Commercialization is the process of moving inventions from discovery to product. In an academic setting, this process invariably involves technology transfer. The discovery may come from a researcher at university or hospital, but over time the center of gravity regarding the transformation of that technology into a product shifts to a company setting.
The path to commercialization is quite different for every technology, even within the healthcare/life sciences industry. This module focuses on three key areas: biotechnology and pharmaceutical discoveries, medical devices, and diagnostics. In any scenario, a technology will need to attract the resources required to bring the technology to market. Investors and companies regard technologies differently based on their stage, development risk and cost, and the size of the market they address.
Key considerations that determine the commercial value of inventions include:
- Market size - annual market for the proposed product
- Market structure – how difficult will it be to address that market and how concentrated are the users of the product (for example, primary care versus specialized surgery)
- Regulatory pathway
- Clinical trial design – time, expense, ability to recruit patients
- Strength of intellectual property
- Reimbursement pathway
- Potential adoption curve
- Degree of unmet medical need
- Stage of development of the technology
By far the longest time to market and most capital intensive of the three areas, drug discovery is also the most widely invested area given the high potential for financial reward. A typical drug development program can take ten years and hundreds of millions of dollars for FDA approval. A telling statistic regarding the difficulty in drug discovery is that 50% of drugs that enter Phase III clinical trials (the last step before FDA approval) fail. The attrition rate is higher at earlier stages, and is extremely high for pre-clinical compounds. The high failure rate and capital intensive nature of drug discovery drive licensing and funding decisions around specific areas:
- Markets – Markets must be large (in dollar terms) relative to the investment required to bring the drug to market. In general, markets smaller than $1 billion potential would require something special about the technology that would either reduce the risk or development cost of the potential therapeutic.
- Market Structure – How will the drug be detailed? Are the likely prescribers reachable at a reasonable cost? Drugs that target primary care providers are the most difficult.
- Competition – How well served is the market? If a market is served even modestly well, considerable clinical experience with the drug will be required to drive physician adoption.
- Regulatory pathway/clinical trial design – This is usually the most important concern in drug development. What does a trial design look like for FDA approval? What will be the time and expense to get to some data on clinical efficacy? What preclinical work is required? While there are always exceptions, a good rule of thumb is that most venture capitalists and companies will not invest in products that are more than two years away from IND. Trials that require outcomes data (cardiac indications, cancer recurrence are examples) are difficult and expensive as they require large numbers of patients followed over a long period of time.
- Intellectual property – Composition of matter is by far the strongest type of intellectual property and is a requirement in many cases. Novel formulations can be acceptable but there is a general feeling that these can be engineered around. Patents showing a new use for an approved drug (that do not require a reformulation for the new use) are of little value, because substitution is too easy and companies would never sue prescribers for patent infringement.
- Reimbursement pathway – Less of a concern for most drug discovery programs. Drugs that show modest improvements over a product that is already generic may have difficulty getting coverage.
- Adoption curve – Dependent on unmet medical need. For example, a new statin or treatment for rheumatoid arthritis would likely have a very slow adoption curve as clinicians are generally satisfied by existing alternatives.
- As mentioned previously, it is generally hard to find investors for drug discovery programs that are more than two years from IND. Demonstrated efficacy in relevant animal models is required, and targets or assays that have not been validated in vivo have little value.
In general, a rule of thumb is that a startup in the field of drug discovery will require an investment (over multiple rounds) of $100 million to reach an exit for investors. This exit (through IPO or sale to a larger company) generally happens years before product approval. Newer models that are focused on a single product development require less capital but have an increased risk (a single shot on goal). This model has been adopted by a few venture capital funds but remains unproven. When evaluating potential partnerships with a biotech company, large pharmaceutical companies generally prefer to see at least Phase IIb data.
Medical devices refers to a very broad range of technologies, ranging from complex, invasive devices with a high burden of demonstrated safety and efficacy to improvements on existing devices with a relatively simple pathway to approval. Unlike drug discovery, the FDA approval process is generally not a significant burden (unless the device is so novel it requires a Pre-Market Approval or PMA, which is a long and expensive path). FDA approval is generally not sufficient to drive adoption, however, and a greater focus is placed on issues such as reimbursement. A new technology must also have enough clinical data to demonstrate to clinicians that they should change their practice of medicine and adopt the new device – this burden is usually far higher than that required by the FDA.
The medical device industry generally uses an outsourced R&D model. The larger companies tend to innovate in new areas by acquiring smaller startup companies. While there are device companies that are acquired based on technology promise alone, the large majority of companies need to demonstrate significant sales before becoming attractive to potential buyers. The sales ramp-up is usually far more capital intensive than the technology development phase of a new company’s existence. Total investment required from company formation to exit is generally over $30 million and requires three to five years.
