Poor baseline oxygenation does not preclude utilization of donor lungs for transplantation

Y. Bai¹, K. Farahnak¹, B. Heiden¹, A. Delhi¹, Y. Yan¹, S. Chang¹, C. Hamilton¹, R. R. Hachem¹, C. A. Witt¹, R. Vazquez Guillamet¹, D. Byers¹, G. F. Marklin², M. Hartwig³, B. Goldstein³, D. Kreisel¹, R. G.. Nava¹, B. Meyers¹, B. D. Kozower⁴, V. Puri^5
1Washington University in St. Louis, St. Louis, Missouri ²Mid-America Transplant, St Louis, Missouri ³Duke University Medical Center, Durham, North Carolina ⁴Washington University School of Medicine, St. Louis, Missouri ⁵Washington University School of Medicine, Saint Louis, Missouri

Purpose: A chronic, persistent shortage of suitable donor lungs is the rate-limiting factor in lung transplantation (LT). Although previous work has shown that donor PaO2 prior to procurement is highly predictive of lung utilization, we aimed to better understand the relationship between the initial PaO2 and lung utilization.

Methods: With collaborative agreements, we used a uniquely-compiled, prospectively-maintained, multi-institutional database from three US organ procurement organizations to abstract demographic, clinical, and radiologic data on all brain-dead donors managed between 1/2014 and 6/2020. Initial PaO2 was defined as the first value measured on FiO2 of 1.0 after declaration of brain death. Donors were dichotomized into Low (initial PaO2 < 300 mmHg) and High (initial PaO2≥300 mmHg) groups. Factors associated with lung utilization rates (LUR, number of lung donors/all donors) and improvement in PaO2 were studied in multivariable models. These models were adjusted for variables such as age, cause of death, body mass index, smoking history, past medical and surgical history, hepatitis C status, creatinine, presence of infection and positive cultures, vasopressor and inotropic drug use, blood transfusions, number of bronchoscopies and bronchoscopic abnormalities, number of non-lung organs harvested, days of mechanical ventilation, and positive findings on chest X-ray and computed tomography scans.

Results: Among 4854 donors, 3702 (76.3%) were in the Low group and 1152 (23.7%) in the High group. The overall LUR was 31.4% (1522/4854): Low group, 28.9% versus High group, 39.1% (p < 0.001).

In the Low group, 1380/3702 (37.3%) donors had an improvement in final PaO2 to ≥300 mmHg, with a LUR of 65.9% (OR 17.59, 95% CI [3.89-22.44], compared to those without improvement in PaO2). In this group, management in a specialized donor care facility (SDCF, OR 1.42 [1.01-1.98]), use of inotropes (OR 2.51 [1.91-3.31]) and antibiotics (OR 1.45 [1.08-1.95]), and more frequent bronchoscopy (OR 1.51 [1.36-1.68]) were associated with higher lung utilization. Massive transfusion (>10 units) was associated with lower lung utilization (OR 0.53 [0.31-0.92]).

In the High group, 615/1152 (53.4%) donors had a decline in final PaO2 to < 300 mmHg, with a LUR of 10.9% (OR 0.06 [0.04-0.09], compared to those who remained above 300 mmHg). In this group, longer duration of mechanical ventilation was associated with lower lung utilization (OR 0.90 [0.84-0.96]).

Additionally, 2861/4854 (58.9%) donors had very low initial PaO2 ( < 200 mmHg), yet a LUR of 26.6%. In this group, factors associated with lung utilization were similar to those in the Low initial group (PaO2 < 300 mmHg).

Conclusion: With appropriate management, low initial PaO2 does not preclude donor lung utilization for LT. Conversely, a significant proportion of donor lungs with high initial PaO2 deteriorate and are not transplanted. Aggressive pulmonary toilet, use of a SDCF model, and minimizing duration of mechanical ventilation may improve LUR in brain-dead donors.

Identify the source of the funding for this research project: 1R01HL146856-01A1