on the best available evidence and consensus from the highly experienced specialists among our 1,355 panel members across 27 leading academic cancer centers.
Numerous studies in a variety of cancer types have shown that adhering to NCCN guidelines saves lives and money. Our number one priority is making sure that all people with cancer have access to the treatments that experts agree are most likely to extend the length and improve the quality of their lives.
References:
Affronti ML, et al. Support Care Cancer. 2014;doi:10.1007/s00520-014-2136-0.
Bristow RE, et al. Obstet Gynecol. 2013;doi:10.1097/aog.0b013e3182922a17.
Bristow RE, et al. J Natl Cancer Inst. 2013;doi:10.1093/jnci/djt065.
Foster JE, et al. Ann Surg Oncol. 2008;doi:10.1245/s10434-008-0021-0.
Lee JY, et al. Eur J Surg Oncol. 2015;doi:10.1016/j.ejso.2015.01.006.
Mearis M, et al. J Pain Symptom Manage. 2014;doi:10.1016/j.jpainsymman.2013.09.016.
Molena, et al. J Gastrointest Surg. 2015;doi:10.1007/s11605-015-2899-8.
Newcomer LN, et al. J Oncol Pract. 2017;doi:10.1200/jop.2016.015198.
Schwam ZG, et al. Clin Oncol (R Coll Radiol). 2016;doi:10.1016/j.clon.2016.01.010.
Visser BC, et al. HPB (Oxford). 2012;doi:10.1111/j.1477-2574.2012.00496.x.
Worhunsky DJ, et al. J Natl Compr Canc Netw. 2015;13:319-325.
Click Here to Manage Email Alerts