Analyzing real-world clinical data may transform cancer treatment

By David Delaney, MD, and Kevin Fitzpatrick

The world’s leading cancer centers have made significant progress in collecting and storing massive amounts of valuable healthcare data. Moreover, during the past few years the volume of published cancer treatment data has increased exponentially, filling data warehouses.

David Delaney, MD, SAP.

David Delaney, MD, SAP.

As a result, data analysts, oncologists, researchers, and scientists are now examining not just digestible bytes of information, but rather petabytes of data. By doing so, they hope to identify algorithms and meaningful relationships that can help clinicians to gain insights and provide more effective—and personalized—cancer therapies and treatments.

In tandem, bioscience researchers have developed new methods to cost-effectively collect and analyze mass volumes of patient biomedical information, including DNA sequencing data. With today’s technologies, sequencing a full human genome costs about $1,500 and takes only about 1 day. In 27 hours, researchers can get the entire genetic blueprint that makes a person unique. Human genome research can substantially contribute to our knowledge of human health and disease, and, increasingly, it can help oncologists better understand the genesis and even optimal treatment for certain cancers.

However, much of this vital cancer data—such as patients’ individual characteristics, genome data, treatment regimens, and even outcomes—is locked away in information silos at universities, oncology research centers, community hospitals, and oncology clinics. As a result, physicians are not able to leverage this latent institutional knowledge to drive better decisions. So, treatment decisions for real-world patients, many of whom would have been excluded from clinical trials, often rely on an uneasy blend of evidence-based medicine extrapolated to account for specific patient factors, combined with the lessons from cases that practitioners identify as similar based on their intuition and experience.

Kevin Fitzpatrick, CancerLinQ.

Kevin Fitzpatrick, CancerLinQ.

Improving on this paradigm is the goal of a new set of technology solutions powered by rich datasets and high-performance, in-memory computing capabilities. The new systems have the potential to facilitate cancer care transformation by connecting physicians to vast amounts of health data. By building platforms and networks that enable oncologists and researchers to harness and leverage the collective experience of entire organizations, systems, regions, and countries, physicians can make treatment choices based on research from a broad group of cancer patients more similarly matched than ever before.

In short, such new systems can provide physicians and the care team with a great number of evidence-based decision-support tools to enhance a physician’s decisionmaking power. With a focus on personalized, data-based decisions, this new approach is remaking healthcare—and more specifically, changing how oncologists treat cancer.


Currently, information about individual cancer patients is stored in siloed databases within each practice or treatment center across the nation. While oncologists can leverage clinical trial data and use it in decisionmaking for an individual patient during his or her treatment, too often the evidence being applied is limited to the experience and judgment of the individual clinician. Outside the small minority of patients enrolled in clinical trials, the ability of clinicians to leverage data about previously treated patients in aggregate, or to continuously learn from and improve real-world results, is severely limited.

One unique platform aiming to aggregate cancer treatment results from oncologists is CancerLinQ. An initiative launched in 2015 by the nonprofit American Society of Clinical Oncologists (ASCO), CancerLinQ is a health information technology (HIT) platform that is aggregating and analyzing a growing amount of patient data in order to uncover insights and trends, and to measure physicians’ care against that of their peers and recommended guidelines. (For more information, see “CancerLinQ Puts Patient Data in Play.”)

Figure 1. Everyday patients tend to be older, less healthy, and more diverse than clinical trial patients. Graphic courtesy CancerLinQ. Click to expand.

Figure 1. Everyday patients tend to be older, less healthy, and more diverse than clinical trial patients. Graphic courtesy CancerLinQ. Click to expand.

The data that oncologists typically leverage for determining treatments are based on a tiny subset—only 3%—of clinical trial patients. This proportion is startlingly low, considering that more than 1.7 million people in the United States are diagnosed with cancer each year.1 This might be okay if the enrolled patients represented the general population of cancer patients. However, real-world cancer patients tend to be older, sicker, and more ethnically diverse than typical clinical study patients, who often tend to be in good health—except for their cancer (see Figure 1).2–4

With CancerLinQ, every patient and their specific type of cancer is accounted for in the data warehouse. CancerLinQ will unlock knowledge and value from the 97% of cancer patients not involved in clinical trials, to help clinicians deliver better, more data-driven decisionmaking based on real-world results from patients closely matched to the patient at hand.


