Paige Turners highlights technology innovators, medical trailblazers, and business visionaries making an impact in the digital pathology, artificial intelligence, and cancer care space.
In this edition, we spoke with Dr. David Klimstra, Founder and Chief Medical Officer at Paige.
Dr. David Klimstra has been on the cutting-edge of pathology since he got his start in the field over 30 years ago. At this time, immunohistochemistry had just revolutionized the practice, enabling him to determine cancer types and subtypes with greater efficiency. Soon after, he was able to adopt advanced technologies like high-throughput sequencing and other molecular techniques to extract deep insight into the unique characteristics of a tumor and help oncologists make precise treatment decisions. Then, just as was happening across many sectors and throughout daily life, digitization was introduced to pathology.
Memorial Sloan Kettering Cancer Center (MSKCC), where Dr. Klimstra had been working for many years, began using whole-slide images scanners to build a large archive of pathology images; An archive that he and colleagues Drs. Thomas Fuchs and Peter Schuffler saw as an opportunity. They wanted to leverage the power of this immense and diverse dataset to train an AI model to have diagnostic capabilities, which would bring never-before-seen efficiency and decision intelligence to the practice of pathology. However, it quickly became clear that the scale necessary to build a model of the size this would require, and the resources needed to bring such a product to the clinical market where it could help patients, exceeded the capabilities of an academic medical center. Thus, Paige was born – built on the desire to take the transformative AI being developed at MSKCC beyond the research environment and deliver it to pathologists to use clinically.
Of course, novel technology is not built and adopted without challenges. At first, the team found that the resources required to create a reliable, patient-safe AI model exceeded even the substantial grant funding they were able to obtain internally from MSKCC and from the Warren Alpert Foundation. Therefore, they needed to convince investors that there was a viable business proposition, despite the fact that they had yet to fully develop or get approval for any products. Thankfully, they were able to overcome these challenges and launch Paige, which has today evolved to offer a full-scale digital pathology platform and robust set of diagnostic and biomarker AI applications. But even now, challenges persist. Pathology remains slow to digital adoption, in large part because of the significant costs required to initiate it, the operational challenges adoption presents, and the lack of reimbursement for using digital pathology. Couple this with a user group that is not motivated to change their pattern of practice, which has served them well for generations, and you have a perfect storm that has prevented AI from realizing its full potential to make a dramatic impact on patient lives.
Dr. Klimstra, along with the entirety of the team at Paige, remains steadfast in his mission to change that. As Chief Medical Officer, he guides the team in building products that will not only help free pathologists from routine, time-consuming tasks so they can focus their skills on the most impactful aspects of diagnosis, but that will help make them even more confident in that diagnosis, which of course would benefit patients. This is achieved by ensuring each Paige product is trained on tens of thousands of slides, rigorously tested and validated in research and in real-world settings, and used in concert with the opinion of the pathologist, so the value is multiplied by the expertise already at hand.
Paige Breast Mitosis, our AI application that supports the automatic detection and quantification of mitoses within invasive breast cancer tissue, is a perfect example. Dr. Klimstra notes that mitotic detection when done manually is laborious and quite subjective, but is important to get right, as it has a direct impact on how patients are treated. With Paige Breast Mitosis, a pathologist can identify a hotspot and quantify the mitoses within that hotspot with unparalleled efficiency and confidence.
This, he says, could have been immensely helpful when he was starting out and throughout his professional years at MSKCC. “I have seen many, many very challenging cases – some of which I could not sort out, and probably some I got wrong. If I could have back all the time I spent scanning lymph nodes, counting mitoses, and looking for vascular invasion, I could have done a lot more research! AI can improve the professional experience and enable pathologists to do so much more with their professional time.”
That is why, beyond working on the development of the most advanced AI tools in the industry, he works hard to push change in the industry, urging labs to undergo digital transformation and tap into the benefits of both digital pathology itself and the AI it enables. It is also why he is so excited for the next era of pathology – one that he and the team at Paige are helping to bring to fruition. Specifically, he is eager to see how prognostic and predictive biomarkers, which have the potential to empower pathologists to provide critical information that cannot be derived from routine slides today, will impact not just the practice of pathology, but importantly, patient treatment and outcomes.
As for the future, he believes that the inevitable consequence of these new technologies will be that an explosion of data, from advanced immunomorphology platforms, genomics, proteomics, discrete clinical data, radiologic studies, etc., will now be available to pathologists, who can extract meaningful clinical information from it that can be practically applied (as opposed to leaving it in clinical records where it will be unclear how and whether it is used). He hopes to see Paige in the center of this next seismic shift in medicine.
Ultimately, he says, his goal has always been to help change how pathologists practice, and to enable them to provide more impactful data for patient management through the use of novel technology, integrated diagnostics, and standardization of diagnostic approaches – and this is the legacy he strives to leave.