Scaling up cancer care

By Mark Lee (Global Head of Personalised Healthcare, Product Development) and Ron Park (Global Head of Personalised Healthcare, Global Product Strategy)

Personalised healthcare is hardly a new concept in oncology. It is rooted in a need to understand the differences between individual patients and translate those differences into more effective treatment options. The first medicines aimed at patients with particular biomarkers and genomic alterations were approved more than 20 years ago, laying the foundation for a new approach to cancer treatment. We are now in the midst of a data revolution, with the emergence of new technologies, analytical methods and sources of high-quality data that is placing us at another exciting inflection point in the evolution of personalised healthcare.

A year ago, we described our vision for how this next chapter of personalised healthcare would unfold. Today, our ability to collect and analyse data on a massive scale has matured to the point that it is becoming increasingly possible to deliver the right medicine to the right patient at the right time.

This new chapter in personalised healthcare is being driven by what we call MDAS, or meaningful data at scale. MDAS represents large amounts of high-quality information – more data, across more patients, than ever before – that allow us to visualise differences between patients with unprecedented resolution. Foremost is the “meaningful” part: MDAS includes patient data carefully selected to inform us about actionable differences across the spectrum of diseases. This data is obtained from a rich mix of sources, from diverse clinical outcomes to genomics, pathology, imaging and health history. Of course, all uses of data must be consistent with applicable healthcare and privacy laws to ensure that scientific insights are revealed without compromising patient privacy.

This brings us to the “at scale” part: to make sense of these ultra-high-resolution portraits of patients, we need data across large populations of individuals. In recent years, sources of MDAS and especially real-world data have exploded in both quantity and quality. At the same time, new tools have been developed to produce insights from that information. MDAS has reached the point where it can be used to help patients across the cancer care continuum and beyond, from generating new discoveries in the lab, to improving clinical trial design and treatment at the point of care, to helping policymakers at a systems level.

Here are some ways we see that happening.

In the lab, MDAS reveals differences between patients that would have been difficult to discern before. For example, checkpoint inhibitor immunotherapies have had success against certain types of cancer, but they don’t work for everyone. Researchers are working hard to understand why only about 20 to 30% of people benefit from this class of medicine. Understanding how to convert these “non-responders” into “responders” could improve the lives of thousands of patients, and this is exactly the type of problem MDAS can help solve.

In a clinical trial of one of our cancer immunotherapies, the use of deep molecular profiling data combined with clinical data revealed that some non-responders had an up-regulated gene signature associated with a protein called TGF-beta. This inspired researchers to combine the cancer immunotherapy with an investigational antibody that blocks TGF-beta in preclinical models designed to mimic the biology of the non-responders, which resulted in improved anti-tumor activity. Similarly, we have been examining the relationship of tumor genomic profiles with treatment response and outcomes in the real-world. Through our collaborations with Flatiron Health and Foundation Medicine, a real-world, de-identified clinico-genomic database is enabling us to probe this question at a new scale.

These examples highlight how real-world data from the clinic can accelerate basic discovery in the lab, and potentially lead to the improvement of existing medicines, the discovery of new combinations or the development of new treatments.

Personalising cancer care further depends on high-quality diagnostics that match patients to precisely targeted therapies. But ensuring patients actually receive the best treatment for them requires infrastructure to facilitate a match. In the last few years, multiple tools have been developed that help oncology care teams find the best medicine for each individual patient.

For example, we have launched the navify Tumor Board software solution to help healthcare providers manage and interpret all the diverse information available for a patient, from medical history to biomarkers, tumor information, radiology images, pathology reports and treatment notes. The software helps standardise the process, allows tumor board members to explore more options for each patient, and facilitates collaboration with experts in remote locations. It also informs the physicians of clinical trials that a patient may qualify for, highlighting options in cases where no currently approved medicines are available.

We are continuing to explore additional partnerships focused on other tools that may help improve care along the patient care continuum.

