Guidelines for donor lung selection: time for revision?

Few data support current guidelines for donor selection in lung transplantation. We determined degree of compliance with current donor guidelines, effect of these and variances on survival, and other donor factors predicting survival.

From July 1999 to June 2008, 10,333 primary transplants were performed in the US, with United Network for Organ Sharing data available for age, ABO type, chest radiograph, arterial difference in partial pressure of oxygen (PaO(2)) greater than 300 on 100% fraction of inspired oxygen, smoking, absence of aspiration/sepsis, and purulent secretions. Multivariable survival methods were used to determine relevance of these and new variables, adjusted for recipient risk factors.

In 56% of transplants, variance from at least one guideline was observed: chest radiograph, 41%; smoking, 21%; and PaO(2), 18%; but rarely ABO compatibility (0.06%). Practice within guidelines was not associated with increased mortality. Common variances from guidelines; eg, PaO(2)/fraction of inspired oxygen down to 230, were not associated with increased mortality, but smoking (p = 0.02) was. New donor variables associated with increased mortality were diabetes (p = 0.001), presence of cytomegalovirus antibodies (p < 0.0001), recent smoking history (p = 0.02), African-American (p = 0.005), blood type A (p = 0.02), death other than from head trauma (p = 0.02), and gender (p = 0.02), race (p = 0.03), and size (p = 0.002) discordances.

Variance from current donor guidelines for lung transplantation is frequent; analysis suggests that donor PaO(2) ranges can be widened and a suspicious chest radiograph, evidence of sepsis, and purulent bronchial secretions ignored. Older age and smoking history appear to have a minor impact. New and possibly important factors identified suggest the need to better understand the impact of a wider range of donor variables on recipient outcomes.