CancerLinQ provides the key to unlocking information in data silos by deidentifying and aggregating patient data gathered from across the country, thereby uncovering patterns and insights that will improve patient care. On the individual patient level, doctors can enter in patient characteristics—including age, symptoms, and genomic type—and receive feedback about how patients with similar characteristics are being treated across the country (see Figure 2).

As of August 2016, CancerLinQ has more than 60 vanguard practices using the platform, with nearly 1 million cancer patient records stored in the system. Participants range from small private practices to some of the nation’s leading cancer centers.

Figure 2. Workflow of the CancerLinQ system, which permits doctors to enter in their patient’s characteristics and receive feedback about how patients with similar characteristics are being treated across the country. Graphic courtesy CancerLinQ. Click to expand.

Figure 2. Workflow of the CancerLinQ system, which permits doctors to enter in their patient’s characteristics and receive feedback about how patients with similar characteristics are being treated across the country. Graphic courtesy CancerLinQ. Click to expand.

While it is still early days, CancerLinQ has the potential to develop one of the most accurate depictions of clinical practice patterns. Oncologists using CancerLinQ now have growing amounts of usable, searchable, real-world cancer information to help them provide better quality assessments, care coordination, case management, and other healthcare activities.

CancerLinQ amasses structured and unstructured information coming from disparate electronic health record (EHR) systems. The format of each data element varies based on the source system, ranging from unstructured text to highly structured elements. The persistent challenge for the platform is to normalize the deluge of data used in cancer care.

An eventual goal of CancerLinQ is to work with the oncology community and EHR vendors to standardize the way data are described, making systems more interoperable and better able to exchange information.


Using data analytics to improve the quality of patient care is a relatively new concept for physicians, so we are constantly gathering user feedback about CancerLinQ. “On the individual patient level, providers currently make recommendations for interventions based on a small number of individuals participating in clinical trials,” says Robin Zon, MD, FACP, FASCO, a medical oncologist at Michiana Hematology PC, one of the practices that have signed agreements to participate in CancerLinQ. “However, our patients very often differ from research participants, and we have no other reliable data on which to base our recommendations.

Robin Zon, MD, FACP, FASCO, Michiana Hematology.

Robin Zon, MD, FACP, FASCO, Michiana Hematology.

“CancerLinQ allows us to learn from patients beyond clinical trials, and discover whether we are treating our patients in the best manner possible,” adds Zon. “We will now be able to answer a question so often asked by the patient: ‘How did patients like me do on this treatment?’ For my patients’ sake, I look forward to using CancerLinQ to learn what I have been doing correctly, and to learn what needs to be changed or further explored, so that I can enhance the quality of my patients’ care and experience during their cancer journey.”


The CancerLinQ platform provides several unique features, including real-time monitoring and reporting of clinical quality measures. In addition, CancerLinQ allows clinicians to make better use of EHRs through advanced data visualization capabilities. CancerLinQ runs on SAP Connected Health and was built on the SAP HANA platform, a multipurpose data management and application platform. The CancerLinQ system has four chief capabilities:

  • Continual performance tracking for comparison with clinical quality measures.
  • Trend evaluation based on deidentified patient data.
  • Patient cohort identification based on shared characteristics.
  • Individual patient timeline construction based on treatments, side effects, and outcomes.

CancerLinQ assigns an implementation team to work with the participating practice and to develop and execute an implementation plan, timeline, and technical approach for uploading the practice’s patient data.


CancerLinQ is among the first of many initiatives that that are aggregating real-world oncology datasets to improve patient care and advance knowledge. The potential for precision medicine to improve outcomes while reducing cost is too great for healthcare groups to ignore. The government and corporate entities funding healthcare and the awakening giant of patient consumerism are escalating demands to accelerate this transformation.

