Tempering Expectations of Breakthrough Therapy Designated Drugs
A recent study published in the Journal of Clinical Oncology (online June 20, 2018; doi:10.1200/JCO/2017.77.1592) sought to evaluate the United States Food and Drug Administration (FDA) breakthrough therapy designation for novel cancer agents. The major conclusion of the study was that little evidence exists showing that breakthrough therapy designated drugs offer improvements in efficacy, safety, or novelty compared with non-breakthrough therapy designated drugs.
Journal of Clinical Pathways spoke with Jonathan J Darrow, SJD, JD, MBA, a faculty member at Harvard and the senior author of the study, to shed further light on this finding and divulge its larger implications.
What was the inspiration for evaluating efficacy, safety, and time to approval for breakthrough designated drug approvals?
Dr Darrow: We had a terrific group of authors, and each probably had his or her own motivation for exploring the specific topic of the paper. For me, the general idea of looking at efficacy levels of new drugs arose out of my dissertation work at Harvard, even before the breakthrough designation was created. I studied what is required, in terms of drug benefit, for a drug to be approved in the US. The answer, I discovered to my surprise, was “very little.” Under current law, it turns out, there is no minimum quantum of benefit that is required for a drug to be approved. Drugs must cross the “zero benefit” threshold, based on substantial evidence, but there is no particular amount above zero that is required. This means that some expensive new drugs may provide patients with little more than a placebo effect.1
But many people expect new drugs to offer substantial benefits.2 For breakthrough therapy designation, the expectation is even higher – that drugs approved will offer much more than the average drug. For the recent article in JCO3, we wanted to investigate what the data showed for breakthrough-designated oncology treatments, in part because these drugs tend to be particularly expensive.
Can you briefly provide an overview of your data? What was the larger conclusion that your data led you to? How do breakthrough therapy designated cancer drug approvals relate to non-breakthrough therapy designations?
Dr Darrow: We examined all new oncology drugs approved between 2012 and 2017. There were 58 in total, 43% of which received a breakthrough therapy designation. Since 2012 and 2013 were transition years for the breakthrough program, if you look at just 2014-2017, more than half (59%) of oncology treatments received a breakthrough designation. We divided drugs into two groups: those that received breakthrough therapy designation and those that did not. It turns out that there was no statistically significant difference between breakthrough-designated and non-designated oncology drugs across a number of measures, including efficacy, safety and innovativeness – which was measured by the percentage of drugs that had a novel mechanism of action.
The drugs were approved based on different endpoints, such as progression-free survival or overall survival. But the most common measure was some form of “response rate” – and there was no statistically significant difference between breakthrough and non-breakthrough drugs in terms of these response rates. In fact, the numerical trend actually favored non-breakthrough drugs.
A similar trend favoring non-breakthrough drugs was observed for rates of death not caused by disease progression; a higher percentage of subjects taking breakthrough drugs died, although again the difference was not statistically significant. A similar trend was also found when we looked at serious adverse events.
There was one notable trend that favored the breakthrough therapy group: differences in progression-free survival. The difference in that measure was 8.6 months for breakthrough therapy designated drugs vs 4.0 months for non-breakthrough therapy designated drugs. The difference was not statistically significant, but numerically does seem to favor the breakthrough therapy group. The caveat, though, is that most of the cancer drugs were not approved on the basis of progression-free survival, so it was a minority of drugs that fell within that category.
The other caveat—and it is a big one—is that even for the minority of drugs that were approved on the basis of progression-free survival, it cannot be assumed that the progression-free survival gains of 8.6 months will translate into survival gains of 8.6 months until and unless the evidence shows that to be the case. It would be premature to conclude that the apparent, non-statistically-significant advantage in the surrogate endpoint actually signals a benefit that is relevant to patients. A previous study has shown a poor correlation between surrogate endpoints and survival, specifically in the area of cancer.
In light of these findings, is there any value to the breakthrough therapy designation?
