Holy Grail Of Precision Medicine – Co-Development Of Therapeutics And Companion Diagnostics

At Illumina’s Genomics Forum conference this fall, former President Obama spoke about how precision medicine has a “considerable role to play in making the health system more efficient and producing better outcomes,” but that it has not yet realized its full potential.

Indeed, precision medicine holds tremendous promise, but still has a long way to go, in particular with respect to efficiently connecting the dots between diagnostics and therapeutics during clinical development.

Clinical validity and utility

Precision medicine relies on the clinical validity of diagnostic tests, or the accurate identification of patients with certain biomarker alterations. Clinical validity is, however, a necessary and not a sufficient condition for successful implementation of personalized medicine. To achieve success, clinical utility is needed, which means that biomarker test results are actionable in terms of guiding decision-making – prescribing of specific medicines, for example – and this in turn leads to positive health outcomes.

But even in oncology, where precision medicine is the most advanced, levels of biomarker testing differ greatly across clinical practice settings and tumor types, according to a recently published article in Precision Oncology. In non-small cell lung cancer, for instance, it’s estimated more than 70% of patients have tumors with biomarker alterations that are in some way related to the efficacy and safety of different therapeutic options. However, many patients still do not receive biomarker testing. And when lung cancer patients do get tested and have actionable results, nearly 30% are not prescribed targeted therapies indicated as appropriate by the tests.

It’s unclear why this is. Challenges exist around testing itself, clinician buy-in, and also payer payment for tests. Perhaps the most important hurdle to overcome, however, is the frequent disconnection between clinical validity and utility.

Co-development of diagnostics and therapeutics

Companion diagnostic tests have become increasingly important for optimizing drug use, especially in oncology. This means tailoring the use of certain therapeutics to those who stand a better chance of benefiting or less of a chance of harmful side effects. For many cancers and an increasing number of other diseases, customizing treatment is based on diagnostics which identify specific biomarkers.

Impediments remain, however, with respect to the most efficient way of personalizing medicine; co-development of diagnostics and therapeutics. There are clinical and practical challenges related to co-development, but also economic issues, such as lack of comprehensive insurance coverage for many companion and complementary diagnostics. Indeed, this continues to be a major obstacle to the realization of personalized medicine more generally.

As knowledge around biomarkers grows, more drugs are coming onto the market with Food and Drug Administration (FDA) approved labels indicating their use with a particular diagnostic test. Nevertheless, most of these diagnostics are post-hoc, that is, the approved therapeutic comes first and then a matching new diagnostic is developed. Or, an already existing diagnostic is found to correspond to a therapeutic after its approval.

By contrast, drug and diagnostic co-development involves parallel identification, development and testing of a therapeutic and its corresponding diagnostic. Precision medicine’s holy grail is co-development of therapeutics and companion diagnostics, given that it facilitates the drug development process and tailors drugs – prior to marketing approval – to patients who stand to benefit most, or have fewer side effects. In turn, this simplifies post-approval reimbursement decision-making, usually favoring positive coverage decisions for both therapeutics and diagnostics.

Regulatory incentives for co-development have been in place for a while. In an FDA guidance published in August 2014, the agency stated: “Ideally, a therapeutic product and its corresponding in-vitro companion diagnostic device should be developed contemporaneously, with the clinical performance and clinical significance of the in-vitro companion diagnostic established using data from the clinical development program of the corresponding therapeutic product.”

Three examples of co-developed companion diagnostics are:

• IDH1 assay, developed by Abbott, in conjunction with ivosidenib (Tibsovo); a Servier-sponsored drug indicated for acute myeloid leukemia
• CDx BRCA assay, developed by Foundation Medicine, in conjunction with rucaparib (Rubraca); a Clovis-sponsored drug indicated for recurrent epithelial ovarian cancer
• ALK CDx assay, developed by Abbott, in conjunction with crizotinib (Xalkori); a Pfizer-sponsored drug indicated for non-small cell lung cancer

But, among the current list of FDA-approved companion diagnostics, most are not co-developed with therapeutics.

Co-development challenges

Twenty years ago, the mapping of the human genome promised new opportunities for genomic or genetic tests to identify diseases and guide treatment modalities, custom-designed for specific patient sub-populations. Completion of the genome project raised the possibility that co-development would become de rigeur. However, in reality co-development has faced numerous clinical, economic, and practical challenges.

Presently, there are thousands of genetic and genomic tests on the market for more than 10,000 conditions. In addition, hundreds of drugs have genetic markers on their FDA-approved labels, and at least one-third of drugs in the pipeline include genetic markers. These can be purely informative in that they inform of a (possible) connection between a particular genetic marker and an indication or disease. Alternatively, the presence of genetic markers on the label can inform prescribers and patients of potential adverse events in certain susceptible sub-populations, or, ideally – as is the case with co-developed drugs and diagnostics – directly tie in the use of a therapeutic with a biomarker test.

Note the difference, however, between identifying genes associated with a disease and developing a successful new drug that is contingent on the use of a companion diagnostic. Genetic discoveries have certainly provided clinicians with important insights into the pathology of disease. But the complexity of genetic expression involving multiple causal pathways – with numerous possible confounding factors – implies there often isn’t tight linkage between a specific set of genotypes and the ways in which disease progresses. In turn, this can make it more difficult to co-develop therapeutics and genomic-based diagnostics.

For example, checkpoint inhibitors – immunotherapies that target cytotoxic T lymphocyte antigen 4 or the programmed cell death 1 (PD1) pathway – have achieved success in the treatment of different cancer types. Yet only a comparatively small subset of patients derive clinical benefit. Immune checkpoint inhibitor efficacy is affected by a multitude of factors involving tumor genomics, host germline genetics, PD1 ligand 1 levels, but also environmental features, including the gut microbiome. Ultimately, a better understanding of precisely how all these variables interact would be required to optimize prescribing of checkpoint inhibitors.

Besides clinical issues, practical matters hamper co-development. There’s the timing aspect, for example. Ideally, a drug company would like to simultaneously launch a therapeutic and diagnostic. But this implies the diagnostic must be developed and ready to be deployed in clinical trials; preferably by Phase 2, and certainly prior to Phase 3.

Lack of early engagement between stakeholders in drug–diagnostic co-development models is yet another considerable challenge. At present, most manufacturers choose not to develop their own companion diagnostics. Instead, they partner with outside firms.

Even then, if partnering externally, the financial incentives to developing companion diagnostics are sometimes hampered by limited intellectual property protections. Diagnostics don’t enjoy the same kind of patent protections therapeutics do. Also, diagnostics generally face lower margins and higher reimbursement hurdles than pharmaceuticals.

In sum, developing a novel biomarker test in parallel with a therapeutic – with the expectation that the biomarker test is going to be included in the FDA-approved label – has been accompanied by a persistent set of challenges in the clinical and economic realms.