By Earl Gillespie, Engagement Manager; Laura Hatton, Engagement Manager; Carrie Jones, Vice President; and Andrew Funderburk, Partner
One of the major challenges of orphan drug development is proving efficacy given the small number of patients available for trials. As a result, regulatory agencies allow orphan drugs to submit data from fewer patients in part to allow for cost-effective drug development. While companies may benefit from trialing drugs in fewer patients, the per patient costs are often orders of magnitude higher in orphan drug development due to high costs associated with manufacturing small batches of drug product and high recruitment costs, amongst other expenses. This leads us to ask to what extent orphan drug development is truly cheaper than traditional drug development overall, when accounting for these higher per patient costs. Our analysis of trial costs from orphan drug assets from four companies (Amicus, BioMarin, Spark Therapeutics, and Ultragenyx) suggests that while orphan drug development remains an expensive proposition, it is still on average less expensive than trialing a traditional drug.
There are a number of orphan assets in development that have very small development programs in terms of patient numbers, yet these small programs still require significant spending. For example, our sample contains seven enzyme replacement therapies (ERTs) and gene therapies. Each asset has been trialed in a small population (sizes range from 12-33 patients), but has large costs per patient (~$2MM-~$5.7MM) associated with it. These costs likely arise from the costs of manufacturing small batches of active drug product, which are notably high for these types of assets. Ultimately, the median cost for the largest trial (whether Phase 2 or Phase 3) for these assets was ~$100MM (range $42MM-$175MM) despite the small patient numbers, which belies the notion that investments in orphan drug development are small. These figures do not include most preclinical development and formulation costs, which were likely substantial for many of these programs (though not well-delineated by program in public data).
Our study also contains five small molecules, one protein, and one polypeptide. For these agents, trial sizes were generally larger than the ERTs and gene therapies and costs per patient were lower (although still high compared to most other non-orphan indications). Six out of seven assets came in between $137,000-$743,000 per patient, with the last asset costing ~$2.2MM per patient. Because of the larger trial sizes (largest trials ranged from 85-261 patients), the overall costs for the largest trial were similar to the ERTs and gene therapies: median $63MM, range $23MM-$270MM.
For this analysis, Health Advances compared company-reported research and development program costs for individual assets with trials reported on clinicaltrials.gov to estimate trial costs per patient. Patients were distributed evenly over the reported length of the trial (e.g., if a 30-patient trial is reported to run from July 2015 to December 2016, we allocated 10 patient-equivalents to 2015 and 20 patient-equivalents to 2016). If a company broke out internal or personnel costs separate from external program spending, internal costs were allocated based on the ratio of external spending by program.
While our analysis found per patient costs to range from $137,000 to over $5MM in advanced orphan drug trials, an analysis published by PhRMA and Battelle estimated average per patient phase 3 trial costs to be $42,000 based on an analysis of 1,680 phase 3 trials. Estimates ranged from a low of $18,000 per patient in infectious disease trials to $69,000 per patient for oncology trials (PhRMA 2015). While per patient costs are significantly lower for these more traditional drug trials, the large number of patients required usually result in higher overall trial costs when compared to orphan drug development.
An analysis of 42 phase 3 trials conducted by the Tufts Center for the Study of Drug Development, found the median cost of phase 3 trials for traditional drug development programs to be $200MM, with a mean cost of $255MM (DiMasi, et al. 2016). This is significantly higher than the $63-100MM median cost found in our analysis of orphan drug trials. Although the per patient costs are often considerably higher for orphan drug development than that found in traditional drug development, the much smaller trial size does contribute to an overall average lower cost of orphan drug trials, though variance is high in both categories.
In summary, orphan drug development is characterized by very high costs per patient, up to several million dollars for gene therapies and ERTs in very small populations, and per-trial costs that can often exceed $100MM.
|Health Advances Expertise: Developing Orphan Disease Market Opportunities
For more information on commercial planning considerations for rare disease products or this data, please reach out to our orphan disease practice at email@example.com or visit https://healthadvances.com/orphan-diseases/.
For more information on the commercial opportunity for orphan drugs by region, please see our recent blog post: “Orphan Disease Drug Sales – How Much Opportunity is There by Region?”
For additional information about the review process for rare disease products in England, please see our blog post: “Findings from the First Five Years of the NICE HST Program.” It summarizes our ISPOR (International Society for Pharmacoeconomics and Outcomes Research) podium presentation on the first 7 recommendations NICE has made using this pathway for highly specialized technologies for rare conditions, with recommendations for sponsors on how to prepare.
Health Advances analysis, company 10-Ks, clinicaltrials.gov, DiMasi 2016 Health Economics, PhRMA 2015.