Changing the way to look at cancer: SpIntellx introduces Spatial Intelligence and Explainable Artificial Intelligence to improve pathology and biopharma research

1.6 million breast biopsies are performed each year in the US to diagnose cancer as a follow-up to suspicious findings through diagnostic tests, such as mammograms1. During a breast biopsy, a tissue specimen from the concerning area is removed, chemically and physically processed, fixed on a slide and stained to be examined under a microscope by a pathologist. These histological samples collected from patients fall into a wide spectrum, ranging from benign to high-risk lesions to cancer, and this characterization process can be highly subjective depending on the pathologist making the decision. According to a recent study supported by the National Cancer Institute and the National Cancer Institute-funded Breast Cancer Surveillance Consortium, there is substantial discordance between pathologists’ diagnoses – particularly for the high-risk lesions1. Therefore, there is need for more consistent and reliable methods of characterization to avoid chances of misdiagnosis, which can result in over or under-treating the patient.

There have been efforts to develop pathology platforms that digitize pathology images taken through a microscope that would be traditionally viewed by a pathologist. These same platforms can use digital analytics tools or advanced computational pathology techniques that apply artificial intelligence (AI) and machine learning to these digitized data. The goal is to improve the process of decision-making in pathology in order to deliver the best care to patients. The challenge is that tumor growth is a complex and dynamic process with many biological and chemical constituents, including different cell types, vascular networks consisting of the lymphatic and blood vessels, and the components of the extracellular matrix. The interaction among these cellular and non-cellular structures within the tumor niche is defined as the tumor microenvironment (TME). TME is now known to play a critical role in the development of malignancies. When the TME is in a healthy state, it protects against cancer, and when it’s not, it can support cancer. Existing methods of digital pathology are limited in the ability to characterize these critical changes in the TME with a spatial context, in other words how the cells and the biological components are oriented to each other. With greater spatial knowledge of the tumor microenvironment, there will be a higher level of breast cancer prognostic and diagnostic capability.
A team of clinicians and scientists at the University of Pittsburgh School of Medicine, consisting of S. Chakra Chennubhotla, PhD, D. Lansing Taylor, PhD, Michael J. Becich, MD, PhD, Jeffrey L. Fine, MD and A. Burak Tosun, PhD, has been developing tools called HistoMaprTM and TumorMaprTM that successfully address and overcome these issues. What is novel about their technology?

    1. They build their computational models based on spatial intelligence, i.e. understanding of spatial relationships among the cellular and non-cellular structures in a TME, going beyond methods that just identify structures and count cells in a digitized image of a pathology sample.

    2. They apply explainable AI (xAI) to these digitized pathology images. xAI is an emerging concept referring to “AI that can justify its results with data” to help the pathologist understand the mechanism by which they make a decision towards a diagnosis, prognosis or a therapeutic strategy. xAI introduces transparency and causality to these tools.
“Much of the current artificial intelligence in healthcare is primarily black box: algorithms make a prediction, but they don’t explain why. With xAI, and that’s what our team brings to the table, the algorithms make their prediction and inform the pathologist or the oncologist why they made that prediction. Our team has led the pack in the implementation of xAI in biomedical sciences, specifically in pathology.“

D. Lansing Taylor, PhD, Chairman and Co-founder of SpIntellx and the Director of the University of Pittsburgh Drug Discovery Institute.

HistoMaprTM: A computational guide for pathologists

HistoMapr has been developed for analysis of microscopic structures across cross sections of pathology samples. It quickly identifies regions of interest of diagnostic significance. By using xAI, HistoMapr can explain why these regions are considered as high-risk lesions, whether the patient is at risk of developing breast cancer or not. Currently, HistoMapr is used as a guide or a training tool to aid pathologists’ decision-making in cancer diagnosis.
“The idea is triaging the patient biopsy samples so they are passed to the expert pathologist to make the final read. HistoMapr can also be used as a training system to support pathologists’ decision making with the help of its xAI feature. This feature could be a game changer for the pathologists in adopting digital pathology.”

Chakra Chennubhotla, PhD, CEO and Co-founder of SpIntellx and Associate Professor at Department of Computational and Systems Biology at the University of Pittsburgh School of Medicine.

