Recently, while preparing the Special Report on Organ Models that appears elsewhere in this issue (starting on page 20), I had the opportunity to interview Don Ingber, founding director of the Wyss Institute and scientific founder of Emulate.
In the final couple of minutes of our conversation, Ingber mentioned that the FDA had requested his team develop chips that incorporated cells from women, as well as from men.
“The FDA is very concerned that there is not enough female representation,” he said.
We continued our conversation for a few more minutes before he was called to attend other duties.
Although that specific comment never made it into that article, the thought stuck with me.
Safely ensconced in my home office, face buried in my laptop or reading the latest scientific literature, it is relatively easy for me to block out the external world. Although I am well aware of the various responses to the COVID-19 pandemic and the racial tensions that have exploded around much of the world, these events and social tectonics don’t always have an immediate impact.
Instead, I remain in my safe place, discussing cell culture technologies and dynamic analytics. My conversation is high-throughput vs. high-content, and the needs of human organ models when animal models simply cannot recapitulate the myriad rare diseases to which humans are prone.
But then, in an instant, I am reminded of Ingber’s comment and the FDA’s concern about female representation.
And I am forced to ask, who else is being missed in our probative efforts?
I and many others have written in the past about representation in clinical trials, that so many trials have historically recruited white men.
Women’s hormonal cycles would be unnecessary confounding factors in any study, the refrain rang out. What truly were children other than short adults? And with rare exceptions, a given disease was that given disease regardless of who required treatment for it.
Achingly slowly, this ground was ceded by the research community. Graph after graph, chart after chart, and anecdote after anecdote showed the belief to be a false one in many instances.
A man was not a woman was not a child. A Caucasian American was not an African American was not a Latin American. The limitations of past clinical trials were acknowledged, and more effort has been put into, if not always expanding representation across a study, at least being aware of its lack.
But in the preclinical space, a lung is a lung, right? A cardiomyocyte is a cardiomyocyte?
In discussing the use of stem cell technologies with organ chip design, Olivier Frey, head of technologies and platforms at InSphero, talked about efforts to deal with donor-to-donor variation in preclinical research.
He referenced “labs that use specific donor pools at the early stage, where you try to homogenize the cell types so that you don’t have a high heterogeneity in your results.”
In this way, he suggested, you effectively average out the noise from the individual variations to get a basic understanding of what is going on.
But who are those donors?
When we start a cellular assay, when we develop a high-throughput microtissue screen, when we build an organ chip model, what do we know about the cellular source material? Perhaps I am making too much of this, over-complicating something that is already complicated enough—a common theme in any DDN Special Report. Maybe the heterogeneity within a racial group is as great or greater than the heterogeneity between groups.
I don’t know. I am not sure anyone has thought to ask.
But if our clinical assumptions—that a man was a woman was a child; that a Caucasian American was an African American was a Latin American—were wrong, should we not at least check it out?
A huge impetus of organ chip—and ultimately, human body chip—models is to improve the translation of preclinical results to clinical outcomes.
Cannot part of that translation be a clearer understanding of the materials on which our efforts rest, of the faces behind those cells and tissues?
I gave this commentary the same title as I gave the Special Report because I think both articles are asking the same questions, simply from different perspectives. When it comes to preclinical modeling, are we all in?
Randall C Willis can be reached at willis@ddn-news.com
ADDED July 22, 2020:
RESPONSE FROM: Donald E. Ingber, M.D., Ph.D.
I have always communicated that this is the whole point of human organ chips that animal models (which by definition are genetically homogeneous for them to be commercialized) can not address. I usually end my talks by explaining how big pharma often spend millions of $ and years to do clinical trials with thousands of patients, and that they usually fail. But then they have their statisticians do number crunching, and if they are lucky, they find a small genetically similar subgroup. And then they do another smaller targeted trial, and if they are lucky again, they get it approved for this narrow application.
With human OOC, you can imagine identifying a small genetically similar subgroup up front (e.g., Hispanic women with asthma who are susceptible to airborne particulates), and then make Lung chips from their cells (iPS cells or biopsy samples could be used as sources). Then you test drugs on those chips and if you find ones that work well, you use those same women as patients in your targeted clinical trial. This could revolutionize drug development by decreasing costs dramatically, shortening the time for development, and greatly increasing the likelihood of success.
Best,
Don
