Activating Patients: The Path to Personalizing Cancer Treatment

by Brad Power

Today’s cancer patients desperately want access to faster and better treatment options, but despite success stories and great potential, the actual use of diagnostic tests and increased data about individuals to tailor their treatment is amazingly low. Barriers (especially in clinical processes and payment) stand in the way of the rapid adoption that is needed. The path to more rapid adoption is through activated patients and startups trying to hack health care.

The Promise of Personalized Cancer Treatment: Amazing cures.

In 2016, Akshay Mehta,1 then 35 years old, was diagnosed with pancreatic cancer. He was treated with several rounds of gruelling chemotherapy, leaving him very weak from side effects and without significant improvement in his disease progression. But then he arranged for a molecular analysis of his tumor through a patient advocacy group (the Pancreatic Action Network). He learned that he had a very rare genetic mutation (an “ALK fusion”) and that there was a drug that is designed to attack tumors with this mutation, though for other cancer types. He got the drug, and the tumor shrank so much that he was able to undergo surgery in 2017, which allowed him to return to work and spend quality time with his family.

Akshay Mehta’s story demonstrates the power of targeting a cancer treatment to an individual based on a diagnostic test. These targeted therapies avoid treatments that won’t work and find therapies that will work more reliably and with fewer side effects. There is an explosion of diagnostic tests that can provide input to treatment, and targeted therapies are regularly coming to market. According to the Personalized Medicine Coalition (PMC),2 over 40 percent of the new drugs approved by the U.S. Food and Drug Administration in 2018 were personalized medicines.

The Stark Reality: Few cancer patients get a personalized therapy.

In a post3 published last year on PMC’s blog, Education & Advocacy, I shared my story of being at the Personalized Medicine Conference and hearing a leader from the prestigious cancer center where I am being treated say that all of their patients get their tumor tissue analyzed. I stood up and contrasted his claim with my own frustrating experience of repeated requests to get my tumor analyzed. It took over a year for me to get the results. I have since learned that my experience was an example of a broader phenomenon: getting diagnostics and targeted treatments based on the results is still exceedingly rare.

The breadth of the problem emerged further for me last winter when I attended a conference on precision medicine in cancer at the Massachusetts Institute of Technology.4 During the conference, Kenna Mills Shaw, a Ph.D. at the prestigious cancer research center MD Anderson, described how she and her colleagues had analyzed data from 2,000 patients to see whether their tumor DNA was sequenced and whether they got targeted treatments based on the analysis. The bottom line: few did (a percentage in single digits). And this was at MD Anderson, one of the premier academic research centers in the world!

Kenna Shaw’s results were validated by several other studies I found. The percentage of U.S. patients with cancer estimated5 to benefit from genome-targeted therapy was less than 1 percent in 2006 and still less than 5 percent in 2018. In the first phase of a 2015 study6 by the U.S. National Institutes of Health of therapy based on the genetic characteristics of tumors, 5 percent of 795 patients treated over the course of three months received recommendations for targeted treatment assignments, with 2.5 percent electing to accept those recommendations.

A review7 of patients with metastatic colorectal cancer found that less than half were tested for common DNA mutations (RAS and BRAF), despite the recognition of such testing in national guidelines as essential to guide appropriate metastatic colorectal cancer therapy selection. A study commissioned by the Personalized Medicine Coalition (see JCO Clinical Cancer Informatics8) showed that although 30 percent of 5,000 non-small cell lung cancer patients had actionable mutations, only about 19 percent of the 5,000 with actionable mutations (2/3) received targeted treatments.

We are in a vicious cycle: People aren’t getting tested as much as they should be; and even when they are, we aren’t using the data for treatment decision-making as much as we should be. If we don’t use the data, we can’t demonstrate impacts on care and outcomes, and then we can’t get payers to pay for tests.

The difficulty of getting diagnostic tests built into the standard of care isn’t a new phenomenon with more advanced DNA analysis of tumors (“next-generation sequencing”). For decades, there have been tests available to determine a patient’s likelihood of a toxic response to a chemotherapy (such as 5FU and Tamoxifen) before it is prescribed. Yet these tests are not part of the standard of care, even though most cancer patients still get some type of chemotherapy.

From these studies, I propose three reasons why patients aren’t being directed to the best targeted options for them:

  1. Converting Science and Diagnostic Breakthroughs to Breakthrough Drugs: Many genetic mutations are being identified for which there are no associated therapies.
  2. A Lack of Incentives to Adopt: Doctors and patients are not pursuing targeted treatment regimens. A confluence of factors is likely to blame, including:
    • Many doctors and patients are unaware of the rapidly evolving possibilities in personalized medicine, which are complex and hard to understand;
    • Doctors are not always trained to interpret the test results;
    • There is such a strong institutional bias to safety that anything new can look too risky (e.g., doctors are afraid they will be sued if there is a bad outcome, insurers and regulators wait for clear evidence from randomized clinical trials);
    • Targeted strategies are not (yet) enshrined in the standard of care;
    • A lack of incentives or motivations to move doctors to prescribe tests and precision drugs.
  1. Institutional Legacies and Practices: Diagnostics, pharmaceutical, and government agencies tend to have entrenched methods.
    “Fee for service” (vs. “pay for quality”) payment systems motivate doctors to steer patients “in house.” For example, some physicians (pathologists) can be reticent to send samples out for testing that their hospital doesn’t offer because of the loss of revenue;
    • Current diagnostic tests (e.g., gene mutations, “tumor mutational burden”9) haven’t demonstrated adequate predictive power of patients’ response to treatment or adverse effects;
    • Data-sharing is still relatively uncommon;
    • Analytical process challenges (sample problems) are complicating testing workflows;
    • Personalized cancer treatments disrupt the “blockbuster” drug business model since they address each patient uniquely with direct delivery of a targeted therapy (thereby missing potential lost opportunity to develop innovative medicine);
    • Manufacturers are reluctant to provide investigational drugs for testing in combination with drugs from other makers (unless market driven);
    • The reimbursement landscape can make it difficult for patients to get or afford recommended treatments, especially because payers are motivated in part to reduce costs;
    • The clinical trial system is slow (many years) and costly (can be a billion dollars), which is especially problematic for people who have a cancer without a current treatment or a rare cancer.

