IVD manufacturers must overcome numerous challenges to develop companion diagnostics for targeted drug therapies.
The cancer market is presently the most advanced area with regard to personalized medicine. This article reviews the development of cancer companion diagnostics and identifies options for the IVD industry to capitalize on this emerging opportunity. The major challenge is the cost of bringing a companion diagnostic to market, which must be justified by the value of the information from the resulting product. The key to a successful outcome is to recognize who will benefit from this information and use this understanding for strategic decision making.
Personalized medicine aims to target drugs to individuals who are likely to respond well to the treatments. It is most useful when the commonly used trial and error approach to drug therapy is ineffective due to the high cost, high risk, or low efficacy of a particular therapy. The type of drugs that best meet the criteria for personalized medicines are the new generation of targeted cancer therapies. Such drugs block the growth and spread of cancer by interfering with specific molecules involved in carcinogenesis and tumor growth.
Since cancer has a variable etiology, such targeted drugs will only be effective in tumors in which their target is involved in the disease process. Furthermore, since such treatments show moderately high levels of adverse events and tend to be expensive, they would benefit from biomarkers that would allow them to be effectively targeted to likely responders.
Table I. (click to enlarge) Table of FDA-approved cancer drugs, target, and indication.
Knowledge of the types, functions, and interactions of the genes and gene products involved in cancer cell growth pathways has substantially increased the number of cancer drug targets. Some of the important cancer drugs, targets, and their associated biomarkers are shown in Table I.
Figure 1. (click to enlarge) EGF receptor signal transduction pathway. (Source: Sigma-Aldrich Co.)
One of the most important cancer cell growth pathways is the EGFR pathway (see Figure 1). This figure illustrates the complexity of one of the pathways involved in cancer and the multiplicity of routes from receptor to nucleus and hence growth promotion. While the EGF receptor is an important drug target, an anti-EGFR therapy would be effective only if this pathway were activated in a particular cancer and if there were no mutations in genes downstream of EGFR that would activate the pathway regardless of whether the EGF receptor was inhibited or not. It follows that it would be useful if there were biomarkers available to determine whether a particular pathway is regulated in a specific tumor prior to initiating therapy.
The Challenge of Clinical Utility
Table II. (click to enlarge) The eight requirements of a successful companion diagnostic.
Successfully developing and introducing a companion diagnostic alongside a cancer therapy requires the coordination of many elements (see Table II). The most important factor is demonstrating clinical utility, i.e., showing that using the assay will improve patient outcomes.
Generating clinical utility data requires a clinical trial to demonstrate the usefulness of an IVD. For a traditional diagnostic that identifies the presence of a disease state or predicts its outcome, the appropriate clinical trial population is a group of individuals with or without the disease in question.
However, clinical trials for companion diagnostics are different. Since such tests are designed to predict drug response, they must be applied to a population taking a therapy and demonstrating differential drug response. The type of trial that is required includes a pharmaceutical element with all of the attendant expenses.
In 2003, the average price of the type of phase III clinical drug trial that would be needed to show utility of a new companion diagnostic was approximately $119 million.1 The biggest challenge of personalized medicine is justifying the cost of demonstrating clinical utility in terms of the potential returns on an IVD and the information it provides.
Clinical Benefit versus Commercial Reality
Given the substantial costs of bringing a companion diagnostic to market, having a good clinical and scientific rationale is not sufficient to drive the development process. Without a sound business plan, the prospects for developing a companion diagnostic are poor.
Four main players interact and influence the decision making regarding cancer companion diagnostics: the pharmaceutical industry, the IVD industry, regulatory authorities, and healthcare providers. Three of these players (the pharmaceutical and IVD industries, and healthcare providers) have a direct commercial interest in combining drugs and cancer companion diagnostics that could justify the costs of implementing a personalized medicine approach.
The drug industry wants to increase the value of its therapies, the IVD industry is driven by the opportunity to boost the price and volume of its products, and healthcare providers strive to provide more cost-effective-healthcare. The successful development and introduction of a companion diagnostic requires that at least one of these three players should believe there is sufficient value in the proposition to justify the costs of development.
Pharmaceutical Industry Perspective
In the past, pharmaceutical companies have resisted using IVD tests to target their therapies. Such resistance was due to a view that the added complexity of requiring an IVD assay that would stratify the target market would result in lower product sales. At least in the field of cancer medicine, this attitude has shifted to a position in which all cancer drugs being developed have accompanying biomarker programs with the potential for companion diagnostics.
Such changes in attitude have been partly driven by the need for improved clinical trials and the realization that properly targeted therapies can achieve greater sales than the trial-and-error drug therapy approach. Other reasons for such changes include the following:
The drug industry wants to increase the value of its products, which is neither the same as nor as simple as increased drug sales. If a pharmaceutical company determines that a companion diagnostic is required to support one of its therapies, it is in an excellent position to facilitate the process. In particular, if a clinical trial designed to show a drug's effectiveness can also demonstrate an IVD's utility, one of the major costs of IVD development has been eliminated. Since drug companies are not interested in developing and promoting their own IVDs, forming partnerships with IVD companies is the normal route to bring a companion diagnostic to market.
