UCLA Liver Transplant Model Predicts Good Outcome: A Newsmaker Interview
With Rafik M. Ghobrial, MD
Laurie Barclay, MD
Nov. 13, 2002 - Editor's Note: The desperate shortage of donor livers available
for transplant is forcing closer scrutiny of transplantation criteria. The Model
for Endstage Liver Disease (MELD) developed at the Mayo Clinic falls short of
the mark, according to a presentation on Nov. 4 by
James F. Trotter, MD, from the University of Colorado Health Sciences Center, at
the 53rd Annual Meeting of the American Association for the Study of Liver
Diseases in Boston, Massachusetts. In a study of 66 liver transplant recipients
with a right lobe transplant from a living donor,
preoperative MELD scores failed to predict which patients or grafts would
survive for at least one year after surgery.
Investigators from the University of California at Los Angeles (UCLA) have
developed a new model which is a far better predictor of outcome, according to a
report in the September issue of the Annals of Surgery. In 46,942
patients with orthotopic transplants over the last 10 years, variables
identified as independent predictors for recipient survival include recipient
age and recipient creatinine, donor age, sex, total bilirubin, prothrombin time,
retransplantation, and warm and cold ischemia times. To learn more about the
UCLA model and its implications for liver transplantation, Medscape's Laurie
Barclay interviewed lead author Rafik M. Ghobrial, MD, an associate professor of
surgery in liver and pancreas transplantation at the Dumont-UCLA Transplant
Center and the David Geffen School of Medicine at UCLA. Dr. Ghobrial is also the
first author of a study of donor and recipient outcomes in right lobe adult
living donor liver transplantation (LDLT), which appears in the October issue of
Liver Transplantation.
Medscape: How well does the MELD model predict recipient survival, graft
survival, and overall outcome after liver transplant?
Dr. Ghobrial: Actually, the MELD model is not a predictor of survival but of
death, that is, how quickly the recipient is likely to die while awaiting
transplant.The MELD model was designed to give priority to the sickest patients
in greatest need of transplantation. So, for example, if a score of 1 represents
the sickest patients who are currently in the ICU, and a score of 3 represents
those at home but still in dire need of a transplant, patients scored 1 are
considered better candidates even though patients scored 3 are more likely to
survive after transplantation.
Medscape: How does the UCLA model differ from the MELD model?
Dr. Ghobrial: The criteria to determine suitability for transplantation are
always changing because there aren't enough organs to go around, so we're trying
to maximize distribution of this precious resource. MELD factors in both waiting
time for transplantation and how sick the patient is. But with time, how long
the patient has been waiting for transplantation loses its effectiveness as a
criterion. The MELD assigns a score to every patient, and
predicts with some accuracy the chance of the patient dying while waiting for a
transplant, as well as the chance of dying after transplantation. The UCLA model
differs in its overall focus and in its specific criteria. It uses a combination
of recipient, donor, and operative factors to predict good outcome. For example,
we know that renal failure negatively impacts survival because some patients may
actually need a double organ tranplant, so the UCLA model includes recipient
creatinine. There's nothing in the MELD model to indicate that a patient has
gone too far to justify transplantation. No one wants to be unsympathetic to a
dying patient, or to tell his family that he can't be transplanted because he's
likely to die anyway. But at the end of the day, using the MELD model leads to
transplanting the sickest patients, while the UCLA model attempts to select
those patients who will make the most use of the organ.
Medscape: What are the advantages of the UCLA model over the MELD model in
predicting outcome? Dr. Ghobrial: The UCLA model was developed over many years
in hepatitis C patients, and it accurately predicts outcome and survival. It
selects those patients most likely to survive the operation, to have a quick
recovery, to have survival of the transplanted organ, and to live longer after
transplantation. The battle between the two models is like a tug-of-war, but
we feel that the UCLA model achieves a better balance through a more logical and
practical approach. We have to offset the urgency of the recipient's need with
the predicted efficiency of organ utilization.
