Computer scientists from Glasgow University are playing a key role in the NHS’s living-donor kidney-transplant programme, which saw the first multiple kidney transplants involving three couples take place at the end of last year.
They have devised a computer program that matches donors to patients whose original willing donor is clinically incompatible, enabling organs to be swapped anonymously between people on the same day in different parts of the UK.
One of the first three-way donations to be carried out took place in December 2009 at three different hospitals across the UK, involving six people. All are now in good health following surgery.
The ability to carry out live donations between people who are not genetically or emotionally related only became possible following the introduction of the Human Tissue Act 2004 and the Human Tissue (Scotland) Act 20062.
Before then patients could only receive organs that had been removed from donors who had died, or via a live donation from a compatible family member or friend. If no suitable organ was found in this instance, the patient would have to continue with kidney dialysis while they waited for a donated organ.
Now, however, patients and their donors can register on a paired donation database managed by NHS Blood and Transplant (NHSBT), where matching runs are carried out approximately every three months in a bid to identify potential paired or multiple matches. Currently, around 150 patients and their donors are listed on the database, and it is hoped that this figure will continue to grow.
Their information – including age, blood group and human leukocyte antigen (HLA) antibodies, as well as their location, is fed into a computer by NHSBT staff.
Seconds after receiving an anonymised version of the data that indicates only compatible donor and recipient matches and their ’scores’, the Glasgow program provides information on potential two- and three-way exchange matches, which are then double and triple checked, and returned to NHSBT the same day.
The computer program used by the university to match donors and patients evolved from one originally written by honours student Kirstin MacDonald, under supervision from Dr David Manlove, senior lecturer in the Department of Computing Science, and Dr Péter Biró, a research assistant.
The department has now received a contract from NHSBT to develop the program further so they can use it in house. This would provide results even quicker because it would be on site and work directly from NHSBT’s own database, which could also be updated immediately.
The project to develop the program further will be carried out by Dr Manlove and research assistant Dr Gregg O’Malley, with further involvement from Dr Péter Biró.
The program, written in Matlab, must avoid trying out every possible solution, as the number of permutations can be astronomically large. Dr Manlove said that a key challenge going forward will be to ensure the program continues to work quickly when the data set is much larger.
’When we first started working with the NHSBT in June 2007 it could take 16 hours to find suitable matches involving two-way donations from 100 pairs, so we’ve really cut that time drastically. Because this paired donation-matching program is just one part of a very large process, it wouldn’t be helpful for it to cause unnecessary delays. You really don’t want to wait more than a few minutes, so we’ll need to investigate new techniques and algorithms that will address this challenge,’ he said.