IVD Industry Perspective
While the IVD industry is excited by the potential revenues from the sale of companion diagnostics, it is hampered by the reimbursement rules established in the major healthcare systems around the world. Such rules cap the prices of IVDs, so increased volume is often the principal route to increased value for IVDs.
The challenge for the IVD industry is that the potential value of a companion diagnostic is often not large enough to justify independent development. As a hypothetical example, imagine an assay that identifies good responders to a drug for metastatic breast cancer and can detect the 50% of the population that may respond to the drug and differentiate it from the other 50% in which a response is unlikely.
The market size is the incidence of the disease, which for this example is about 350,000 patients in Europe and the United States, and the value to the end-users is getting a better chance of response to therapy in patients who test positive. The clinical arguments for using the pharmacodiagnostic are compelling. However, if the assay is reimbursed at a typical level for a nucleic acid–based test ($200) and the IVD manufacturer estimated capturing 25% of the market, the total potential annual revenues would be $17.5 million. While this is a reasonably sized opportunity for an IVD company, it is still not sufficient to justify the cost of a clinical trial needed to demonstrate an assay's value.
The implications of this analysis are that IVD companies should limit their pharmacodiagnostic development plans in oncology to situations in which they can collaborate with drug companies. Alternatively, they should either find other ways to demonstrate clinical utility or obtain higher reimbursement levels that more accurately reflect the value of their products.
Healthcare Provider Perspective
From a pharmacoeconomic perspective, the use of a companion diagnostic to target a therapy is almost always good value; therefore the potential savings from more-cost-effective drug treatments could be used to justify the cost of companion diagnostic development
Since healthcare providers around the world are under pressure to provide more-cost-effective healthcare, they perhaps have the most to gain from personalized medicine. Even the most facile pharmacoeconomic analysis demonstrates that appropriate use of companion diagnostics increases the cost-effectiveness of healthcare. But healthcare organizations do not have the financial resources to invest in the studies needed to demonstrate clinical utility, so they depend on studies or data from the drug and IVD industries. However, such organizations can employ strategies to increase the likelihood of introducing personalized medicine.
In summary, while the drug industry has many reasons to invest in developing companion diagnostics, the opportunities for the IVD industry and healthcare providers are limited by the economic realities of current healthcare provisions (e.g., reimbursement). Accordingly, a companion diagnostic could be expected to emerge in the following situations:
This list is exemplified by the examples of cancer personalised medicines currently available in the market as illustrated below.
Pharmaceutical Industry Demand
The best-known personalized medicine example is Herceptin by Genentech (South San Francisco, CA), a breast cancer therapy used in both adjuvant and metastatic settings. The efficacy of Herceptin in an unselected population is low, and using an unselected population would have required a very large clinical trial to demonstrate its benefits. By preselecting patients with elevated HER2, a much smaller trial was used to demonstrate its efficacy.
During clinical development, Genentech developed the assay used to identify the HER2-positive patients. However, since the assay was deemed inadequate for commercialization, Genentech approached Dako Corp. (Glostrup, Denmark) to codevelop the HercepTest, a commercial immunohistochemistry kit. The kit was validated by demonstrating equivalence to the clinical trial assay. In September 1998, FDA approved both Herceptin and HercepTest.
In this example, the cost of the trial to demonstrate the clinical utility of the IVD alongside the drug was borne by the drug company, and the IVD partner was brought in later to bring the assay to market. In 2006, while sales of Herceptin totaled $1.2 billion, sales of HercepTest reached only tens of millions of dollars.
A similar story emerges with other examples of drug-company-driven clinical trials, with IVD companies undertaking the clinical trials and providing the IVD assays. One example is Dako's EGFR protein expression assay, which must be performed before patients are eligible for treatment with Cetuximab, a humanized EGFR antibody, by ImClone Systems Inc. (New York City). Another example is the K-RAS mutation assay by DxS Ltd. (Manchester, UK), which must be conducted prior to treatment with Panitumumab by Amgen Inc. (Thousand Oaks, CA).
While providing cancer companion diagnostics to serve the needs of the pharmaceutical industry is a good business opportunity for IVD companies, the value of the drug sales that are predicated on the availability of the IVD is not reflected in the value of the IVD itself.
IVD Industry Push
One of the benefits of molecular diagnostic tests is that they tend to be high-margin products. That is, the cost of manufacturing them is low compared with their price. Broadly speaking, a molecular diagnostic's value can be considered to be the product of its price and volume. Any given cancer is not a particularly high-incidence disease, and companion diagnostic tests tend to be used only once per patient. Consequently, the key to justifying the investment needed by an IVD manufacturer to independently bring a novel companion diagnostic to market is ensuring a high price for the product.