- Markets – Markets must be large (in dollar terms) relative to the investment required to bring the device to market. In general, markets smaller than several hundred million dollar potential have difficulty attracting capital. Investors also are concerned about the perceived value of each procedure, and as a result dramatically favor therapeutic devices over diagnostic devices. Devices that have a disposable component are considerably more attractive than those that do not, given the ability to resell to the same customer repeatedly.
- Market Structure – Extremely important in medical devices due to the need for the small company to build a sales force. Highly concentrated procedures (but still with enough customers) are ideal as the company can reach a large percentage of the market with a relatively small sales force.
- Competition – How well served is the market? How difficult will it be to displace existing equipment? What will the competitive reaction be from existing suppliers?
- Regulatory pathway/clinical trial design – Investors prefer devices that only require 510(k) (substantial equivalence) FDA approval. Pre-Market Approvals (PMA) are a long and expensive process, and require a very attractive new market (drug-coated stents, for example). Given the general use of 510(k), regulatory requirements are not onerous and are not seen as a major risk.
- Intellectual property – Broad claims are rare. Need to be able to defend against direct competition, but there is a general understanding that intellectual property will only slow down, but not block, new entrants.
- Reimbursement pathway – Often the path to reimbursement with clinical use is the most important area of diligence for new medical devices. Companies submit for new reimbursement codes around the same time they submit their 510(k), but new reimbursement codes often take two years. If no reimbursement is possible in the short run, market launch is very difficult. Ideally, codes that can be used already exist. Devices that are reimbursed under general procedure disease-related group (DRG) codes require significant data to make the economic case to a provider that outcomes are improved and costs reduced by adopting the new device.
- Adoption curve – Dependent on unmet medical need, clinical evidence, and reimbursement status. This phase of a startup consumes the most capital and slow adoption is the most common cause of failure of new medical device companies. Capital equipment is particularly difficult, given their very long sales cycles.
- Stage – Devices should have at a minimum an early prototype with some demonstration of in vivo efficacy.
Drug/device combination products are becoming increasingly common. The FDA treats the approval pathway for a new product based on its view of the primary mode of action. Almost all device-centric combination products use existing, FDA approved (and usually generic) drugs. Use of a novel compound would require an approval process similar to an new drug.
Novel diagnostics (along with healthcare information technology) is one of the more difficult categories of technologies to commercialize. With important exceptions, they are seen as difficult to bring to market (large and increasing regulatory and reimbursement hurdles and burden of clinical evidence) and lower value (lower reimbursement compared to therapeutics). Many venture capital funds have policies not to invest in any diagnostic technologies, due to the structure of the market. Despite these odds, there have been some dramatic successes in recent years.
Markets – Markets must be large (in dollar terms) relative to the investment required to bring the diagnostic to market. In general, markets smaller than several hundred million dollar potential have difficulty attracting capital. Investors favor technologies that characterize known or suspected disease over population screening technologies, due to the reduced statistical challenges. Results from the new diagnostic test must change medical decisions rather than confirm suspected disease without a change in care based on the test results.
- Market Structure – Similar to medical devices, highly concentrated procedures (but still with enough customers) are ideal as the company can reach a large percentage of the market with a relatively small sales force.
- Competition – How well served is the market? Often the strongest competition is the judgment of the physician.
- Regulatory pathway/clinical trial design – Regulatory requirements are very important in diagnostics. How will the new technology demonstrate to physicians that there is sufficient evidence to base care decisions on the new technology. Companies that provide tests as a reference lab (“home brew”) do not require FDA approval, although the FDA is tightening these regulations. Companies that provide kits require 510(k) or PMA approval.
- Intellectual property – Similar to drug discovery, fundamental, defensible intellectual property is required to attract investment.
- Reimbursement pathway – As with devices, a key to a diagnostic company success is the potential for reimbursement for clinical use. Reimbursement decisions often follow not only FDA approval, but inclusion in care guidelines by the relevant physician organization. For screening tests, there is a medical economics requirement: tests must show that they are able to cost less than a certain amount per quality year of life saved.
- Adoption curve – A key area. Inclusion of a test in treatment guidelines by relevant physician organization can be the single most powerful driver of adoption for new diagnostic tests.
- Stage – Retrospective analysis of market data is usually not sufficient. Some prospective data is generally required, although in certain cases a blinded retrospective study can be sufficient as a start.