Clinical Researcher—December 2020 (Volume 34, Issue 10)

GOOD MANAGEMENT PRACTICE

David Messina, PhD

 

Scientific and medical research is an investment in our future. We’re all keen to live long, healthy lives and most of us are willing to empower physicians to make educated decisions about our care if and when we do become ill. Further, as we’ve seen from hundreds of years of scientific advancement, medicine is not a short-term investment. When a new drug or medical device comes to market, it builds on decades of basic science research, translational research and development, and clinical studies.

In the early 2000s, economists evaluated federal investments in medical and health research and found “that the returns from the national investment in medical research—both in the past and what is likely to be delivered in the future—are exceptional and far greater than is appreciated by either policy makers or the public.” That said, public and private investors alike know there is no fast-track to success; no magic pathway that results in breakthroughs faster or more often.

Scientific and medical research is rigorous, methodical, and fraught with failure. Often, however, these failures come with great learning opportunities, which help yield future successes.

Learning from Clinical Studies

One such learning opportunity is centered on the clinical studies that help move therapeutics and diagnostics out of the research lab and into clinical practice. Drug trials have exploded in the last decade, particularly in immune oncology, with more than 1,000 trials initiated in the U.S. alone last year, according to Kantar Health. This has spurred significant innovation in the therapies available for patients.

However, as we enter an era of more precise medicine, many of the therapies being developed work only in a subset of the patient population. This means there is a need for the development of a diagnostic approach for selecting patient populations who will respond well to the therapy.

At first, the industry prioritized this at the same time as the drug, developing the companion diagnostic in parallel and resulting in an on-label test that is required for safe use of the drug. An example of this is the use of a PD-L1 immunohistochemistry (IHC) assay to select patients for anti-PD-1 immunotherapies. While the intent behind this approach is sound, I would argue that it has unintentionally laid the foundation for a lack of innovation in diagnostic development.

For example, in some indications where PD-L1 IHC is currently the approved companion diagnostic device, such as recurrent and metastatic squamous cell carcinoma of the head and neck, clinicians have little to no confidence in its ability to predict response. So, rather than guiding treatment decisions, it’s seen as necessary but unproductive.

Despite broad awareness of this situation, the industry has overwhelmingly focused on expanding the indications approved for treatment with anti-PD-1 therapies, and has not given sufficient consideration to improving our ability to predict response and replacing the on-label diagnostic. So it is that, as with many other facets of medical research, our failures have taught us that there is room to improve. The future of diagnostics requires a new trial approach—particularly in oncology.

Driven by Innovation

While IHC has provided decades of invaluable information, it is far from being the most sensitive or specific methodology we have at our disposal. Further, the reproducibility of the technology poses major challenges, as was evaluated in the Blueprint Study. So, when we design clinical studies—either for companion diagnostics or independent diagnostics—I would argue we should be driven by the most innovative, informative tools we have, not just what has worked well in the past and represents a “safe bet.”

There are many other technologies available that have demonstrated they can generate valuable biological insights for oncology biomarkers. By moving these from the research space and into our suite of options for diagnostic development, we are expanding our arsenal in our fight against cancer. Further, when we look to more advanced technologies, we have the benefit of moving from a single-analyte snapshot to a more multidimensional, multi-analyte approach, which provides a more holistic view of the patient’s disease.

Examples of other technologies that should be considered include other imaging technologies with improved sensitivity and specificity, namely, immunofluorescence and mass cytometry. If spatial information is not biologically relevant to the biomarker to be measured, then technologies using next-generation sequencing (NGS) or polymerase chain reaction (PCR) are an excellent option. This umbrella is vast and covers DNA measurements such as tumor mutational burden (TMB) and microsatellite instability (MSI), as well as RNA measurements including gene expression, immune profiling, and predictive immune modeling.

Each of these have plentiful translational research that demonstrates their potential use as oncology biomarkers, and should be considered for diagnostic development in both the companion and predictive diagnostic setting. Examples include TMB across multiple solid tumor types,  co-testing MSI by PCR and dMMR by IHC, and the relationships between MSI, TMB, and PD-1/PD-L1 expression.

With the goal of precision medicine comes the need for multidimensional technologies. In the immunotherapy example, measuring holistic immune response at the site of the solid tumor is paramount to improving our ability to predict tumor response and has been shown to perform better than IHC alone. The future of diagnostic trials must leverage innovative technologies that can provide better insight into the complex biology of each patient.

Decentralized, Yet Harmonized

Diagnostic clinical studies are challenged by the same barriers to success that have been described for all clinical trials, including recruiting sufficient diversity in cohorts to represent the general patient population, ensuring compliance of patients, and confirming sites are following protocols identically. It’s certainly clear that decentralized trials are most effective for recruiting a more diverse patient population, but you might be concerned that this approach increases the latter challenge of site management.

In fact, each of the barriers listed can be reduced when a sponsor partners with a contract research organization (CRO) that is well-versed in running highly virtual trials. These CROs are equipped with platforms to provide remote monitoring, electronic patient consenting, and electronic data capture. This model even enables individual investigators, who may be passionate about science and improving patient outcomes but are not located in academic institutions with dedicated clinical trial staff, to enroll patients for participation in a clinical trial.

A virtual trial platform also allows for sponsors to leverage direct-to-patient engagement. By extending trial sites beyond large academic centers to sites local to patients, we enable maximal diversity in recruitment, streamline participant engagement, and help ensure improvements to patient compliance. This approach to diagnostic trials, especially those that are non-interventional and may not require additional hospital visits, would allow patients to participate in a clinical trial no matter where they live, what their socioeconomic status is, and who their treating physician is.

The future of diagnostic trials must be decentralized and supported by CRO partners who can keep sites and protocols harmonized.

Independent and Equally Impactful

Building a diagnostic that improves patients’ care paths and clinical outcomes should not be considered an afterthought, or only considered when bringing a new drug to market. Lessons learned during the development of predictive diagnostics for new therapies demonstrate that this approach has value for all therapies available. By developing predictive diagnostics for decision points along multiple care paths rather than for only one therapy, we move closer to the precision medicine paradigm where these technologies will empower physicians to understand the potential outcomes for the myriad of therapeutic options available to them.

Robust clinical studies for biomarkers should be prioritized not only during drug development, but also for therapies already on the market. The diagnostics being evaluated for predicting tumor response to anti-PD-1 therapy represent great progress post-therapy approval. However, we should not stop with building diagnostics for immunotherapies alone. Helping physicians make decisions about chemotherapy, radiation therapy, and combination therapies using diagnostic tools will not only improve patient outcomes, but will provide financial advantages for payers, patients, and the entire healthcare ecosystem.

Conclusion

As members of the clinical research community, we’re all aware that the stakes of our investment in medical and healthcare research have never been higher. Scrutiny in how funds are spent, the rising cost of healthcare, and our aging population require us to find new ways of rising to meet these challenges.

A fixation on novel therapies alone will not allow us to meet our goals of matching patients with the most impactful treatment regimen. Expanding our efforts to focus on diagnostic innovation, and even on how and when we conduct the clinical studies that bring diagnostic technologies to market, is essential to delivering successful outcomes and closing the precision medicine gap.

David Messina, PhD, has spent the last 20 years in computational biology and human genetics. He contributed to the Human Genome Project at Washington University in Saint Louis, mapped disease genes at the University of Chicago, and co-developed the first comprehensive atlas of human transcription factor genes. As COO of Cofactor Genomics, he is the lead on all regulatory and reimbursement efforts for the company, driving the implementation of RNA-based diagnostics and their clinical application.