Decision Analysis: Commercializing Anti-Vascular Endothelial Growth Factor Monoclonal Antibody or Implantable Medical Device
Master of Health Administration Candidate, 2010
University of Southern California
Price School of Public Policy
Drug/medical device development requires detailed planning and analysis because the commercialization process is time consuming, capital intensive, and highly regulated by the FDA. Therefore, Decision Analysis1 was used to determine the best alternative for SCLERA LLC, the proposed client of this decision analysis for purposes of this assignment. SCLERA LLC has the potential to launch either an anti-VEGF (Vascular Endothelial Growth Factor) monoclonal antibody designed to treat Age-related Macular Degeneration (AMD), or a sustained-release extraoccular drug delivery device designed for the delivery of these anti-VEGF therapies into the eye, in order to treat the same disease.
[All research completed for this article was conducted for purposes of coursework, and not on behalf of SCLERA LLC.]
Although commercializing a monoclonal antibody would result in a higher payoff than launching a medical device, adjusting for outcome probabilities and costs, choosing to commercialize a medical device by first applying for Class I/Class II certification, is a better decision than choosing to launch either an anti-VEGF monoclonal antibody, or a medical device by first applying for Class III certification. Class I/Class II medical devices require lower requirements for review and approval than do Class III devices. Devices are classified as Class I/II if they are “substantially equivalent” to existing devices. “Substantial equivalence” is defined as having the same use and technological characteristics as an existing approved device. To earn approval, a 510k pre-market notification (PMA) must be filed, asking the FDA to certify the device as Class I/II. This process is usually completed over 3-6 months2. On the other hand, however, Class III medical devices require higher levels of critique and observation than do Class I/II medical devices. The applicant must file an “Investigational Device Exemption” (IDE) to earn permission to perform human trials. Assuming that the clinical trials yield favorable results, a PMA may be filed to allow for marketing of the device.2
THE PROBLEM & ALTERNATIVES
Corinne G. Wong, PhD. is the owner, principal, and Chief Scientific Officer (CSO) of a contract research organization, SCLERA LLC. Therefore, she is the sole decision-maker of her company. Dr. Wong has the ability to commercialize two products, but can only commercialize one, due to her time constraints. Therefore, she is looking to launch the product that has the best payoff, relative to risks and costs, over a five year period. The payoff over a five year period was chosen because SCLERA LLC assumes that it will be acquired by a larger biotechnology or medical device company after it has proven successful in that time period.3,4
The first product is a therapeutic agent, an anti-VEGF antibody that targets the blood vessels of patients with Age-related Macular Degeneration (AMD). Anti-VEGF antibodies are used to reverse the detrimental effects of AMD. Genentech Inc. has already captured over 88% of the AMD market share with its product, Lucentis®, because of its superiority. However, Lucentis® is not infallible. The price of the product is very expensive at $2,000 per injection, which must be injected once/month. In fact, Genentech Inc. notes that another one of its products, Avastin®, a much cheaper ($100/injection) biochemical “cousin” of Lucentis®, is stealing revenue away from Lucentis®.5
Therefore, there is a need in the market for a much cheaper product, and SCLERA LLC has the ability to meet that demand. In effect, the payoff of commercializing a therapeutic agent is much higher than that of commercializing a medical device ($233,119,150 for the therapeutic compared to $12,695,438 and $6,702,913 for a Class III medical device and a Class I/Class II medical device, respectively).5,7 However, as lucrative as this opportunity may seem, the financial risks are higher and the probability of success is much lower than if SCLERA LLC was to launch a medical device. For instance, preclinical trials and patenting of the therapeutic agent costs $1,000,000, with only a 80% chance of approval by the FDA. 3,4 Furthermore, the cost of conducting Phase I clinical trials is $14,610,000, with only a 75% chance of approval.1,2 Moreover, Phase II clinical trials are riskier and more expensive than Phase I clinical trials; Phase II trials cost $39.35 million, with only a 50% chance of FDA approval.1,2
