Economic Analysis of Availability of Follow-on Protein Products. VI. Study and data limitations

07/27/2009

A. General Limitations

This study was completed in two phases (i.e., selection of candidates and economic analysis). While there are limitations specific to each phase, there are also limitations that applied to the entire study. As in any model, there is an inherent difficulty in predicting the future. Determining which candidate biologics manufacturers will choose to pursue with FoPPs and how many of these FoPPs will be produced for each originator drug is an imprecise evaluation subject to many factors. Unforeseen future changes (e.g., in patent extensions, additional indications covered, advances in technology) could affect the attractiveness of an originator drug and the ability of a company to create a FoPP.

Similarly, there is currently no approval pathway for FoPPs under the PHS Act. (Omnitrope® and glucagons were approved under the FDCA.) Delays in creation of an approval pathway would affect the outcome of our analysis, pushing back market entry for some of the first FoPPs by several years. Even if one of the proposed bills were approved, variations between these bills with regard to the period of market exclusivity for the first FoPP, jurisdiction for determinations of interchangeability, and the level of evidence necessary to make a ruling of biosimilarity would affect outcomes predicted by our model. Additionally, the stringency of regulations, once approved, may make it more difficult to produce FoPPs for some types of biologics compared to others. For example, the case-by-case approach of the EMEA requires different levels of evidence for different biologic products. A similar model in the US might deter manufacturers from pursuing a FoPP for biologics, requiring higher standards of evidence for biosimilarity due to the associated costs of production and clinical trials.

B. Limitations Related to the Selection of Candidates

Many of the limitations of our study are specific to the selection of candidate biologics. The lack of available and consistent information about patent expiry dates was one of the limitations to our study. Patent expiry dates, which were drawn from market research reports, public corporate documents, and other sources, are often inconsistent. This derives in part from inherent uncertainty of intellectual property law and claims, various court decisions, and business decisions.[149]

Similarly, due to the complexity of biologics, there are generally several patents protecting a manufacturer's exclusivity rights for any given drug. While the patent protecting the drug itself may expire in a given year, the formulation, technology involved in manufacturing, or cell line used to create the biologic itself may not expire until later. Patent challenges also make patent expiry a variable that is difficult to determine for this analysis. Successful patent challenges by generic manufacturers could potentially open the market to FoPPs several years ahead of the projected patent expiry, while patent extensions would delay the introduction of a FoPP. This ambiguity makes it difficult to accurately select the most likely candidates for FoPPs based on patent information.

In our discussion with experts, there was no consensus as to what the most important selection criteria for a candidate biologic should be. While some argue that the complexity of the molecule will limit the number of FoPP entrants to the market, others claim that, if the market is large enough, generic manufacturers will find a way to overcome scientific hurdles. The lack of consensus regarding the ranking of selection criteria limits our study, as we may have chosen some candidate biologics for our analysis that may have been excluded given different criteria.

C. Limitations Related to Economic Analysis

As in all analyses of this type, the validity of the final estimates relies on the validity of the underlying assumptions. It is because of the high degree of uncertainty that we performed and describe the results of several sensitivity analyses.

We highlight the key role played by several assumptions here. First, our estimates of market size beginning in 2012 are a simple linear extrapolation of market projections over the period 2008-2012. To the extent that growth potential in these markets is under- or over-estimated, our projected cost savings will also be under- or over-estimated. We explore a related matter in Section 6.5, where we model the effects of a "second-generation drug," which significantly reduces market revenue for the specified drug and associated FoPPs.

Second, as noted in earlier sections, information on patent expiry is difficult to find and not always consistent; as a result, our estimates of market opening to entry are also subject to uncertainty. Here too, we investigate the effect of assuming markets opening earlier and later than our base-case estimate.

Additionally, the first step in our model is the estimation of market entry, which depends heavily on the estimated fixed costs of entry to each market. Due to the lack of empirical data, this estimate is calibrated within the model to produce an estimated number of entrants with face validity. However, the method of calibration is also subject to uncertainty and we therefore perform multiple sensitivity analyses on the fixed cost of entry estimate and the projected number of market entrants.

Finally, our models of market entry, market pricing, and market uptake are based in part on studies performed in the small molecule market, which differs in important ways from the biologic market. Although we have attempted to adjust our estimates to account for these differences, there is still uncertainty as to their applicability. Therefore, we perform additional sensitivity analyses in which we vary each of the parameters in turn and explore the effect on the projected cost savings. Even so, these variables interact in complex ways, and single-way deterministic analyses do not fully account for these potential differences.

View full report

Preview
Download

"report.pdf" (pdf, 575.91Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®