Earlier dosimetry estimates for [O-15]water assumed its instantaneous
equilibrium with total body water. This assumption leads to an underes
timation of the absorbed doses to organs with high blood flows, since
the biodistribution of this short-lived radiopharmaceutical is depende
nt upon blood flow to organs. We have developed a physiologically base
d whole body blood flow model (WBBFM) using a commercially available i
con-driven mathematical simulation software package and applied it to
the reevaluation of [O-15]water dosimetry in humans. The WBBFM uses mu
ltiple parallel compartments to represent organs, heart chambers, the
injection site for [O-15]water, and blood sampling sites (arterial and
venous). Input values to the WBBFM include organ blood flows, organ m
asses, organ water volumes, organ:blood partition coefficients, inject
ed activity and S-values of [O-15]. The WBBFM is based on the same ass
umptions that are used in calculating regional blood flow using [O-15]
water and simulates the human body closely in its physiologic response
. The activity in each organ is derived from the simulation and is use
d to calculate absorbed doses. The WBBFM calculated absorbed doses in
mu Gy/MBq (mrad/mCi) to various organs are as follows: heart-2.66 (9.8
4), kidneys-2.20 (8.15), thyroid-1.83 (6.78), brain-1.66 (6.13), ovari
es-1.25 (4.61), breast-1.24 (4.59), and small intestine-1.03 (3.83). T
hese values are approximately two- to threefold higher than the earlie
r estimates of Kearfott [J. Nucl. Med. 23, 1031-1037 (1982)] and simil
ar to the recent findings of Herscovitch et al. [J. Nucl. Med. 34, 155
P (1983)]. We believe this approach yields more realistic dosimetry es
timates for [O-15]water. Accordingly, we have revised the amount of [O
-15]water administered during regional blood flow studies at our insti
tution. The relative ease and accuracy of this approach suggests its u
sefulness in dosimetry estimation for other freely diffusible radiopha
rmaceuticals. (C) 1996 American Association of Physicists in Medicine.