Last year, the White House committed the nation to a $215 million investment in precision medicine. And in his 2016 presidential address, President Obama announced the National Cancer Moonshot, a $1 billion government initiative created to accelerate cancer research efforts, promote data sharing, and facilitate collaborations to advance cancer prevention, treatment, and care. The National Cancer Moonshot initiative is led by Vice President Joe Biden, who has catapulted the project forward this summer by hosting a national summit as well as many regional summits.

Clifford A. Hudis, MD, FACP, ASCO.

Clifford A. Hudis, MD, FACP, ASCO.

CancerLinQ is working toward solving one of the major challenges to the Cancer Moonshot outlined by Vice President Biden: the need to break down the information silos that contain critical science, data, and cancer research results, and to deliver game-changing treatments faster to those who need them. “One of the goals of the Cancer Moonshot is to share existing knowledge and new information more freely,” says Clifford A. Hudis, MD, FACP, CEO of ASCO. “CancerLinQ directly addresses this challenge by helping our community learn from everyday practice and share real-world data to speed the development and dissemination of the latest information and best treatment approaches.”

For this dissemination to happen, technology companies, oncologists, cancer research centers, and policymakers have to join forces. Earlier this year, ASCO released a comprehensive report, The State of Cancer Care in America: 2016, in which researchers outlined several strategies to improve cancer care delivery and accelerate progress. 5 One of the key elements of the path forward will be to advance health information technology that supports efficient, coordinated care. The ASCO report notes that the adoption of EHRs has led to the rise of big data analytics platforms, including CancerLinQ. ASCO is now calling on policymakers to take steps to make data sharing fast, efficient, and secure, so that these new initiatives can achieve their potential for patients.


While the healthcare industry is poised to embrace powerful data insights and improve cancer care outcomes, the technology has to keep pace. The volume of existing data on both cancer research and specific patients is immense—but often unstructured. With initiatives such as CancerLinQ and momentum from policymakers behind the National Cancer Moonshot, we can address the urgent need to break down barriers that impede progress in our fight against cancer.

By analyzing patient medical records and using in-memory computing technology to accelerate the uncovering of trends among millions of cancer patients—and sharing that knowledge—we can enable more data-driven decisionmaking. Oncologists and all caregivers in the cancer treatment ecosystem charged with deciding on a course of treatment, can get new insights in seconds, not years.

More hurdles remain, including the ongoing need for innovation, investment, quality assessment, and data sharing. But CancerLinQ is already proving how partners in the collaborative community can harness data to make decisions that support enhanced care for cancer patients and will ultimately improve the lives of millions of people around the world.

David Delaney, MD, is chief medical officer for healthcare at SAP, and Kevin Fitzpatrick, is CEO of CancerLinQ LLC, a wholly owned subsidiary of the American Society of Clinical Oncology. For further information, contact CLP chief editor Steve Halasey via [email protected].


  1. CancerLinQ [home page]. Alexandria, Va: CancerLinQ, 2016. Available at Accessed August 22, 2016.
  2. Lewis JH, Kilgore ML, Goldman DP, et al. Participation of patients 65 years of age or older in cancer clinical trials. J Clin Oncol. 2003;21(7):1383–1389; doi: 10.1200/jco.2003.08.010.
  3. Greenwood A. Taking action to diversify clinical cancer research. NCI Cancer Bulletin, Special Issue: Clinical Trials Enrollment. Available at: Accessed August 29, 2016.
  4. Mitchell AP, Harrison MR, George DJ, et al. Clinical trial subjects compared to “real-world” patients: generalizability of renal cell carcinoma trials. Presentation at the 2014 annual meeting of the American Society of Clinical Oncology [abstract]. J Clin Oncol. 2014;32(suppl):abstract 6510. Accessed August 29, 2016.
  5. The state of cancer care in America: 2016. Alexandria, Va: American Society of Clinical Oncology, 2016. Available at: Accessed August 22, 2016.