So much of cancer care is about time. When a clinician diagnoses a patient with cancer – particularly advanced cancer — there’s no time to waste. Treatment decisions that give patients the best chance at extending and improving life must be made quickly. New, novel treatments have the potential to enhance patients’ lives, but without the information needed to match patients with those treatments, time can be lost. Clinicians need as much information about a patient’s cancer as soon as possible, yet too many are forced to make difficult treatment decisions without critical insights about the molecular drivers of their patient’s cancer.

Every person with a cancer diagnosis deserves access to genomic insights about the unique profile of their disease so that they – and their physician – can be armed for the essential first step in developing informed treatment strategies. However, approximately two out of three people with advanced cancer do not get the broad genomic insights about their cancer that can help them find the best treatment for their specific cancer. [1] And even for those that do have their genomes analyzed, not everyone will get insights that point to a treatment plan with options tailored to their disease. More options are needed. That’s where real-world data can play a role.

Real-world data allows us to identify patterns in clinical practice, specifically and importantly outside of the controlled environments of clinical trials. These data and patterns help fuel the cycle of research and analysis that produces insights to help improve clinical decision-making and, ultimately, lead to new targeted therapies and more options that benefit patients. At Foundation Medicine, where I serve as the senior director of data and insights delivery, we’re working with real-world data every day to help improve drug development research and clinical treatment decisions. How do we do this? We gather and harmonize genomic data collected through comprehensive genomic profiling (CGP) done in the clinic as part of standard-of-care practices. CGP provides physicians with valuable insights today about their patient’s specific cancer. Brought together, these data can then be used in research to generate insights that can lead to better cancer care for tomorrow.

Genomic data alone can be a powerful tool but connecting it to clinical outcomes unlocks many other opportunities for insights. In partnership with Flatiron Health, we developed a world-class database of patient-level genomic and clinical data called the Clinico-Genomic Database (CGDB), which links our de-identified CGP data with Flatiron Health’s de-identified clinical data captured by electronic health records. This real-world database provides researchers with diverse data from more than 97,000 patients across all tumor types. The number of patients in this dataset grew by approximately 40% in 2021. Importantly, the CGDB is representative of a broader patient population receiving cancer care in the U.S. — opening endless possibilities for research that drives advancements in precision oncology.

The Two-Way Street Between Real-World Data and CGP
Research shows that people with advanced cancer may experience better outcomes when genomic testing is used to match them to a therapeutic approach. [2][3][4][5] Not only does it help clinicians and patients make treatment decisions by matching patients with targeted therapies that may be more effective, but CGP data also powers research to further our understanding of cancer.

Advertisement #3

It allows researchers to see how patients are responding to treatments based on their specific genomic signature, which helps inform the development of new targeted therapies and identifies treatments that should or shouldn’t be considered for a patient based on the data. By combining CGP and clinical data, the CGDB aids researchers in conducting retrospective studies that zoom in on patient populations that would be difficult to do in – or are more representative than – traditional clinical trials.

Real-world data like aggregate CGP results and CGDB present a unique opportunity for us to look for patterns or signals in previously unappreciated driver mutations. Early in our company’s journey, we found that variants affecting splicing in the MET gene were occurring much more frequently in non-small cell lung cancer (NSCLC) than any other tumor type. NSCLC accounts for approximately 85% of lung cancer diagnoses, and 3 to 4% of which are associated with MET exon 14 splicing (METex14) [6][7]

Later, through analysis of CGP data across thousands of patients, we found more than 200 alterations that affect METex14, almost all in lung cancer patients. This discovery was by far the largest characterization of METex14 frequency, diversity, and co-occurring genomic alterations. And, as the development of METex14-targeting therapies evolved, CGDB played a role in characterizing the real-world treatment patterns of METex14 patients. The first step in translating this knowledge into improved clinical outcomes for patients is identifying these METex14 alterations through CGP, which allows clinicians to match patients with the best available treatment – including newer targeted therapy – a critical step in the treatment decision-making process.

By analyzing real-world data, we’re able to identify patterns emerging in clinical practice. We can see which patients are receiving CGP, when they receive it in their treatment course and how this impacts treatment decisions and, ultimately, patient outcomes. These testing patterns also allow us to spotlight barriers patients are facing in accessing precision medicine.

For example, in partnership with our collaborators at Sylvester Comprehensive Cancer Center at the University of Miami Miller School of Medicine, the University of Michigan, and Harvard Medical School, we presented a study at the 2021 American Society of Clinical Oncology Virtual Scientific Program (ASCO21) analyzing the genomic landscape, CGP utilization, and treatment patterns among more than 11,000 men with advanced prostate cancer – what is believed to be the largest known cohort of its kind. Using data from the CGDB, researchers found that despite similar rates of actionable gene alterations between men of European and African ancestry, men of African ancestry were less likely to receive CGP early in their treatment course. These findings are an uncomfortable reality that we have to face if we’re going to improve disparities in access to testing, and, ultimately, disparities in outcomes.

This is why I’m passionate about working with real-world data. Studies like this are a critical step in identifying barriers patients face in accessing quality cancer care so that we can deliver on our mission of transforming cancer care by helping people facing a cancer diagnosis get access to the best available tests and treatments.

What’s Next for Real-World Data? Growth.
Real-world data provides the opportunity to help provide better care for more patients. If we understand the implications of treatment, we can develop more options for patients. Further, if we identify barriers to access, we can inform and open those channels so that more patients have the best available options when facing a cancer diagnosis.

Again, cancer is about time. Patients don’t have time to run into roadblocks to access valuable information that could inform their care. They don’t have time to take a trial-and-error method to their treatment plan. They don’t have time for their physicians to lack the resources to help them find the best available treatment for their specific cancer. That’s why it’s imperative that the entire cancer community fuel the cycle of real-world data that will truly make an impact in clinical practice – and patients’ lives. This means we need to continue investing in tools that identify valuable real-world data and, most importantly, raise awareness that these tools exist.

Researchers need quality real-world data to advance the way we think about and treat cancer. Information is key – and capturing this diverse patient-level data will continue to lead to novel therapies, validation of new targeted treatments, and a better understanding of which treatments should or shouldn’t be considered for a specific patient. Real-world data contains crucial information that can enable us to refine our thinking and move closer to understanding the best available treatment for each patient.

[1] L.E.K. interviews and analysis FMI-purchased Kantar data, SEER, MedPanel ‘Assessment of Market for Pan-Cancer Testing Wave 2’, IMS Health Medical Claims Data (DX) dataset (April 2015 to Dec 2016).
[2] Schwaederle M, et al. J Clin Oncol. 2015;33(32):3817-3825. Accessed Oct. 7, 2021.
[3] Wheler JJ, et al. Cancer Res. 2016;76(13):3690-3701. Accessed Oct. 7, 2021.
[4] Hainsworth JD, et al. J Clin Oncol. 2018;36(6):536-542. Accessed Oct. 7, 2021.
[5] Schwaederle M, et al. JAMA Oncol. 2016;2(11):1452-1459. Accessed Oct. 7, 2021.
[6] American Cancer Society. Key Statistics for Lung Cancer. Available at
[7] Data on file. Novartis Calculation. Kantar Health. CancerMPact: lung (non-small cell) stage IV incidence and newly recurrent. Updated December 15, 2018.

Featured Image: Doctor consulting with a patient, working on diagnostic examination.Photo courtesy: © 2016 – 2021 Fotolia/Adobe. Used with permission

Advertisement #5