Finally, Phase III clinical trials are just as risky as Phase I clinical trials at a 75% chance of FDA approval, but are seven times as expensive, costing $98,310,000.1,2 Additionally, there are many potential competitors that will come to market within the next few years that are much stronger, scientifically, financially, operationally, and managerially, than SCLERA LLC.6
On the other hand, the second product is a sustained-release drug delivery device designed for Lucentis®, Genentech’s AMD product, and other potential AMD products. SCLERA LLC will charge $200 for the device that will release Lucentis® or other anti-VEGF therapies into the eye, over a period of eight months. Therefore, for only $300/year ($200/8 months) and the cost of 1.5 injections of Lucentis®, a patient can observe the same effects as those of eight injections of Lucentis®. Thus, the patient saves nearly $21,000/year. Not surprisingly, the Center for Medicare and Medicaid Services (CMS) will adopt this immediately because it is responsible for 80% of the cost of Lucentis®. Although this device will reduce health care expenditures, the payoff to SCLERA LLC is much lower than that of a therapeutic agent ($233,119,150 for the therapeutic compared to $12,695,438 and $6,702,913 for a Class III medical device and a Class I or II medical device, respectively).5,7 However, the financial risk and probability of failure is lower for a device than it is for a therapeutic ($-400,000 for a Class I/Class II medical device, $-1,900,000 for a Class III medical device and $-153,000,000 for a therapeutic).
When selecting to commercialize the medical device, SCLERA LLC has the option to choose to apply for Class I/Class II certification or Class III certification. Class I/Class II certification is given to those devices that pose less risk to patients, and thus require less oversight by the FDA.9 introduce this concept earlier Class III medical devices, on the other hand, are riskier to the patient, and require more oversight than Class I/Class II medical devices. Class I/Class II certification may be granted by the FDA to SCLERA LLC only in the case that there is another medical device in the marketplace that has already been certified by the FDA as a Class I/Class II product. In fact, there is a product on the market that resembles SCLERA LLC’s sustained-release drug delivery device. If SCLERA LLC chooses to apply for Class I/Class II certification, it must pass the “Substantial Equivalence” (SE) test. This test measures whether or not SCLERA LLC’s product closely resembles that which is in the market already. The FDA has reported that the approval rate for products that apply for Class I/Class II medical devices is 90%.9
The benefits of applying for Class I/Class II certification are that the FDA approval process only requires the filing of a FDA 510(k) document, and not the Pre-Market Approval application (PMA).9 Therefore, SCLERA LLC is not required to conduct clinical trials that demonstrate the safety of the medical device. In essence, SCLERA LLC must only conduct preclinical trials. The drawbacks of Class I/Class II certification are that other competitors may enter the marketplace relatively easily because they only need to file an FDA 510(k) and conduct preclinical trials, which are relatively cheap, and show “Substantial Equivalence” to SCLERA LLC’s medical device. Therefore, the costs for commercializing a Class I/Class II medical device are $400,000, and the probability of approval is 90%.3,4,9 On the other hand, the payoff is $6,702,913.5 This payoff is much lower than that of a Class III medical device, because the barrier of entry for competitors is much lower than it would be if SCLERA LLC’s medical device was Class III certified.
The benefit of Class III certification is that a competitor must file a FDA 510(k) and then a PMA, and conduct clinical trials before it is granted approval to commercialize its product. Completing a PMA is riskier, and requires much more time and energy than completing a FDA 510(k), alone.9 Therefore, SCLERA LLC may enjoy market exclusivity and a high barrier to entry, which is reflected in its revenues of $12,695,438.5 However, the drawbacks to obtaining Class III certification instead of Class I/Class II certification are that approval is much riskier (90% approval of Class I/Class II, and 67% for Class III medical devices and a subsequent 78% for PMAs for those Class III medical devices) and more stringent (FDA 510(k) for Class I/Class II, and FDA 510(k) and a PMA for Class III), and that commercializing is more financially burdensome ($-400,000 for Class I/Class II, and $-1,500,000 for Class III). The results describe how decision analysis was critical to evaluating the best alternative for SCLERA LLC.
