© 2002 American Thoracic Society
Cystic FibrosisWhen to Refer for Lung Transplantation–Is the Answer Clear?Divisions of Pulmonary Medicine and Cardiothoracic Surgery University of North Carolina School of Medicine Chapel Hill, North Carolina Patients with cystic fibrosis (CF) lung disease comprise the second largest group of patients currently undergoing lung transplantation in the United States (1). Time accrued while awaiting organs (an average of about 2 years in the United States) dictates the allocation of donor lungs, rather than urgency of need. Because CF may have an unpredictable course, timing of referral for transplantation has become a critical issue. In 1998, a consensus statement published by a panel of national experts declared "prognostication of survival is inexact, and the timing of referral is difficult," in contrast to idiopathic pulmonary fibrosis that has relatively good predictors of mortality and also may require lung transplantation (2, 3). In theory, for CF, one could compute and compare the mortality probability for a patient with and without transplantation and evaluate the risk to benefit ratio. Although posttransplant CF mortality data are available from the United Network for Organ Sharing, predicting the natural history of CF is less precise. Can one predict short-term mortality in CF? That is the question that Mayer-Hamblett and colleagues (4) attempt to answer in a study published in the current issue of AJRCCM (pp. 1550–1555). They developed and validated predictors of 2-year mortality from the largest CF patient data set to date (n = 14,572 patients), gathered from the 1996 National Cystic Fibrosis Foundation Patient Registry. Two-year mortality was chosen because this is the average time in the United States between listing and transplantation, and accurate predictors might facilitate the transplant referral process. To test the hypothesis, they compared the new model to the currently used lung transplant–referral criterion, the FEV1. Significant predictors of 2-year mortality were increasing numbers of hospitalizations, increasing use of intravenous antibiotics, colonization with Burkholderia cepacia and Pseudomonas aeruginosa, increased height and age, and lower lung function (FEV1% predicted). The highest odds of dying in 2 years (4.1) were in patients infected with both B. cepacia and P. aeruginosa. Despite its sophistication, the model did no better at predicting 2-year mortality than use of the FEV1 criterion alone (see Figure 1 in Reference 4). Both the new model and FEV1 had high negative predictive values and low positive predictive values; that is, both can predict reasonably well who will survive 2 years, but neither can predict well who will die within 2 years. Thus, one could argue that the most important message to emerge from the study was negative, and the reader is left wondering how this study matches up against other published data, and what is new?
Using a fairly large data set of patients (n = 673) followed at the Toronto CF Center for several years, Kerem and colleagues suggested that an FEV1 of less than 30% predicted was the most significant predictor of mortality (about 50% 2-year mortality) (5). A more sophisticated approach was used by Liou and colleagues, who developed a multivariate model to predict mortality in a larger national data set of patients (n = Mayer-Hamblett and coworkers, by way of introduction and discussion, declare, "premature referral for lung transplantation could shorten life expectancy" (4). It is important to distinguish premature referral from premature transplantation. Evaluation and wait listing does not commit the patient to "premature transplantation." It is far better for a lung transplant team to see the patient too early rather than too late, so that the patient is not too sick to wait nor too sick to undergo intensive rehabilitation and preparation for the surgery. If the patient stabilizes because of intensified treatment when sufficient time has accrued for seniority, they can be declared "inactive," with the option of reactivating once the clinical situation demands it. The large numbers of patients with CF who die waiting for lung transplant attests to the problem of referral "too late" rather than "too early" (9). The lung allocation algorithm is currently under review by a subcommittee of the United Network for Organ Sharing Thoracic Organ Committee, to comply with the "Final Rule" promulgated by the Department of Health and Human Services, which requires the Organ Procurement and Tissue Network contractor (United Network for Organ Sharing) to distribute organs to those most in need, minimizing effects of geography and waiting time. The subcommittee is attempting to create an algorithm based on risk of dying on the waiting list, balanced by probability of survival post-transplant (9, 10). This type of allocation system would benefit from studies like that of Mayer-Hamblett and colleagues that identify factors that help predict mortality, to direct lungs to those most in need. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
5,800) (
