DEVELOPMENT TIMELINES – Drug Development Times, What it Takes – Part 3
This is the third and final article analyzing the Development and Review Times of the US FDA’s new drug approvals (NDA and BLA) for the period 2010 through 2018. The first article1 provided an overview of development and review times while the second article2 looked at New Molecular Entity (NME) product approvals. This article looks at the development and review times associated with the larger and much more mixed group of products based on Previously Approved Actives (PAA).
PREVIOUSLY APPROVED ACTIVE APPROVALS
NDA approvals for products incorporating PAA are assigned NDA Classification Codes3 by the FDA. Table 1 summarizes the products available for Development and/or Review Time analysis. Unfortunately, not all approved products can be analyzed because of missing or redacted information in the FDA’s published review documents. None of the products in this analysis are Abbreviated New Drug Approvals (ANDA).
All PAA products are categorized as 505(b)(1) or 505(b)(2) approvals. A 505(b)(1) submission requires the sponsor to have complete rights of reference to the underlying active ingredient. This is generally the case when the filing is from the originator or a licensee. A 505(b)(2) submission does not have the same rights of reference to the previously approved pharmaceutical active. A total of 149 PAA products were 505(b)(1) approvals, the balance were 505(b)(2) approvals.
The Review Time analysis includes 545 PAA products. Mean and median Review Time averages by Type are presented in Table 2 (Categories with less than 20 approvals are excluded).
A quick review of the Review Times summarized in Table 2 suggests there is little difference between the median Review Times for Previously Approved Active and New Molecular Entity products. There is a difference in the mean Review Times of about half a year in favor of NME approvals that may reflect the greater regulatory experience of companies developing NME products. There is little reason to believe that the regulatory review hurdles and timelines for PAA products should be greater than for NME products, although NME products are more likely to receive some sort of accelerated review. That the shorter Review Times are related to regulatory expertise is suggested by the one-third year difference in the median review times for PAA products that were approved through the 505(b)(1) or 505(b)(2) process. The PAA 505(b)(1) approvals are more likely to be associated with companies that originally developed the active. Review delays were noted in the approval documents for 28% of the 505(b)(1) and 43% of 505(b)(2) filings, further suggesting the greater regulatory experience of companies submitting 505(b)(1) applications.
The figures in Table 3 reveal the considerable difference in Development Time between NME- and PAA-based products of 3.5 years or more (mean and median). The median values perhaps best represent the most realistic benchmark regarding the time spent in the clinical stage of development for a product using a Previously Approved Active. While there is limited opportunity to significantly shorten the clinical development process, there is a seemingly unlimited opportunity to extend it.
The mean average and median Development Times for PAA product approvals are remarkably consistent as a function of Type. The exception is found with Type 4 Combination Products that have required about a year less. This may simply be a function of relatively more Type 4 products (41%) being developed and approved on the basis of bioequivalence programs.
The figures in Table 4 suggest why, on average, Development Times for New Molecular Entity products are much greater than Previously Approved Active products. While NME approvals require full efficacy and safety studies, PAA products can be approved on the basis of more limited efficacy and safety studies, or even just bioequivalence studies.
PAA products approved solely on the basis of bioequivalence data complete the clinical development portion of their development on average in 3 years, with a median of 2.6 years. This is half the 6 (mean) and 5.1 (median) years, respectively, of PAA products developed on the basis of efficacy and safety indicating trials. PAA products approved on the basis of efficacy and safety trials still require about 2.5 years less Development Time than NME products.
The differences noted earlier between 505(b)(1) and 505(b)(2) review times are also seen with clinical development times. 505(b)(1) programs show as much as a one-year shorter Development Time.
PUTTING IT ALL TOGETHER
The numbers paint a rather consistent picture as presented in Table 5 that compiles median Development and Review times for NME and PAA products. The differences matter not only when estimating the overall timeline to complete a clinical development program through to approval, but also when deciding what path to take when developing a Previously Approved Active.
The factors underlining these timelines are not restricted to clinical development competence or bad luck during the development process. Equally important are issues, such as funding gaps that force development pauses, unexpected issues with patient recruitment, or an unexpected difference of opinion with the FDA on trial design and patient numbers.
Clinical development and regulatory plans will always suggest that products can be developed in less time than these averages. Of course they can, but on average they don’t. Any company or team suggesting a development program will be completed more quickly than the benchmarks of Table 5 needs to clearly explain how they expect to “beat the averages.”
About a year ago, I was speaking to a former colleague about an interesting new product development his small company was working on with a previously approved active for a well understood and established indication. The product offered a clinically interesting twist on the usual treatment paradigm. The product also had positive efficacy data by virtue of earlier investigator IND work. Based on their clinical development plan, which would include efficacy and safety trials, they expected to be submitting their product to the FDA in 2 or 3 years. My immediate response, without reviewing their development plan, was that they were seriously underestimating the timelines. Was I questioning their competence? Of course not, these are top notch folks. Experience suggested that the time would be double or more what they were estimating. Catching up a year later, I was told that a few months after our discussion the FDA had come back with a requirement for more preclinical toxicology data that would delay the program by a year and add in more expenses. I suspect this is only the start of the surprises that will hopefully not delay the program beyond the medians in Table 5. Stuff happens, even to the best.
The Development and Review times presented in these three articles are not the final word. They should be the starting point for discussions and the stress testing of development plans. There are two key questions that need to be asked. How do we plan to beat the averages? What will it cost to achieve this exceptional, average beating, performance?
There are many lessons to be learned in the Product Approval documents available at the FDA website. It takes a bit of work to go through the many documents, but like a good detective story, there are clues that point to important evidence of what happened and why. Understanding how others came to be where they are is an important step in understanding what is ahead of us. “History doesn’t repeat itself, but it does rhyme.”4
- NDA Classification Codes. https://www.fda.gov/media/94381/download.
- Generally attributed to American humorist Mark Twain (Samuel Clemens).
Dr. Josef Bossart serves as Managing Director at The Pharmanumbers Group. He has 4 decades of experience in the biopharmaceutical industry, having held senior sales, marketing, operations, and business development positions within Big Pharma and emerging Specialty Pharma companies. His activities include analyzing corporate, technology, and product development strategies in the area of Drug Delivery. Dr. Bossart earned his PhD in Medicinal Chemistry from The Ohio State University.
Total Page Views: 3153