Digital Pathology, Computational Biology, Organ Transplants, & the Future of Medicine
Posted 23rd March 2020 by Liv Sewell
Both transplant outcomes and lab methods have stagnated over the last 40 years. Ishita Moghe and Professor Kim Solez comment upon the rapidly changing landscape of medical research and the potential of digital pathology for transforming patient outcomes.
The relationship between lab methods and transplant outcomes
Organ transplants became routine in the 1980s with the breakthrough development of immune-suppressant drugs, imaging techniques, anti-microbial agents and donor-recipient matching. Since the 1980s organ transplants have radically improved short-term survival for those patients who have been able to receive them. More than 50% of patients receiving liver and heart transplants will survive for 7 years, only 25% of patients on the waiting list will.2
Yet long-term survival duration has not changed for organ transplantations. According to Abbas Rana (MD) and Elizabeth Louise Godfrey (BSBE), a patient was just as likely to survive 5, 10, or even 20 years with an 1980s-era organ transplant as is a patient with a transplant in recent years.3 Causes of death have remained the same, with no significant change in deaths from graft failure or infection, and death from malignancy increasing somewhat over time.4
In this presentation Ishita Moghe (BSc) and Kim Solez (MD) highlight how methods of organ analysis have also stagnated over the last 40 years. Could transplant outcomes be linked to lab methods? What could the technologies which are transforming pathology mean for organ transplantation outcomes?
Organ analysis in the digital age: single-cell sequencing & machine learning
Moghe and Solez hail a renaissance in medical research made possible by artificial intelligence, machine learning, and the huge amount of data generated by scRNA-seq, digital whole slide imaging, and electronic patient records.
Single-cell RNA sequencing has found new types of human blood dendritic cells, monocytes, and progenitors resulting in a doubling of the number of known cell types. Moghe and Solez predict analogously, that as scRNA-seq is applied further, new cell subtypes will be found in all organs, doubling the number of known cell types in every organ. This will lead to better disease classification and in transplantation, better donor-matching.
Donor matching is key because graft failure is one of the main reasons for death in transplantation patients. 7% of patients’ kidney transplants have failed within 1 year and 17% of patients’ kidney transplants have failed within 3 years.5 Currently, it is difficult to predict transplantation outcomes because such a wide variety of diverse factors in donor & recipient histories contribute. Massive multi-factor analysis of donor and recipient history could be applied to better predict prognosis and reduce graft failure.
Digital pathology and the future of medicine
For Moghe and Solez, putting newly available data through AI structures is going to yield actionable, integrated information on gene expression, risk factors, precision phenotypes, diagnostics and targeted treatment in organ transplantation and across healthcare.
In this presentation, Moghe and Solez situate digital pathology in its larger context, the future of medicine, and open up new horizons for pathology classification, organ transplantation and healthcare more widely.
What will the data renaissance mean for you?
Watch Ishita Moghe (BSc) and Kim Solez (MD) present on ‘Kidney Medicine in the Digital Age: Single Cell Sequencing and Artificial Intelligence’
1 Rana, A., & Godfrey, E. L. (2019). Outcomes in Solid-Organ Transplantation: Success and Stagnation. Texas Heart Institute Journal, 46(1), 75.
5 Lara E. Tushla, ‘When a Transplant Fails’, National Kidney Foundation Website, available at https://www.kidney.org/transplantation/transaction/TC/summer09/TCsm09_TransplantFails (accessed on 20/11/2019).
Leave a Reply