A controversy has erupted at prestigious cancer center Memorial Sloan Kettering (MSK) about the granting of exclusive access to its surgical pathology slide archive and database to a startup company, Paige.AI. It was been revealed in a recent NYT article that the hospital holds an equity stake in the company. There is also participation in the startup by some hospital insiders (see: Sloan Kettering’s Cozy Deal With Start-Up Ignites a New Uproar). Below is an excerpt from the article:
An artificial intelligence start-up founded by three insiders at Memorial Sloan Kettering Cancer Center debuted with great fanfare in February, with $25 million in venture capital and the promise that it might one day transform how cancer is diagnosed. The company, Paige.AI, is one in a burgeoning field of start-ups that are applying artificial intelligence to health care, yet it has an advantage over many competitors: The company has an exclusive deal to use the cancer center’s vast archive of 25 million patient tissue slides, along with decades of work by its world-renowned pathologists. Memorial Sloan Kettering holds an equity stake in Paige.AI, as does a member of the cancer center’s executive board, the chairman of its pathology department and the head of one of its research laboratories. Three other board members are investors....
Doctors [at MSK} said they were concerned about a lack of communication from hospital leadership, and one said patients were nervous that their health data was being commercialized by the institution. Hospital pathologists have strongly objected to the Paige.AI deal, saying it is unfair that the founders received equity stakes in a company that relies on the pathologists’ expertise and work amassed over 60 years. They also questioned the use of patients’ data — even if it is anonymous — without their knowledge in a profit-driven venture. In addition, experts in nonprofit law and corporate governance have questioned whether Memorial Sloan Kettering, one of the nation’s leading cancer centers, complied with federal and state law governing nonprofits when it set up the deal. The experts pointed out that charitable institutions like Memorial Sloan Kettering must show that they didn’t provide assets to insiders for less than the fair market value.
Computerized analysis of the scanned images of surgical pathology slides is an important and ongoing research area that will lead to more accurate and faster diagnosis of specimens. Algorithms will not replace pathologists any time soon but will certainly improve the accuracy and speed of diagnoses. An essential requirement for such research is access to images of specimens in addition to the diagnoses assigned by pathologists. Hence, exclusive access to superb cancer databases such as the one at MSK is enormously valuable. Here is a short but revealing quote about this process copied from the Paige.AI web site suggesting that Paige is also scanning the MSK slides:
Medical artificial intelligence at an unprecedented scale: Memorial Sloan Kettering Cancer Center scans up to 30K pathology slides each month, an enormous treasure of training data for powerful predictive models. PAIGE is ramping up to scan 100K archive slides per month.
The issue under consideration here is not whether artificial intelligence (AI) can and should be used to improve surgical pathology performance but rather whether it is appropriate for MSK to provide exclusive access to its pathology database to Paige.AI. I can't improve on the following two comments about this issue that were posted on the API listserve on 9/20/2018 by Dr. Brian Jackson. He is an informatician of note and also Medical Director of Support Services, IT and Business Development, at ARUP Laboratories. Boldface emphasis is mine.
- In my opinion, tissue archives (ideally having accompanying patient consent, even if they’re de-identified) should be treated as public resources for the advancement of science broadly. Exclusive licenses for tissue archives, just like patents on DNA sequences, are just flat out immoral. Even if you ignore the embarrassing conflicts of interest. [side note: the pathology department chair gave up his equity stake after the NYTimes reporters started asking questions.]....
- The problem is not commercialization per se. The problem is when scientists and physicians naively act as if commercialization excuses them from basic ethical principles. It’s completely possible to set up companies, even for-profit ones, that are fully transparent and respect scientific and medical ethical values. And that don’t have glaring conflicts of interest. In this case, I’m guessing that eliminating the exclusivity and the personal equity stakes for faculty members might have solved most of the problems. If that meant that certain investors wouldn’t have wanted to invest, then that’s what STTR grants are for. Or maybe a philanthropist. Or bootstrapping using departmental funds. But you don’t prostitute resources that you have stewardship over and then pretend that it’s ok just because it’s “innovation”.