His program could speed U.S. kidney transplants
Share with others:
For six years, Carnegie Mellon University computer scientist Tuomas Sandholm has been refining algorithms for a computer program that would make a truly national living donor kidney exchange program work, potentially saving hundreds of lives a year.
On Dec. 6 living donors and recipients -- who didn't know each other -- in St. Louis and Lebanon, N.H., exchanged kidneys that had been matched by Dr. Sandholm's program.
Planes with donor kidneys flew from St. Louis and Lebanon to Allegheny County Airport in West Mifflin -- which just happened to be about halfway between the two spots -- where the kidneys on ice were traded and then flown back to be transplanted.
"I think that was awesome," Dr. Sandholm said this week about the transplants. "It is amazing. There were politics along the way and people had differing opinions about how it should be done, but now it has finally started."
"It is a big reward to be able to help humanity," he said.
They were just the first two transplants of what Dr. Sandholm and others involved in the program hope are eventually 2,000 to 3,000 living donor kidney transplants from unrelated people each year. That's a much needed increase when 93,000 people are on the national waiting list for kidneys and only about 17,000 a year get transplants.
"This is one of the most positive things we have going on now in our effort to try to allocate these scarce resources," said Ken Andreoni, an Ohio State University Medical Center transplant surgeon. "It was a long time in coming."
For five years, until last year, Dr. Andreoni ran the United Network for Organ Sharing's committee that oversaw the creation of what is now a pilot program for "paired exchanges," that is, living donor kidney transplants between unrelated people who don't know each other.
If the current pilot program -- involving 77 of the nation's live donor kidney transplant centers with hundreds of potential donors and recipients -- goes well, it could be expanded to all 234 such transplant centers and their thousands of potential donors and recipients, in one to two years.
"And that is the way it should be," said Surendra Shenoy, the transplant surgeon at Barnes-Jewish Hospital in St. Louis who removed the kidney from a living donor there to be flown to New Hampshire. "So now when you have a lot of pairs [of a donor and a recipient] you can do lots of exchanges."
"This is a landmark event in the history of kidney transplants," he said.
Paired exchanges came about because so many people who need kidneys are able to find someone they know to donate a kidney, but many of those possible donors are incompatible for a variety of physical reasons, including mismatched blood types.
The idea behind paired exchanges is to get all those people willing to donate a kidney, but incompatible with their friend or family member, to, instead, give it to a stranger identified by the national network.
Early on in the work to develop the national network, Dr. Andreoni said, it was obvious "that, really, the game is mathematics."
To dramatically increase the numbers -- to increase the possible matches -- they would need top level computer programmers and mathematicians to help them.
That's because if you have, say, 100 potential donors and 100 recipients, the number of variables involved -- from matching blood types, to accounting for which recipients are highly sensitized because of their antibodies, to which donors don't want to leave town to make a donation, and others -- quickly evolves into the billions.
In 2004, when Dr. Andreoni's committee was just starting its work, there were just 34 paired exchange kidney transplants across the country.
That number has increased to 301 paired exchanges so far this year, nearly all of them organized by the regional networks, not the national program.
Starting in 2004, several top computer scientists, economists and mathematicians took note of the problem.
One of them was Harvard economist Alvin Roth, an expert on game theory who co-founded one of the first pair exchange networks, the New England Program for Kidney Exchange.
He was speaking at a conference in 2004 when Dr. Sandholm heard him talk about using game theory -- which is used to find the best solutions when there are many, many alternatives -- in organ exchange networks.
Dr. Sandholm, an expert on game theory, was intrigued.
Other computer programs that were being developed to address the problem could handle only several hundred pairs of donors and recipients. Any more than that "and the computer would break down," said Michael Rees, founder of the Alliance for Paired Donations in Maumee, Ohio, who worked with Dr. Sandholm early on.
Dr. Sandholm's breakthrough was to develop algorithms using game theory that broke through the barrier of just hundreds of pairs and instead allowed up to 10,000 pairs -- enough for a national network to be properly analyzed with all of the variables.
The concept may soon go international. Canada is considering creating a similar system. Holland has already asked about Dr. Sandholm's program specifically, and India and Australia have inquired about the idea.
The pilot program now runs matches every four to five weeks. The first run in October initially found three, two-way exchanges, but two of the three sets or "cycles" of exchanges broke down because of compatibility problems.
A second run of matches last week found four, three-way exchanges, meaning 12 people would get new kidneys from live donors if they all worked out. But one of those exchanges has already broken down, too.
Clearly it's a work in progress.
Kathy Niedzwiecki thinks it works just great already.
She is the 52-year-old Pelham, N.H., woman who got a kidney from a woman in St. Louis, thanks to Dr. Sandholm's program.
"I didn't realize I was such a big part of history when I did this. I just wanted to improve my life, be able to help raise my kids and do more with my husband."
First Published December 18, 2010 12:00 am