A Machine with a Thousand Moving Parts that Don’t Work
Entrepreneurs reveal the systemic reasons why cancer patients cannot access emerging therapies
(Part 2 of a 3-part series)
A few years ago, at the Partnering for Cures conference, Kathy Giusti, the powerhouse leader of the Multiple Myeloma Research Foundation and a myeloma patient herself, was asked what advice she would give patients. She said that patients need to know that they have to be their own best advocates. They cannot expect the system to give them the answers, because there is no functional system. Nowhere is that truer than in the realm of clinical trial matching.
However, there is little that patients, or even the hardest-working physicians, can do to find a trial if they do not have access to clear information or a way to sift through it. That was the mission of the startup that I was running at the time, called Cure Forward. For each individual patient that came through our service, we assembled their medical records to identify trials that could be suitable for them, and then we presented those patients to trial sponsors to confirm the matches. It worked, sometimes. But not enough to make the business work — we closed the doors in 2017, but not without learning a lot about cancer trials.
I’m not the only one to attack that problem. There are many others like me. We have all encountered the same systemic issues, and we leverage our own expertise and networks to attempt new solutions. In my new role at Biden Cancer Initiative, I’ve been talking with every tech innovator I can find, and it turns out that just about everyone starts by creating the same building blocks, like clinical-trial-relevant terminology dictionaries and selection rule libraries. It’s pretty dry stuff, but people take pride in doing it well. It’s the table-stakes to do clinical trial matching, and people only do it because the existing system doesn’t provide the resources they need. These efforts are redundant and will lead to further fragmentation.
Those entrepreneurs all then encounter the next big problem. They use their meticulously structured databases, based on the information available in public resources like clinicaltrials.gov (managed by the National Library of Medicine) or the Clinical Trial Reporting Program (managed by NCI and available to tech developers), to generate patient-trial matches. They present patient “leads” to investigators and trial “opportunities” to patients or their doctors, but those matches are quickly invalidated, because of inaccurate or incomplete information in those public sources. This leads to further frustration for patients and for investigators. One innovator I spoke with did a systematic survey and found a clinically consequential data deficiency in 50% of public trial postings.
One thing to know about entrepreneurs, especially ones motivated by a mission to save the lives of cancer patients, is that they always find a way. At Cure Forward, we called this “Always Forward”, and it’s something that still motivates me today.
For trial matching, the way past the public-information quality problem is to obtain the scientific protocols that are used to run the trials. These are confidential documents, held by trial sponsors and investigators. Entrepreneurial companies start partnerships with biopharma companies or with medical institutions, and they obtain the protocols under confidentiality for the purpose of patient-trial matching.
When they use these accurate clinical trial details in their matching algorithms, the tech innovators generate matches that hold up to scrutiny. Time for champagne? Well, no. Those celebrations are short-lived. While some of the matches unlock access to previously unknown and exciting therapies, that’s usually not the case.
And now we get into the next real problem. The hot trials — the ones for the exciting new therapies –aren’t available to hospitals outside the elite academic medical centers like Dana-Farber Cancer Institute or MD Anderson Cancer Center. Those centers have massive collections of clinical trials, and they routinely match their patients to trials. Trial sponsors believe that hospitals besides the academic medical centers aren’t a reliable source of patients, because their trial infrastructure, including matching processes, often isn’t as robust. Instead, sponsors activate their trials with the investigators they know, at the centers with the most mature and intensive patient-matching and trial-execution apparatus.
In reviewing the clinical trial offerings at several community-based multi-hospital health systems, it becomes apparent that the menu is pretty thin. Some of the more advanced community-based systems with research departments offer the trials administered by NCI Clinical Trials Cooperative Groups. These grant-funded trials sometimes look at new ways of applying existing treatments like chemotherapies. They pose important research questions, but they just aren’t as the same as a chance to access the newest immunotherapy or cell therapy trials. Community hospital systems sometimes offer some of the new therapy trials, but these are often limited to the ones that need to cast a very wide net to find very rare patient types, which activate those sites on the narrow possibility that one of them might find a patient. As such, the chances of actually matching to one of those trials is pretty slim. The result is that even when those patients go through rigorous trial matching among the trials available at their health system, they just don’t get very clinically desirable options, so they don’t enroll. For the entrepreneurs, this is a miserable outcome — their solution works as it should, but it doesn’t make a difference! Perhaps if clinical trial capabilities, including recruitment resources, were more evenly distributed, hospitals beyond the academic medical centers would also be able to operate trials, and patients at those community hospitals would have improved access to potentially effective new treatments. Furthermore, this could result in patients from more diverse socioeconomic and racial backgrounds participating in trials, resulting in a better understanding of how those new treatments work across subpopulations.
Let’s return to the story of Norman, a fictional cancer patient introduced in the first post of this series. Why did the system let him down? Well, the standard treatment didn’t work for Norman and his health institution probably didn’t have access to the most promising new treatment trials. His care team went above and beyond, looking for trial options. They were limited by the same systemic challenges that entrepreneurs have confronted, but with a lot less time and technology to back them up. It’s no surprise Norman never found the treatment that might have made a difference.
The Biden Cancer initiative has assembled a working group on clinical trial recruitment effectiveness, which I lead. My next post will describe our ideas for how we can break out of this inefficiency, so that we stop failing patients like Norman.