By Kevin E. Noonan --
Professor Henry Grabowski has been cited often in the data exclusivity debate, on the economic justification for a twelve-year data exclusivity period for biologic drugs. Dr. Grabowski (at right), a Professor at The Fuqua School of Business at Duke University, has some credibility (in addition to his academic credentials) for having the only peer-reviewed analysis of the topic (his paper, "Follow-on biologics: data exclusivity and the balance between innovation and competition," was published online in Nature Reviews Drug Discovery on May 12, 2008). Opponents cite his support by the Pharmaceutical Research and Manufacturers of America (PhRMA) as raising issues of impartiality (albeit in the face of frankly partisan contrary opinions that have not faced peer review). Under the circumstances, it is important to the debate that everyone be familiar with the assumptions, analysis, and conclusions presented by Professor Grabowski.
First, the assumptions. Professor Grabowski posits the following realities that may be open to dispute but have sound bases in the experience of innovator biotechnology and pharmaceutical companies:
• "Without a data exclusivity period, there would be little incentive to invest in developing and marketing new product candidates with few remaining years of patent protection or with uncertain forms of protection."
• "[N]ewly approved products with substantial commercial sales would be exposed immediately to legal risks associated with patent challenges and early entry of generic versions."
• "Data exclusivity assumes particular importance for biological entities as compared with chemical entities because many of these products rely on narrow patents that make them more vulnerable to challenges from follow-on competitors."
• "As follow-on biologics will be comparable but not identical to the innovator's molecule . . . they may avoid infringing the innovator's core patents, while still being able to gain regulatory approval through an abbreviated pathway."
• The mean R&D costs for bringing a biologic drug to market is $868 million, with biologic drugs relating to oncology cost an average of $1.016 billion. "[C]apitalized R&D costs for a representative NBE range from $1.24 billion to $1.33 billion when the real cost of capital is 11.5-12.5%," citing DiMasi & Grabowski, 2007, "The cost of biopharmaceutical R&D: Is biotech different," Manag. Decis. Econ. 28: 469-79.
• The "nominal cost of capital" for biotechnology companies is "approximately 15%," assuming an inflation rate of 3.5-5%.
• "[L]onger exclusivity times encourage increased development of NBEs and NCEs as well as additional research on new indications for established products."
• "[F]or industries in which the R&D process is costly and risky, longer exclusivity periods to realize innovation benefits are needed in comparison with those industries in which innovation is easier and less costly."
• "It is common for the development of an NBE to originate in a start-up company financed through venture capital financing."
• "At the initial stages of development, there is a high degree of scientific risk associated with proof of concept." (citing U.S. FDA Center for Drug Evaluation and Research (CDER) special report: "From test tube to patient: improving health through human drugs").
• The overall probability of success in clinical development of an NBE (using a survey of 522 biologic drug candidates) is 30% "(that is, the success rate of candidates that make it as far as trials in humans)," citing DiMasi & Grabowski, 2007. Paradoxically, the article asserts that "[w]hile biologics had a higher overall success rate than chemical drugs, they have had lower success rates in the most expensive Phase III clinical trials . . . after high development costs have been incurred."
• The most commercially-successful biologic drugs are generally "best in class" or "first in class" and are expected to be "the primary targets of generic biologic firm."
• "There is accumulating empirical evidence that new medicines and therapies have played an important role in increased longevity, enhanced quality of life and improved labour-force productivity," citing Lichtenberg, 2001, Int. J. Health Care Finance Econ. 5: 47-73; Nordhaus, 2003, Measuring the Gains from Medical Research: An economic approach; Calfee, 2007, The American March/April, pp. 41-52. "Furthermore, recent studies have found that consumers have appropriated significantly more of the societal benefits than innovators in the case of new therapies for HIV/AIDS, as well as several other new technologies," citing Phillipson et al., 2005; Nordaus et al., 2004.
• NBEs are capable of being investigated for different illnesses related to the same biological pathway, using TNF inhibitors as an example. However, "[t]he development of new indications for established biologics would be particularly vulnerable to early patent challenges by generics firms . . . because obtaining approval for a new indication post-approval can take several years and involve . . . significant costs. The uncertainty surrounding early patent challenges may tilt the risk-return balance against otherwise economically viable investment programs . . . [and] patients would be deprived of health benefits from new indications."