We are accustomed to talking about cancer according to where it arises in the body: breast cancer, lung cancer, colon cancer. But a small percentage of cancers have no known point of origin. Known as cancers of unknown primary, or CUPs, they appear as metastatic growths throughout the body, frustrating physicians with their diffuseness and no indication of how they should be treated. Despite treatment with a broadly-acting chemotherapy regimen, however, most CUP patients live for less than a year from diagnosis.

There is another way to think about CUP, however. Because cancer is driven by DNA mutations that cause cells to grow out of control, each case can be characterised according to the particular genetic signature that drives it. And that means that even for CUP, there is a possibility that a molecularly-targeted treatment is available – we just need MDAS to find it.

This is why we work with Foundation Medicine, a molecular diagnostics company that specialises in comprehensively profiling tumors for their unique genomic makeup. In a multi-site international study that also uses the navify software to select patients, called the CUPISCO trial, we are testing the safety and efficacy of molecular-based therapy for CUP. The hope is that this new way of thinking about CUP will build our clinical and genomic knowledge base in this patient population, and reveal possibilities for physicians and patients to have access to personalised treatment options that they may have never even considered.

MDAS offers tremendous opportunity to improve how we conduct clinical trials by not only making them more accessible but also faster, more efficient and more considerate of the patients who enroll in them. Whereas traditional clinical trials contain both an experimental and control group, having MDAS may enable us to put everyone participating in a trial on an experimental medicine that has the potential to improve upon an existing standard of care.

Through our work with Flatiron Health – which specialises in extracting high-quality data from electronic health records to curate de-identified, research-grade, real-world datasets – we can understand how to compare patients in clinical trials to similar patients receiving standard of care treatment in the real world. This could allow us to supplement or replace the control group in trials with real-world data. The result may allow for faster recruitment, lower costs and treatment of more patients in clinical trials with a potentially more effective treatment than the standard of care.

We recently pursued this approach when applying for regulatory approval of a new cancer medicine targeted to a rare mutation. As is common in rare cancers, it would have taken years to enroll enough patients for both treatment and control groups. Instead, we worked with Flatiron Health to curate a cohort of de-identified patient data to form the control group. Using Flatiron data, we were able to reduce the time required to perform the trial, increase the number of patients in the control arm and potentially accelerate the medicine’s path to approval.

The use of real-world data in regulatory approvals has obvious benefits for the developers of medicines and the patients who are waiting for better treatments. But it also has broad application for regulators in their mission to ensure the safety and effectiveness of new and existing treatments. Two recent pieces of legislation, the Food and Drug Administration Reauthorisation Act and the 21st Century Cures Act, set out milestones for the FDA to explore the use of real-world data in approvals and in expanding indications for medicines already on the market. MDAS in the form of real-world data can also be of enormous value to regulators in conducting post-market surveillance to answer questions about safety in particular patient populations.

We are supporting the FDA’s efforts through our active engagements with a diverse group of external stakeholders. For example, we are core members of the Duke-Margolis Center for Health Policy’s Real-World Evidence Collaborative, which is exploring ways to assess real-world data and integrate it into the regulatory process in the US.

We live in a world that is being transformed by data. Its impact is beginning to be felt in cancer care at every scale, from the most basic molecular biology to the point of care to the level of national regulatory policy. Although we often think of personalised healthcare as something that will come over time, we believe that we are poised for an immediate transformation, with major benefits to come for the many patients who are waiting.

In oncology, we have the ability to access comprehensive genomic profiling, and are developing the rich clinical databases and the infrastructure to access and analyse them quickly. To ensure that patients receive the benefits of this new paradigm,we are committed to partnering across the entire cancer care ecosystem to deliver personalised healthcare, fueled by data, analytics and technology. With these capabilities and collaborations in place, we believe we have the opportunity to improve the lives of every person diagnosed with cancer.