Issue Date: July 4, 2016
Clinical trials by the numbers
Despite decades of advances in pharmaceutical R&D, the odds of getting a drug from Phase I clinical trials through approval are still only around 10%. But a few well-placed bets—on a rare disease, on a biologic versus a small molecule, or on the use of a biomarker—can double or triple the likelihood of approval.
These conclusions emerged from a recent study by the Biotechnology Innovation Organization (BIO), a trade association, that used data supplied by Informa’s BioMedTracker and Amplion’s biomarker tracking services. Looking over the past decade, BIO and the two data-gathering firms analyzed 9,985 transitions through clinical trial and approval stages for 7,455 drug development programs at 1,103 companies and found a cumulative approval rate of 9.6%.
Road to approval
|Success rates (%)||Phase I||Phase II||Phase III||Approval||Phase I to approval||No. of transitions(a)|
|All diseases/drug types(b)||63.2||30.7||58.1||85.3||9.6||9,985|
|New molecular entities||61||26||49||78||6.2||5,858|
|Non-new molecular entities(c)||70.1||48.3||73.9||90.4||22.6||1,524|
|Rare disease therapies||76||50||73||89||25.3||521|
|Highest approval rate: Hematology||73.3||56.6||75||84||26.1||283|
|Lowest approval rate: Oncology||62.8||24.6||40.1||82.4||5.1||3,163|
Similar results were reached in a recent study by the consulting firm McKinsey & Co., which relied on data from 1996 to 2014 from Informa’s Pharmaprojects database. Tracking the progress of more than 9,200 compounds, the firm found an overall success rate of 11.6% for the past three years. That’s an uptick in what had been a 15-year decline in approval rates, the study found.
Beyond their time frames, the studies differed in whether individual products (McKinsey) or disease indications (BIO) were counted and in how biologics were defined. Using Food & Drug Administration classes, BIO reported an overall likelihood of approval of about 12% for biologics and 6% for small molecules. McKinsey’s looser definition yielded 18% and 9%, respectively.
But both studies found that for any drug type in any disease area, the biggest hurdle to overcome was Phase II, the so-called “proof of concept” stage, in which success seldom rose above 50% and averaged just 31% overall. Phase III, which accounts for most clinical development costs, was especially tough for oncology drugs, the only category to see a much less than 50% chance of getting through.
Moreover, although oncology drugs are about one-third of the development pipeline, they had the lowest success rate of passing from Phase I through approval, BIO found: just 5.1%. In fact, subtracting out this large class increases the overall chances of approval to about 12%. Within oncology, hematologic therapies did twice as well as those targeting solid tumors.
Clinical success rates are linked to an understanding of the diseases being treated, points out Justin Burns, a financial analyst at Sagient Research Systems, part of Informa.
“In hematologic conditions, the scientific understanding and ability to measure blood markers is very advanced,” he explains. Focusing on these surrogate endpoints in early-stage trials can be predictive of later success. “Oftentimes, these surrogate endpoints are sufficient for regulatory application,” he adds.
In fact, BIO saw a 26% likelihood of approval when biomarkers were used to select patients. Only 512 clinical transitions, or 5% of the total database, involved biomarkers.
The BIO researchers didn’t find enough Phase I transitions with biomarkers used to reach statistically significant conclusions. That’s to be expected, Burns points out, because “Phase I trials rarely, if ever, run diagnostic screens for patients with the underlying disease.” But there were enough data in later-stage trials to determine that biomarkers have a positive impact on drug approval rates.
In contrast, in oncology, “response rates, the primary measurement of success in early-stage trials, are not accurate predictors of overall survival, the primary regulatory endpoint of choice that is studied in later-stage trials,” Burns says. Neurology, the second-largest category, is similarly plagued by poorly understood mechanisms of action and a difficulty in treating the underlying disease without serious side effects.
It’s not surprising, Burns notes, that the biggest categories have the least success. “Since the mechanisms of action are not as elucidated for many of these diseases, researchers are constantly attempting novel ways to treat these diseases, which is an arduous endeavor with low success rates,” he says.
More broadly, though, McKinsey suggests that the industry’s shift away from the quantitative “shots on goal” method of developing drugs to more targeted techniques is helping improve the quality of the industry’s pipeline and is behind rising approval rates.
Both studies also broke out results for the burgeoning rare disease drug area. BIO did so to probe whether the probability of success was any better, especially when a particular defective gene is the sole cause of a rare disease. It focused on inborn genetic disorders by subtracting out rare oncology indications and found an overall success rate of about 25%.
In fact, rare diseases were an exception at all phases, with success rates around 50% for the challenging Phase II and well above that for the others. McKinsey reported a total success rate of 29% but indicated that this “might be slightly inflated in our analysis owing to a growing share of potentially life-saving products being expedited straight to Phase III trials through adaptive trials and breakthrough mechanisms.”
Burns points out that expedited approval alone does not directly affect the higher success rates for rare diseases, because the drugs would presumably obtain approval anyway. Because the drugs address unmet needs, “FDA tends to be more lenient in their reviews as well as timelier,” he says.
Time is a consideration, given that it takes a drug more than a decade to progress through the clinic and onto the market. The final stage—filing for and getting approval—alone takes one to two years. Oncology therapies that may benefit from expedited approval pathways get through the fastest, whereas neurology drugs progress the slowest. Ultimately, about 86% of drugs get through the approval stage but only 61% make it on the first try.
In the end, about 90% of drugs in development are destined to fail. Beyond clinical failures, BIO points to funding shortfalls, competition, litigation, and strategy shifts as other factors in ending programs. Confirming the impact of strategy decisions, the McKinsey analysis found the steepest decline in overall success rates occurred from 2008 to 2011, when drug firms were heavily restructuring their R&D organizations.
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society