With these assumptions in place, the article addresses some of the developmental realities of bringing biologic drugs to market. The article reports that average development times for biologic drugs have more than doubled, from about 44 months in the period from 1982-1989 to more than 110 months in 2005-2006. The article uses Avastin® (bevacizumab) as an illustrative timeline from compound development to commercial introduction. The article uses the discovery of vascular endothelial growth factor (VEGF) in 1989 as a starting point and FDA approval 15 years later in 2004 as a standard, having an IND filing in 1997, and Phase I (1997), Phase II (1998) and Phase III (2000) clinical trials as benchmarks for how long regulatory approval takes. Professor Grabowski's model assumes an average of 4-5 years of preclinical development, followed by 8 years of clinical trials, and involve 3.3 clinical trial candidates for every approved drug.
Looking at the effects of these considerations on the "break even" point for drug development, the article states that the mean break-even time for NCE drugs is 16 years, based on NCE drug introductions between 1980-1984, defined as the point "where the present value of cumulative after-tax cash flows just intersects the present value of after-tax R&D investment . . . signifying the fact that the firm has recouped its investment plus a return equal to the industry's average cost of capital for that period," citing Grabowski & Vernon, 1994, J. Health Econ. 13: 383-406. For drugs developed between 1990-1994, the time was 15 years, citing Grabowski et al., 2002, PharmacoEcomonics 20 (Supp. 3): 11-29. "In contrast, the average market exclusivity periods . . . for new molecular entities [experiencing generic competition] in the 1996-2005 period generally fluctuated between 12.5 and 15 years." (And only 30% of NCE drugs have revenues that exceed the cost of R&D, a feature of the "blockbuster" drug paradigm where these ultra-successful drugs "pay for" the remaining drugs that are less profitable). Accordingly, Professor Grabowski analyzed the "break even" point for biologic drugs using a "portfolio" model, where there is one "blockbuster" biologic drug product and a number of less commercially-successful drugs in the portfolio. Using this approach, the model predicts that "[s]ales increase at a rapid rate during the early years of the life cycle, reach maturity [at around 9.8 years] and then slowly decline due to product obsolescence." The results of competition from biogenerics would be a "much steeper decline in sales," although this is not included in Professor Grabowski's analysis, which is directed to determining the optimum data exclusivity period without generic competition.
The results of this analysis, widely reported, is that the "break even" point for biologic drugs requires a data exclusivity period from between 12.9 years (assuming a discount rate of 11.5%) to 16.2 years (assuming a 12.5% discount rate). Professor Grabowski cautions that these discount rates are conservative and that "smaller publicly listed biotechnology companies and non-listed private biotechnology firms would generally have a much higher cost of capital," suggesting that the actual "break even" point may require even longer data exclusivity periods. "[I]t must be kept in mind that there are few biotechnology companies that are profitable, and the universe of biotechnology firms is populated with developmental-stage companies whose principal assets are their human capital and intellectual property," which companies would be expected to "realize higher costs to launch a new product" than a more established firm (such as a traditional pharmaceutical company).
Professor Grabowski identifies two different types of errors that can confront the decision-making process in this area. First, "[i]f data exclusivity periods are too short, new product candidates with inadequate or uncertain patent protection will be deterred." "On the other hand, if data exclusivity periods are too long, price competition could be delayed beyond what is necessary to encourage innovation." He concedes that "the United States Congress [will need] to balance the objectives of innovation incentives and price competition" for biologic drugs, just as it has for NCE drugs with the Hatch-Waxman Act. He concludes with the following recommendation:
Legislation on follow-on biologics should be designed to strike a balance between the incentives for price and innovation competition. In particular, the legislative bills without any provisions for a data exclusivity period, or only very nominal periods of exclusivity, would have adverse effects for new biological innovation activities. Under these legislative scenarios, there would probably be an explosion in patent challenges shortly after a new product is introduced. While the right to undertake patent challenges is an integral part of the US intellectual property system, entry through abbreviated filings should be delayed until the representative NBE has had an opportunity to earn risk-adjusted break-even returns. This important concept for innovation incentives is incorporated in the US legislative proposals that provide for a substantial period of data exclusivity.