Decision analysis suggests that SCLERA LLC would earn a better payoff by choosing to commercialize a Class I/Class II medical device than by choosing to commercialize a Class III medical device or a therapeutic agent. The reasoning behind the decision is that although there is potential for market penetration by competitors, the probability of success (90% approval) is substantially higher than that for a Class III medical device (67% for FDA 510(k) for Class III and a subsequent 78% for PMA approval). Additionally, in the case that SCLERA LLC’s medical device does not obtain Class I/Class II certification on its first attempt, it has two options. Primarily, it may file a de Novo Petition (not demonstrated in the decision analysis due to lack of probability/statistical information) that challenges the FDA’s decision. On the other hand, SCLERA LLC may also opt to file a FDA 510(k) for Class III certification. Therefore, the payoff for choosing to apply for FDA 510(k) Class I/Class II certification is higher, $6,568,785, than choosing to apply for FDA 510(k) Class III certification, $5,361,636.
On the other hand, the payoff for choosing to apply for FDA 510 Class I/Class II certification is higher, $6,568,785, than that for choosing to commercialize the anti-VEGF therapeutic, $6,470,809. The reasoning for this is that although the payoff after the completion of Phase III clinical trials is $233,119,150, after adjusting for the costs of preclinical and clinical trials ($153,000,000) and the probabilities of outcomes at each level (preclinical-80%, Phase I-75%, Phase II-50%, Phase III-75%), commercializing a medical device and applying for Class I/Class II certification is a better option for SCLERA LLC. However, since the difference between the two outcomes is less than $100,000 ($6,568,785-$6,470,809), if SCLERA LLC were to negotiate a higher up-front offer from an established pharmaceutical/biotechnology company (Licensee) to sell SCLERA LLC’s product (Licensor), the better decision may be to commercialize a therapeutic agent.
RELEVANCE OF DECISION ANALYSIS
Decision Analysis was used to measure the payoff for the alternatives outlined above. This analytical tool has much relevance to the pharmaceutical/biotechnology/medical device industry, as it has been used in the past by many pharmaceutical companies, including Bayer Pharmaceuticals.2 Because drug development is time consuming, capital intensive, risky, and highly regulated by the FDA, companies in this industry use decision analysis as opposed to intuition alone. In fact, Decision Analysis is used to determine the feasibility and market potential of each of a company’s product candidates.2
Data, including statistics, financial data regarding clinical trials, and outcome probabilities are collected at each level of development in order to get the most precise value for each alternative. Outcome probabilities, costs of each step in the commercialization process and the payoff of each alternative required extensive research and referencing of 10-K reports of relevant companies, primary sources such as interviews with industry experts, and data from publications within the pharmaceutical/biotechnology/medical device industry. Since each step of the clinical trials process has its own cost or payoff, and is contingent on the success/approval of the previous step, Decision Analysis is a valuable tool used to determine the value of each alternative.