Dr Darrow: There is certainly a benefit to the breakthrough therapy designation from the perspective of the manufacturer. It lends the seal of approval of the government’s expert agency; it is, quite literally, a judgment by the Federal Government’s expert drugs agency that a given drug merits a “breakthrough” designation. That can help attract investment, boost utilization, and increase stock prices. It may also increase expectations among patients that the drug will be approved and thereby potentially increase pressure on the FDA to approve the drug. All of these results can benefit the manufacturer.
It is also possible that there are benefits from the patient perspective. Drugs approved under the breakthrough therapy program are approved faster than drugs that are not designated as breakthroughs. It is a substantially faster process: 4.8 vs 8.0 years.4
But the benefit of faster availability is meaningful only if a drug actually offers improved outcomes. Unfortunately, many breakthrough-designated drugs are approved on the basis of surrogate endpoints and uncontrolled trials5 – meaning trials with no concurrent comparison group, making it difficult to determine whether a given drug actually offers an advantage.
What are the larger implications of your findings?
Dr Darrow: In my view, the bottom line is that by labeling new cancer drugs as breakthroughs, when the actual benefit may be neither large nor supported by robust data, the designation is potentially creating expectations among seriously ill patients that are unlikely to be fulfilled.
A related concern is that the term “breakthrough” can be used to justify exceptionally high prices that are far out of proportion to actual therapeutic value.
Cancer is a serious, and often terminal, illness, and everyone wants each new therapy to be a genuine breakthrough. The concern is that, in reality, it may be that patients are sometimes being given false hope.
A final concern is that—perhaps too often—the health care system is expending enormous sums on drugs that offer little, if any, additional benefit.
Are there any other points or parting messages you would like to make that have not been covered?
Dr Darrow: Most breakthrough-designated oncology treatments are approved under the accelerated approval program, which means that they are approved on the basis of surrogate endpoints that are only “reasonably likely” to predict clinical benefit. At the time of approval, the FDA considers the benefit of these accelerated approval drugs to be uncertain, until confirmatory trials are completed years after the drug is approved. Because benefit is uncertain, the FDA excludes these drugs from the definition of “existing therapies,” which is the standard against which designation of future breakthrough therapies is made.
You may have noticed that there is a tension here. The same drug that is granted accelerated approval status, and for which benefit is therefore by definition uncertain, is simultaneously being granted a breakthrough therapy designation. The FDA is on the one hand saying that clinical benefit is uncertain until the confirmatory trials are completed, and on the other hand saying that the drug qualifies as a “breakthrough.” The public—both patients and physicians—may have a difficult time understanding what message the FDA is sending about these drugs.
In addition, because accelerated approval drugs are excluded from the definition of “existing therapies,” subsequent drugs that are approved with a breakthrough therapy designation may often not have to be compared to the most recently-approved, and potentially most relevant, previous breakthrough drugs. That is significant because it means it can be very difficult to determine whether new drugs actually offer benefit.
1. Darrow JJ. Pharmaceutical efficacy: the illusory legal standard. Washington & Lee Law Review 2013;70:2073–136.
2. Schwartz L, Woloshin S, Welch GH. Using a drug fact box to communicate drug benefits and harms. Annals of Internal Med. 2009; 150(8):516-27.
3. Hwang TJ, Franklin JM, Chen CT, Lauffenburger JC, Gyawali B, Kesselheim AS, Darrow JJ. Efficacy, safety, and regulatory approval of Food and Drug Administration-designated breakthrough and non-breakthrough cancer medicines, J Clin Onc 2018;36(18):1805–12.
4. Hwang T, Darrow JJ, Kesselheim AS. The FDA’s expedited programs and clinical development times for novel therapeutics, 2012–2016. JAMA 2017;318(21):2137–8.
5. Pregelj L, Hwang TJ, Hine DC, Siegel EB, Barnard RT, Darrow JJ, Kesselheim AS. Precision medicines have faster approvals based on fewer and smaller trials than other medicines. Health Affairs 2018;37(5):724–31.