In a follow-up to the recent pharma partnering article, ‘Beauty and the Beast – CDx Partnership Disparity’, Bob Holland provides the pharma viewpoint.
It’s a very seductive myth that the co-creation of a novel medicine with a companion diagnostic will provide a good outcome for both the pharma company and the diagnostics company. Surely both companies need each other? Surely their respective products will provide a guaranteed market for the other party’s product?
But is this really so? Let’s consider the drivers of the pharma companies:
Drugs companies have to get their products through regulatory approval (and pricing and reimbursement processes). Here, having a companion diagnostic going through relevant development and regulatory processes at the same time is clearly in the interest of the pharma company. The diagnostic facilitates effective clinical trials, can be required for the drug approval and is necessary for testing to be available for the drug to launch. So in the late development and peri-launch phases, the pharma company clearly needs the diagnostics company to succeed.
However, once a drug is licensed and on the market, pharma companies want their drug to be used by as many patients (who would gain benefit) as possible. Here the interests of the pharma company and the diagnostics company begin to diverge. Marketing success for the drug relies on the ready availability of inexpensive and easily delivered testing. It doesn’t matter to the pharma company which test is used, provided that the test is reliable and the results are accurate. In fact, what really matters is that testing should be simple to organize and that testing interferes minimally in the established patient care pathway. And here, there are a couple of key points to keep in mind: that the patient care pathway can vary widely between territories and healthcare systems; and that a new test has to be integrated, one way or another, into the testing framework that already exists for a given diagnosis and a given kind of patient. So, requiring healthcare providers to utilise precisely and only the companion diagnostic which was co-developed with a novel drug might be detrimental to the uptake of even the best of new medicines.
The point about integrating a new test into a current testing framework will soon become very problematic in the cancer field where multiple tests and multiple drugs either already exist or are in development for many important tumor types. One current example is non-small cell lung cancer (NSCLC) where two targeted therapies are already on the market with different tests (EGFR inhibitors requiring tumor genotyping for activating mutations; and crizotinib – approved with a FISH test to identify tumors with ALK translocations). In this same cancer there are drugs in clinical development requiring immunohistochemistry, others requiring additional genotyping in the EGFR receptor and yet others requiring genotyping of other genes. In just a few years, patients with NSCLC might require half a dozen different tests to identify which targeted therapy they should receive. It is inconceivable that each new drug to be approved will mandate an extra test to be performed—the logistics of this multi-modality testing will become too complex and the costs prohibitive.
So what is likely to happen? The single drug-single companion diagnostic model is unlikely to survive for much longer. The pharma companies don’t need it and patients and healthcare providers don’t want it. Already we see multiple forms of tests being used in the clinic to enable the prescription of specific, targeted medicines. We also see today, mainly in the cancer field and driven by academic medical institutions, the development of panel tests, i.e., tests which identify multiple tumor genetic abnormalities in parallel. It is quite conceivable, in territories where laws and regulations allow, that pharma companies will collaborate with multiple test providers to ensure maximum availability of testing, and with each other and with the academic world, to ensure that the panel tests provide coverage of medicines on the market and medicines in development.
If I were leading a diagnostics company, I’d think hard about using my company’s R&D resources to create a single drug companion diagnostic and I’d probably be looking in detail at the new panel tests and ensuring that my company was going to take advantage of this next evolution of the testing world.
After 27 years working in international pharmaceutical companies across all phases of pre-clinical and clinical research and development, Bob Holland has recently retired from his role as global head of personalized healthcare and biomarkers at AstraZeneca.