VALIDITY, ASSUMPTIONS and LIMITATIONS
Although the recommended solution may seem feasible and logical, there are many uncertainties that challenge its credibility. For instance, the costs of clinical trials for the therapeutic are an average of all those completed to date.i There was no access to Regeneron Pharmaceuticals Inc.’s clinical trials costs, which would have given a more precise estimate of SCLERA LLC’s costs of conducting clinical trials. The reason is that the costs of the test drug and the control drug are similar to those which SCLERA LLC would use in its clinical trial. On the other hand, the probabilities of the outcomes associated with each phase of the clinical trial process are also subject to inaccuracy.ii
The propriety nature of the product precludes access to the FDA’s data on the outcomes of clinical trials for anti-VEGF therapies, then the probabilities for each phase of the clinical trial process would be more precise. Additionally, the costs and probabilities of approval and disapproval of the results of pre-clinical trials (In Vitro/In Vivo) for the therapeutic agent are estimates by expert scientist, Dr.Corinne Wong, and veteran venture capitalist, Dr. Frederick Haney.iii,iv
Most importantly, the outcomes/payoffs for the approval of the therapeutic are based on many assumptions. Primarily, the market size assumes a 10% compounded annual growth rate. This assumption is based on the evidence provided by the increase in sales of Lucentis, Genentech Inc.’s drug that for AMD treatment.v This rate also assumes the possibility of the entry of competitors in this market place. The entry of competitors is also reflected in the market share of the licensee (the company that which SCLERA LLC chooses to license its therapeutic agent).vi However, the % of revenue due to SCLERA LLC is quite accurate, as shares of revenue are usually in the 5-10% range.5 Finally, the payoff for the therapeutic agent also assumes that the pharmaceutical/biotechnology company to which SCLERA LLC will license its product, will pay an up-front cash payment of $180,000,000.7 This was assumed because it was the amount of money given to Regeneron Pharmaceuticals, Inc. by Bayer Pharmaceuticals, Inc. upon FDA clearance.vii
Moreover, there are also limitations to the credibility of the probabilities and outcomes/payoffs for the commercialization of the medical device. For example, the payoff for class III approval by the FDA is contingent on a few assumptions. The number of patients in the market, which is critical in indicating revenue from the device, was based on the market size of the respective year. The number of patients in the market was calculated by dividing the market size for the given year by $24,000. The reason for this is that each patient spends $24,000/year on Lucentis ($2,000/month), and it comprises over 88% of the market share. Since SCLERA LLC will choose to distribute this product itself, it will not require any licensing. However, the market penetration rate assumes that barriers to entry of a competitor or a similar product will be very high. With SCLERA LLC’s device as a Class III product, competitors would need to file Pre-Market Applications (PMAs) and spend roughly $1,500,000 on clinical trials before they enter the market place.3,4 This barrier to entry is reflected in SCLERA LLC’s market penetration rate over the first five years of commercializing the product. This rate was a rough estimate based on the idea that the Center for Medicare and Medicaid Services (CMS) would adopt the product in order to save the money that it was spending every month to treat AMD patients.
On the other hand, if SCLERA LLC was to issue its product as a Class I or Class II medical device, then the barriers to entry would be low. In this case, a competitor would need only to file the FDA 510(k) document, which takes less time to be completed and approved than would the PMA.viii Competitors would only need to show “Substantial Equivalence” (SE) to a Class I or Class II medical device before it is commercialized.ix The market penetration rate given that SCLERA’s product is a Class I/Class II medical device reflects the fact that competitors will enter the market in 2011 (year 3). This estimate is rough, and is subject to high variance. However, in the case of a Class I/Class II or a Class III medical device, the probabilities are very accurate, as they were obtained from the U.S. Government Accountability office.9
iStonebraker, Jeffrey S. “How Bayer Makes Decisions to Develop New Drugs.” INTERFACES 32 (2002): 77-90.
ii DiMasi, Joseph A., Ronald W. Hansen, and Henry G. Grabowski. “The price of innovation: new estimates of drug development costs.” Journal of Health Economics 22 (2003): 151-85.
iii “Interview with Dr. Fred Haney.” Telephone interview. 1 Apr. 2009.
iv “Interview with Dr. Corinne Wong.” Telephone interview. 2 Apr. 2009.
v Genentech 10-K report.
vi Waltz, Emily. “Companies eye slice of age-related macular degeneration market.” Nature Biotechnology 24 (2006): 1041-042.
vii Regeneron Pharmaceuticals Inc. 10-K report.
ix United States. United States Accountability Office. Medical Devices: FDA Should Take Steps to Ensure That High-Risk Device Types Are Approved through the Most Stringent Premarket Review Process. Government Accountability Office, 2009.
The majority of the sources above have been reviewed by the Securities and Exchange Commission (SEC). Additionally, the data on outcome probabilities associated with each stage of the trial are also accurate. On the other hand, clinical trial costs are reliable, but subject to little variance due to the fact that they were evaluated in terms of dollars in the year 2000.1 However, these costs were adjusted for inflation. The interviews with Dr. Fred Haney and Dr. Corinne Wong are also subject to possible variance due to the fact they are expert opinions.