Medscape: How well is the UCLA model accepted in the field? Dr. Ghobrial: It's
very well accepted among groups awaiting transplant because it's an
understandable model. The UCLA model accurately predicted outcome in the 25,000
patients studied retrospectively. Before this model,
we didn't have the data to predict patient survival, but now we can study the
model and validate it prospectively. Prospective validation will take years.
Medscape: How can we best cope with the drastic shortage of available livers for
transplant?
Dr. Ghobrial: Education of the general public, through the Oprah Winfrey show
and other media, has helped increase awareness slightly. But I think education
has gone about as far as it can, and it has not translated into a dramatic
increase in the number of available organs. The key is in changing legislation.
Right now, the legislation is that no one who dies is an organ donor unless the
family agrees to it. We should change that so that everyone who dies is a
potential organ donor unless the family objects. But there are not a lot of
politicians willing to do that.
Medscape: What criteria should factor into deciding who should receive a liver
transplant?
Dr. Ghobrial: If predicted survival based on the UCLA model is less than 70% in
one year, if it falls below threshold criteria, we shouldn't transplant. If
one-year survival probability is at least 80% to 90%, it's okay to
transplant. We need to consider the cost-effectiveness of treating patients who
are very sick, recognizing the pressures of third-party insurers, decreased
reimbursements, and the high cost of drug therapy. The most common cause of
death following transplantation is sepsis, which is very expensive to treat. We
have to take into account that the costs of treating one patient for sepsis -
who ultimately dies - could have been used to treat two
patients who survive. We'll have to study all these factors to help determine
the optimal interventions and decisions. But in general, the sicker the
recipient, the less effective any interventions will be.
Medscape: How difficult is it to apply the UCLA model in practice?
Dr. Ghobrial: If a prospective recipient falls below a certain threshold
indicating low probability of survival, they should be delisted. Everyone agrees
in principle that that's a great idea. But I'm a physician - I've sworn to
preserve life. I am going to do my best for my patient no matter what. If
there's a predetermined national agreement about who should not be transplanted,
it makes my decision easier and more justifiable, even though it's
psychologically, practically, and medicolegally difficult to tell a patient and
his family that he's too sick to qualify for transplant.
Medscape: What are the advantages and disadvantages of LDLT over cadaveric
transplant?
Dr. Ghobrial: One big advantage of LDLT is that you can transplant before the
recipient gets too sick. On the other hand, LDLT is relatively new, and we have
a 15- to 20-year history of cadaveric transplantion, so we have a better
understanding of the complications and operative technique. There has not yet
been a comparably rigorous analysis of LDLT. We also have to consider risk to
the donor with LDLT.
Medscape: What patient selection criteria do you recommend for consideration of
LDLT over cadaveric transplant?
Dr. Ghobrial: Under the right circumstances, given the right donor and recipient
factors, LDLT is clearly a good thing, but it shouldn't be done in hopeless
situations where the recipient is not likely to survive. We should subject
criteria for LDLT to rigorous analysis. When do you get the best
outcome? What is the best utilization of donor tissue? These questions still
need a lot of work before we can find the answers.
Medscape: What are the risks to the donor in LDLT?
Dr. Ghobrial: Death. In the U.S., there have been two perioperative deaths in
healthy donors. Last year, there were about 400 LDLTs performed in the U.S. The
risk of donor death is less than 1.0% - very low, but real. A donor
who is very healthy can end up dying. Depending on how you define complications,
the complication rate in donors ranges from 9% to 67%; probably 20% is a
reasonable overall estimate. It's not like a bone marrow transplant - it's an
extensive procedure. The complications are mostly
minor, but some are life-threatening.
Ann Surg. 2002;236(3):315-322
Liver Transplant. 2002;8(10):901-909
Reviewed by Gary D. Vogin, MD
Related Links
Conference Coverage
53rd Annual Meeting of The American Association for the Study of Liver
Diseases
News
Pre-Op MELD Score Does Not Predict Survival in Living Donor Liver Transplant
Recipients
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.419 / Virus Database: 235 - Release Date: 11/13/2002