shutterstock_458190886.jpgJoe Keller, Chris Crabtree and Sarah Jegasothy co-wrote this article with Malik Kaman. 

If you’ve missed the transformation of the oncology therapeutic landscape over the past five years, you’ve had your head in the sand. While many recent advancements are hugely beneficial, one result of this tidal wave of investment is that some new drugs offer only marginal improvements yet command outrageous price tags. This is a controversial issue and should give everyone in oncology pause. 

Three additional trends include: 

  • Molecular diagnostics and liquid biopsy will transform treatment and monitoring.
  • Big data and artificial intelligence will disrupt healthcare delivery over the next 10 years.
  • Access to clinical trials continues to be a major challenge for patients and physicians.

With these converging trends, it’s time to seriously consider some fundamental changes in the way that we design and execute trials. The unintended consequences of today’s clinical trial process include prioritizing incremental innovation over public health, stakeholders with counterposed incentives, and critical data that remains siloed or difficult to access. If we were to design the system from scratch, it would look dramatically different. 

With that background, we humbly propose the following idea: What if every interested cancer patient gets enrolled into a large-scale adaptive trial supported by a centralized bioinformatics database? Patients would be assigned to appropriate cohorts based on their treatment goals (palliation vs. cure), age, disease stage and molecular/histological subtype. Each cohort would then be part of ongoing, adaptive A/B experiments designed to test and improve upon the current standard of care. Once a given intervention had “proven” superiority over another (or not), it would then appropriately replace the old standard of care for all other future cohorts in the trial. 

In this way, the standard of care for all patients would constantly evolve and improve. Researchers, investigators and pharmaceutical manufacturers could propose new improvements as they do today, but they would stop running their own independent trials. The A/B assessments wouldn’t have to be “winner take all” assessments. It’s actually good to have multiple, equally efficacious options available for patients. Most importantly, the data would be centrally collected and curated (both successes and failures), and made publically available via open source standards for anyone to analyze. 


BLOG POST: Showdown at the 'IO' Corral

BLOG POST: Accelerating Real-World Evidence to Advance the Trial Landscape

The downstream bioinformatics would allow clinicians to make better, more evidence-based decisions for patients. It also would help governments, regulators and payers make much more informed decisions about access and reimbursement. The biopharmaceutical industry would benefit, too. Today, manufacturers shoulder sole responsibility for the most expensive and risky aspects of drug development. This new system would unburden pharma to collaborate much more openly with physicians and institutions about unmet needs and how to address them.

The idea isn’t nearly as far-fetched as you might believe. In fact, there are a few trials starting to discuss this topic already. The National Cancer Institute and the American Society of Clinical Oncology have taken meaningful steps in the direction of precision medicine, but we need to take this idea even further.

The solution lies at the local level. Much of the infrastructure required to begin exploring this type of approach already exists. What we lack is an aligned vision and the political will to make it happen. In this regard, Europe may have an opportunity to assume leadership that America can’t match. Many of Europe’s local health technology assessments (HTAs), payers and hospitals run significantly more centralized processes than exist in the U.S. Many countries could initiate pilots to further develop this concept. It will require significant compromises across stakeholder groups, but it’s certainly possible—and we can’t afford to wait.

The war on cancer started more than 40 years ago, and we’ve made profound progress on some tumor types since then. However, unmet needs remain extremely high for many cancer patients and the prospect for rapid innovation is being kept in check by arcane processes. The time for leadership and change is now. 

We call on any academic institution, HTA, payer or governmental agency to ask, Are we ready to pursue this idea further? If so, let us know. We challenge the industry to dream big, not because it’s easy or convenient, but because the potential for a cure is within reach and we should no longer be willing to delay.

Topics: patient outcomes, oncology, Pharma, drug development, cancer, R&D, clinical trial design, clinical trials, bioinformatics database