Donor-recipient matching in lung transplantation: Which variables are important?

Abstract

OBJECTIVES: Donor to recipient (D/R) matching in lung transplantation (LTx) is usually directed by blood group (identity or compatibility) and predicted total lung capacity (pTLC) based on height and age. Other donor (D) and recipient (R) characteristics such as cytomegalovirus (CMV) serology (±), gender [male (M)/ female (F)] and age are often ignored, but the impact of D/R mismatch for these variables and their combinations on outcome is less investigated. METHODS: The early and late outcomes in 461 lung recipients (149 single-lung and 312 double-lung) transplanted between July 1991 and December 2009 were explored, comparing different D/R combinations for gender (M/F), age (<20/21 to 45/≥45 years), CMV (±), blood group (identical/compatible) and pTLC (-9%Δ, +11%Δ). RESULTS: Overall 5-, 10-, 15- and 20-year survival rates were 69, 50, 37 and 37%, respectively, and were significantly better in females {HR [95% confidence interval (CI)]: 0.5 (0.3-0.9); P = 0.023} and worse in older recipients [HR (95% CI): 1.6 (1.2-2.2); P = 0.003]. On univariate analysis, survival was significantly worse in recipients with gender opposite to that of the donor (39% for mismatch vs 51% for match at 10 years; P = 0.04), but not for other D/R matching variables: age (P = 0.89), pTLC (P = 0.14), CMV (P = 0.15), blood group (P = 0.82) and their combinations. The best survival at 5 years was seen in female donor (DF)/female recipient (RF) (80%), the worst in DF/male recipient (RM) (47%), and intermediate in male donor (DM)/RF (72%) and DM/RM (63%); P = 0.0001. On multivariate analysis, D/R gender mismatch was found to be the sole negative predictive factor for survival with an 80% increased risk of mortality [HR (95% CI): 1.8 (1.1-2.8); P = 0.01]. CONCLUSIONS: In our patient cohort, survival after LTx was superior in female and younger recipients. D/R gender mismatch may be an important prognostic factor for long-term outcome. A gender combination of DF/RM should be avoided. The exact reasons for these differences remain speculative.