Reasons for Differences
Estimated phase transition probabilities and eventual approval success rates differed significantly by therapeutic class in Tufts’ study. For self-originated drugs, systemic anti-infective drugs enjoyed the greatest approval rate (23.9%), followed by musculoskeletal (20.4%), antineoplastic/immunologic (19.4%), respiratory (9.9%), GI/metabolism (9.4%), cardiovascular (8.7%), and CNS drugs lowest (8.2%).
“The distribution of disease categories does differ between large and small molecules. In particular, large molecule development is more concentrated in oncology and immunology, while small molecule development is more concentrated on cardiovascular and CNS conditions,” says Joseph A. DiMasi, Ph.D., director of economic analysis for Tufts CSDD. “However, differences in therapeutic class focus do not appear to explain the differences in success rates.”
What reasons, known or suspected, can explain that difference? “This is a difficult question to answer definitively, but it may be that large molecule development, in general, is more targeted, and, given their therapeutic focus and functions (e.g., immunologics and replacement therapies), safety issues may be less prevalent,” Dr. DiMasi speculates.
BIO’s Thomas also points out an interesting result of the BIO/BioMedTracker study: Oncology drugs had the lowest success rate at 5.6%, meaning just one in 20 cancer drugs advanced from Phase I all the way to approval. That rate combines the 11% success rate of lead cancer indications and 2% success rate of secondary indications.
Thomas shared a possible reason for the success-rate differences between diseases: With infectious disease, it’s easy to take a blood sample and quickly generate a readout that tells whether or not your drug is working. In oncology and cardiology, developing a drug is more difficult given the outcome studies required, with multiple efficacy endpoints, not to mention the complexity of the diseases; a cancer drug may work against several different enzymes, versus infectious disease, where a simpler drug targeting a single virus may suffice.
As more biologic drug candidates are discovered and undergo R&D and review, it will be interesting to see how much the gap with small molecule drugs in terms of attrition rates changes. While some narrowing can be expected, the studies to date have all found more biologics advancing through reviews than NMEs.
No wonder then that big pharma is increasingly thinking large, as in molecule, as it scrambles to squeeze more success from the billions it is spending on R&D. Withers & Rogers’ study in particular bolsters the paradigm that big pharma is shifting its development focus from small molecule chemicals to large molecule biologics in recent years.
Among the top 10 pharma firms it surveyed, Novartis had the most patent applications relating to biological drugs published in 2009, with 260—more than twice as many patents for biological drugs compared to small molecules—followed by Johnson & Johnson and Merck & Co. AstraZeneca ranked lowest with just 15 biologicals patents, or 22% of its total. AstraZeneca and Pfizer had less than half of its patents sought for biological in 2009.
There are, of course, far more small molecule drugs on the market. “It is considerably easier to develop and manufacture small molecule drugs. They come with lower R&D costs and there is an established market infrastructure for them. Therefore, the shift to biologics could result in fewer new products making it to market,” Nicholas Jones, Ph.D., partner at Withers & Rogers, explains.
Future studies should examine more closely the type of biologics most apt to achieve regulatory approval—a topic touched on by the Tufts study. KMR’s Martin says that the firm plans to issue another PBF analysis next winter. Research will have to dig deeper to spell out just which type of biologic works best on what disease.