Spatial pulmonary flow distribution in rabbit isolated lungs is a poor representation of the situation in vivo.

1. Isolated lung preparations are established to investigate effects on pulmonary vascular tone and spatial pulmonary flow (Q (rel)) distribution. In the present study, we hypothesized that Q (rel) distribution in isolated lungs is only poorly correlated with the in vivo situation. 2. Fourteen rabbits were anaesthetized and mechanically ventilated with room air. Animals were held in an upright position for 15 min and Q (rel) was assessed using fluorescent microspheres (Q (rel-in vivo)). A second injection of microspheres was made after isolation of the lungs (Q (rel-ex vivo)). Lungs were dried, cut into 1 cm(3) cubes and spatial Q (rel) distributions were analysed. 3. The mean correlation of Q (rel-in vivo) and Q (rel-ex vivo) was 0.592 +/- 0.188 (95% confidence interval 0.493-0.690). The Q (rel) was redistributed to more ventral (the mean slope of Q (rel) vs the dorsal-ventral axis changed from -0.289 +/- 0.227 to -0.147 +/- 0.114; P = 0.03), cranial (mean slope of Q (rel) vs the caudal-cranial axis changed from -0.386 +/- 0.193 to -0.176 +/- 0.142; P < 0.001) and central (mean slope of Q (rel) vs the hilus-peripheral axis changed from 0.436 +/- 0.133 to -0.236 +/- 0.159; P = 0.003) lung areas. 4. The results obtained from studies investigating Q (rel) distributions in isolated lung models must be interpreted cautiously because the isolated lung set-up significantly affects the spatial distribution of pulmonary flow.