For these reasons, today’s patients, who desperately want access to faster and better treatment, are unlikely to receive it from risk-averse, profit-seeking institutions with legacy cultures, people, and processes accustomed to one-size-fits-all medicine.

Activated Patients and Startups Can Create the Needed Breakthroughs

To achieve the faster, more disruptive breakthroughs that the system desperately needs, I believe that cancer patients should take a more active role in their own care. They (and their close caregivers) need to educate themselves and advocate for themselves. Physicians can help. Below is what I recommend to anyone who wants to accelerate the needed disruption of cancer treatment and what I’m doing:

Educate yourself about your disease, treatment options, and startups that are disrupting the status quo.

Cancer patients should find out about the possibilities associated with predictive diagnostic tests and targeted therapies, including joining online patient communities, getting second opinions (including exploring access to clinical trials) and finding a “patient advocate.” And they should find and take advantage of innovative industry disruptors. As in the cases of Amazon in retail, Uber in taxi driving, or AirBnB in hotels, the “barbarians at the gates” offer breakthrough services directly to consumers

For example, I’ve joined the Leukemia and Lymphoma Society and compiled maps of startups in each of ten “jobs to be done” for cancer patients.

Consider asking your doctor about clinical trials that you could join. Be part of a cohort in a trial to determine the efficacy of a drug with breakthrough status, especially if no treatment is currently available.

Become an advocate: Insist that you and every cancer patient must be treated with the best available knowledge.

Every patient is empowered to ask questions and insist that his/her treatment includes the best appropriate tests and therapies currently available (“n of 1”).  Patients should be tracked to follow the course of their response (a “perpetual trial”). The real-world data from these personalized experiments should become the basis for continuous learning to prioritize the most promising therapies and slash the time and cost of developing drugs and testing clinical hypotheses.

Patient advocacy groups can help channel this advocacy. Cancer patients need to join groups that will advocate for getting their best personalized treatments now (e.g., drugs chosen because they match their molecular and genetic profile – “genotype”, rather than where their cancer appeared – “phenotype”). For example, the pancreatic cancer advocacy group PanCAN’s “Know Your Tumor” effort has shown10that a comprehensive precision medicine system can be implemented in community and academic settings, with highly actionable findings observed in over 25 percent of pancreatic cancers. And in a recent study11 of patients with previously treated metastatic cancers, all had genomic sequencing of their tumors, and almost half got a personalized combination of drugs.

  • Turn your advocacy into action by being an early adopter of new apps and services, especially from startups, to help the revolutionaries. For example,
    • I’m getting my blood and microbiome analyzed by startups (Natera and Viome).
    • I signed up years ago for genetic analysis (23andme) and take their surveys.
    • I’m going to put my health data into a personal data repository offered by a startup (MyCancerDB).
    • I joined a health research project and donated my health data and specimens (All of Us).
  • Help startups that offer disruptive services directly to consumers.
    • For example, I’m a founding member of a startup that is offering cancer services directly to consumers (MyCancerDB).
    • I’m seeing how I can help publicize companies that offer online advice (“second opinions”), such as Cancer Commons.

In 2018, Brad Power was diagnosed with lymphoma and went through a course of chemotherapy. Building on his experience with process reengineering and managing technology, he is focusing on helping people with a cancer diagnosis get information on the best options for their treatment. He seeks to accelerate two big themes: personalization (gathering lots of data about each individual and customizing unique treatments) and self-care (empowering individuals to partner with startups offering services directly to consumers to take an active role in their treatment). He is the founder of Reengineering Cancer Treatment, is a founding member of MyCancerDB, and is collaborating with Cancer Commons.


Discovery of Rare Mutation Gives Survivor Options, Hope,, Pancreatic Cancer Action Network, Pancreatic Cancer News

Personalized Medicine at FDA: A Progress and Outlook Report,, The Personalized Medicine Coalition

Accelerating Personalization in Cancer Treatment: A Patient’s Perspective,, Personalized Medicine Coalition, Education & Advocacy, A Blog on the Field of Personalized Medicine

The Future of Precision Cancer Medicine,, MIT Center for Precision Cancer Medicine, Inaugural Symposium, Koch Institute for Integrative Cancer Research at MIT

Estimation of the Percentage of US Patients with Cancer Who Benefit from Genome-Driven Oncology,, JAMA Oncology

NCI-MATCH Trial (Molecular Analysis for Therapy Choice),, NIH, National Cancer Institute

Genomic Profiling for KRAS, NRAS, BRAF, Microsatellite Instability, and Mismatch Repair Deficiency among Patients with Metastatic Colon Cancer,, JCO Precision Oncology, An American Society of Clinical Oncology Journal

Cost Effectiveness of Multigene Panel Sequencing for Patients With Advanced Non–Small-Cell Lung,, JCO Clinical Cancer Informatics

Tumor mutational burden definition,, NCI Dictionary

Molecular Profiling of Pancreatic Cancer Patients: Initial Results from the Know Your Tumor Initiative,, Clinical Cancer Research

Molecular Profiling of Cancer Patients Enables Personalized Combination Therapy: The I-PREDICT Study,, Nature Medicine