Blood flow in exercising muscles by xenon clearance and by microsphere trapping.

The accuracy of muscle blood flow measurement by the 133Xe clearance method (QXe) was assessed against direct venous outflow (Qv) and microsphere trapping flow (Q mu) determinations in isolated perfused dog gastrocnemius both at rest and during graded stimulation [O2 consumption (VO2) up to 12 ml X 100 g-1 X min-1] and in the gastrocnemius, vastus lateralis, and triceps of intact dogs at rest and while running on a treadmill at varied speeds up to maximum VO2. In 29 measurements performed in 11 isolated muscles, Q mu was in good agreement with Qv at rest and at all stimulation levels (Q mu/Qv = 1.0; r = 0.98). 133Xe clearance yielded much lower blood flows than the venous outflow and the microsphere trapping methods. In 43 measurements in 11 muscles, the mean QXe/Qv ratio was 0.57 +/- 0.03 (SE), independent of blood flow. Similarly, in 65 measurements in 2 intact dogs, the mean QXe/Q mu ratio in all tested muscles was 0.49 +/- 0.02 (SE), independent of blood flow. These results show that the 133Xe clearance method considerably underestimates blood flow in dog muscles.