Justifying such an investment can only be achieved by creating significant demand for the assay and having strong intellectual property protection, which allows the vendor to maintain a high price. This still leaves the problem of the need to demonstrate clinical utility. But if that can be achieved creatively, then there is the possibility of bringing a high-value companion diagnostic to market.
One example is Genomic Health Inc. (Redwood City, CA) and its OncotypeDX test. This IVD assay quantifies the likelihood of breast cancer recurrence in women with newly diagnosed, stage I or II, node-negative, or estrogen receptor-positive breast cancer who will be treated with Tamoxifen.
Genomic Health met all three of the criteria for developing a high-value companion diagnostic product:
The OncotypeDX assay is available from Genomic Health for $3650, which is considerably more than HercepTest's price of around $300.
Regulatory agencies play an important role in introducing pharmacodiagnostics to the market. One recent example is the decision by EMEA to grant marketing authorization for Amgen's Panitumumab, but only in patients who do not carry a K-RAS gene mutation. (Panitumumab is a humanized anti-EGFR antibody.) In September 2006, FDA granted approval to Panitumumab for treating patients with EGFR-expressing, metastatic colorectal carcinoma with disease progression following chemotherapy regimens.
In contrast, when Amgen presented to EMEA in May 2007 the same data that were shown to FDA, the EMEA's Committee for Medical Products for Human Use adopted a negative opinion. The committee recommended refusing the marketing authorization on the grounds that the drug had only a very small effect in increasing the time until either the disease got worse or the patient died, in comparison with best supportive care.
The EGFR gene is overexpressed in colorectal cancer. Panitumumab binds to EGFR and inhibits the growth and survival of cells expressing EGFR. A diagram of the EGF receptor signal transduction pathway is shown in Figure 1, with the EGF receptor highlighted.
One route for the signal from the EGF receptor to be transmitted to the nucleus is through the K-RAS gene product. The K-RAS gene is mutated in 35% of colorectal cancers, and the effect of this mutation is to lock the K-RAS gene product permanently in the “on” position. If K-RAS is downstream of EGFR, inhibiting EGFR will have little effect on tumors in which growth is being driven by K-RAS gene mutations. From this it would be expected that K-RAS–positive tumors would not respond to EGFR inhibitors.
On the basis of this reasoning, Amgen scientists reanalyzed the clinical data to determine if K-RAS mutation status was related to drug response, and presented the data to EMEA. They found that only patients who lacked the K-RAS mutation showed a benefit. In September 2007, EMEA granted a conditional marketing authorization for Panitumumab in patients whose tumor contains a nonmutated K-RAS gene.
This conditional approval has led to the demand for an appropriate companion diagnostic. Having provided the reagents for the clinical trials, DxS also launched the IVD in January 2008 to coincide with the drug's launch.
Healthcare Provider Demand
Given the relatively high costs of cancer drugs and low costs of cancer pharmacodiagnostics, healthcare providers perhaps have the most to gain financially from increased use of pharmacodiagnostics. Despite this benefit, providers do not often drive initiatives to introduce new pharmacodiagnostic tests. Perhaps this is because such organizations do not have a culture of innovation and prefer to take a more reactive role to new technologies that are presented to them.
One recent development that could cause a major shake-up is a decision by the National Institute for Health and Clinical Excellence (NICE; London), a health technology assessment agency. NICE has published a draft guidance to ensure that patients with multiple myeloma are prescribed Velcade by Janssen-Cilag (Beerse, Belgium). If such patients show a full or partial response to Velcade, treatment continues to be funded by the National Health Service (NHS). However, those patients showing a minimum or partial response would be taken off the drug, and the costs would be refunded to the NHS by the drug's manufacturer.
This proposal is interesting because it puts the onus back on the pharmaceutical company to find more-effective ways of identifying likely responders and minimizing the amount to be paid back to NHS. The upshot is that the proposal will encourage pharmaceutical companies to develop tools that can identify likely responders, which will then be made available through the IVD industry.
Stephen Little, PhD, is the chief executive officer at DxS Ltd. (Manchester, UK). He can be reached at stephen.little@
This article has used examples from the cancer market to explore the factors driving the development and deployment of pharmacodiagnostic tests.
Of the five scenarios that were identified, three of them led to the same situation: IVD companies would continue to develop companion diagnostic products only at the request of pharmaceutical companies. While such companion diagnostics would greatly affect the value of drug sales, this would not be reflected in the price of the products themselves. Only the scenarios in which the IVD industry took the initiative resulted in companion diagnostics that managed to capture a reasonable proportion of the value they were creating in the healthcare system.
The challenge for the diagnostics industry is how to translate the value that companion diagnostic products create into a more fairly reimbursed test price. Without a significant overhaul of the reimbursement system, it seems the best routes available are to invest in the acquisition of intellectual property and clinical utility data to allow the development and sale of a high-value product or to establish equitable partnerships with pharmaceutical companies to share the value that the companion diagnostic creates.
1. J DiMasi, R Hansen, and H Grabowski, “The Price of Innovation: New Estimates of Drug Development Costs,” Journal of Health Economics 22 (2003): 151–185.