AI allows donors and recipients to match for liver transplants

Like a matchmaking app, but with a much more important purpose: saving lives. A team from the Reina Sofía University Hospital in Córdoba has developed an artificial intelligence (AI) model that predicts the success of donor-recipient matching in liver transplants .

The system has been tested on patients in a project involving the 24 Spanish adult liver transplant centers and the National Transplant Organization (ONT), coordinated by its director, Beatriz Domínguez Gil.

This research project, which received financial support from the Mutua Madrileña Foundation through the Maimonides Institute for Biomedical Research in Córdoba, was led by Dr. Javier Briceño, Head of the General and Digestive Surgery Department and Head of the Liver and Pancreas Transplant Programs at the Reina Sofía University Hospital in Córdoba. It also included the participation of experts from the University of Córdoba's Computing Unit (César Hervás and Pedro Antonio Gutiérrez) and Dr. Rafael Calleja, from the same hospital.

The project's goal was to develop an AI system that could help match liver donors in asystole (a type of donation known as "heart failure") with recipients on the waiting list, which would, a priori, achieve greater patient survival and avoid problems with graft loss.

The results of the work, published in the scientific journal 'Transplantation', are based on a series of 539 donor-recipient pairs , the largest volume recorded to date worldwide of donors with asystole of this nature.

In any case, Dr. Javier Briceño emphasized that the final decision on the allocation of the transplanted organ will be the physician's.

Asystole donation, or " heart-stopping donation ," refers to organs obtained from donors who are in cardiac arrest, not brain-dead donors. Although these are more difficult to obtain than those obtained from brain-dead donors, their availability has increased the number of organs available for transplant.

The greatest difficulty with these organs is due to the fact that in Spain there is a five-minute period between the donor's heart stopping, the date of death being certified, and the start of the organ harvesting. Since this waiting period can harm the organ's condition, it receives during those minutes what is called "regional normothermic perfusion." This technique consists of "connecting the donor's blood vessels to a heart-lung machine, recirculating the blood in the deceased donor so that the organs are in the best possible condition for transplantation," explains project coordinator Dr. Javier Briceño.

This technique is not used in other European countries or in the United States, "where once death is certified, the organs are extracted very quickly, but this can have repercussions, especially in the liver, and lead to graft viability problems," the expert adds.

The work published in the scientific journal 'Transplantation' is the latest step in a line of work that, until this project, had been conducted on brain-dead donors. The MADRE study (donor-recipient matching model in Spain) was conducted in 2009-2014, which pioneered the analysis of how proposed donors can be assigned to a liver transplant candidate waiting list using algorithms based on artificial intelligence. Eleven national centers participated in the study, including the Reina Sofía University Hospital in Córdoba. This study was subsequently validated in the United Kingdom (2017) and in the United Network of Organ Sharing (UNOS) database in the United States.