Mannitol penetration of liver cells in estimations of extracellular fluid volume.

It has been suggested that since mannitol is excreted in the bile it may enter the liver cell and hence over-estimate extracellular fluid volume if used as a marker. Boluses of D-[14C]mannitol were injected into anaesthetized, nephrectomized dogs. After equilibration, liver biopsy samples were taken. Total liver water was estimated gravimetrically and compared with extracellular fluid volume estimated from the biopsy samples using beta-counting. There was no significant difference between these two estimations. However D-[14C]mannitol and [14C]inulin carboxylic acid spaces for liver were significantly different. This suggests that mannitol penetrates liver cells. A correction factor is derived to enable the technique to be used more accurately in 'whole-body' extracellular fluid volume estimation.