According to A. Burak Tosun, PhD, who recently joined the SpIntellx team as the Lead Engineer, based on his conversations with many pathologists within the last few months:

“What really differentiates HistoMapr from rest of the available the digital pathology tools are both the xAI feature, which comes out as a “why?” button on the user interface, and its ability to identify regions of interest on its own, which makes it extremely time-efficient. xAI helps the pathologists feel more connected with HistoMapr as they see the connection between the computer’s decision and their own decisions.” – A. Burak Tosun, PhD, Lead Engineer for HistoMapr and the Principal Investigator of the Small Business Innovation Research (SBIR) Phase I grant for HistoMapr.
Another team member, co-founder Jeffrey Fine, MD, who is a breast and gynecologic pathologist at UPMC Magee Women’s Hospital with extensive experience in pathology imaging and computational pathology agrees:

"Advanced spatial analytics, based on xAI, will be a key driver for augmenting pathologist diagnosis of individual patients. Pathologists can understand xAI because it is transparent; this means that pathologists have all the information they need to make the very best diagnoses. This will revolutionize precision medicine." – Jeffrey L. Fine, MD, Associate Professor of Pathology and Director of the Advanced Imaging and Image Analysis Subdivision (Pathology Informatics).
HistoMapr has provided 56% faster case review and more accurate (83% vs. 52%) diagnosis of high-risk lesions, which are considered as challenging cases in pathology, in a pilot study on breast tissue biopsies. The SpIntellx team is grateful to the team of pathologists at Ob/Gyn Pathology Department at Magee-Womens Hospital for their participation in these pilot studies2.

TumorMaprTM: A tool to improve personalized therapeutic strategies, diagnostic and prognostic tests for cancer

TumorMapr uses data that has already been generated by third-party platforms, which take the primary tumor sample, digitize it, take fluorescence biomarker pictures and run the sample analytics, such as counting cells.

“It turns this data into knowledge, identifies tumor hotspots with cancer activity, or regions of interests, and then infers the network biology that is driving these regions to be hotspots by applying spatial intelligence on multiplexed and highly multiplexed (hyperplexed) fluorescence images.” – S. Chakra Chennubhotla, PhD.

TumorMapr has the ability to build networks and make connections between these selected hotspots that are positive for cancer biomarkers. The xAI feature in TumorMapr explains why a specific patient would not respond to a certain treatment or what components of the network can be targeted as a therapeutic strategy. This enables precision medicine, as it can have impact on prognosis and therapeutic strategies for each individual case.

“Every patient has a variation in their network biology within these identified tumor hotspots. Existing targeted therapies usually focus on one or two molecules involved in cancer pathways, but by using TumorMapr, we are taking an unbiased approach to see what’s out there, what are the networks involved and how we can use that network information to better build prognostic and diagnostic tests and/or whether we can personalize the treatment. This is a game changer in the field of precision oncology.” – D. Lansing Taylor, PhD.

TumorMapr has been tested on primary tumor tissue samples collected from 432 chemo-naïve colorectal cancer patients. It successfully predicted the 5-year risk of cancer recurrence with significantly higher diagnostic ability compared to existing state-of-the-art methods3.

The path towards commercializing

In September 2017, the team started their own company, SpIntellx, that stands for Spatial Intelligence and xAI. Their first step towards commercializing their technology involved sending an application for early translational research funding through Pitt’s Center for Commercial Applications of Healthcare Data (CCA). Teaming up with sciVelo, they received funding and translational science support that helped to spawn their first pilot study validating HistoMapr on breast cancer samples.

The company was recently awarded with Small Business Innovation Research (SBIR) Phase I funding through the National Science Foundation (NSF) supporting HistoMapr. This SBIR program enabled SpIntellx team to connect with more pathologists and to have access to more pathology data for further validation of their technology.
“The sciVelo team was instrumental in working with the founders to establish a translational science critical path and to gain support from the CCA for a pilot study on breast biopsies. sciVelo helped us frame a high-impact pitch deck for CCA funding, which we continued to adapt and use for many of our subsequent presentations, including the pitch for seed funding and in the writing of our SBIR Phase I proposal on HistoMapr.”

S. Chakra Chennubhotla, PhD.

Upon receiving the first SBIR Phase I funding, the team decided to move the technology out of the university as a licensed spin-out, so that they could accelerate their growth. According to Dr. Chennubhotla, Rob Racunas, JD, MS of the Innovation Institute was a major resource in licensing and commercializing the technology from Pitt. For faculty interested in advancing their science toward commercial applications, and therefore pursuing similar opportunities to the SBIR funding but not sure how to proceed, Dr. Chennubhotla advises to get involved early with Office of Research, Innovation Institute, and sciVelo at the University of Pittsburgh.

So far the SpIntellx team has raised more than $1M, filed 7 patent applications (2 PCT patents issued and 5 pending) and they hold a software license through Pitt’s Innovation Institute. They are continuing to pursue SBIR funding through the National Institute of Health (NIH) to further provide opportunities to run validation studies for use of TumorMapr in precision oncology.

If you are interested in learning more about SpIntellx, click here.

Author/Photographer: Ceren Tuzmen, PhD