A significant limitation of standard bronchoalveolar lavage (BAL) technique is the inability to measure or calculate epithelial lining fluid (ELF) volume and, therefore, in vivo concentrations of substances in the ELF. We evaluated a new rewash BAL procedure with the radiolabeled tracer technetium pertechnetate ((TcO4-)-Tc-99m) that theoretically should be immune to even exaggerated fluid shifts during BAL. To test this theory, we measured ELF volume in control sheep using isosmotic (280 mosm/L) hypoosmotic (140 mosm/L) and hyperosmotic (570 mosm/L) BAL solutions to induce exaggerated fluid shifts during the lavage procedure. The mean ELF volume of the lavaged lung segment was not significantly different for the three solutions (isosmotic, 1.7 +/- 0.8 ml; hypoosmotic, 1.1 +/- 1.2 ml; hyperosmotic, 2.1 +/- 1.6 ml). The slope of the (TcO4-)-Tc-99m disappearance curve, however, was significantly steeper for the hyperosmotic solution (-0.40 +/- 0.04%/min) compared with the other solutions (isosmotic, -0.14 +/- .01%/min; hypoosmotic, -0.12 +/- 0.07%/min). Calculation of ELF volume using sodium as an endogenous tracer gave consistently smaller values with each of the mannitol solutions (isosmotic, 0.21 +/- 0.30 ml; hypoosmotic, 0.02 +/- 0.03 ml; hyperosmotic, 0.18 +/- 0.18 ml). The failure of sodium to provide accurate estimates of the ELF volume may be due to complicated sodium movement in the lung and errors in our assumption of the initial concentration of sodium in the ELF fluid. We conclude that the rewash BAL technique with (TcO4-)-Tc-99m gives values of ELF volume that are not significantly affected by even exaggeration of the fluid flux that invariably accompanies BAL.
