Abstract
543
Objectives The overestimation of plasma clearance (CL) values caused by increased extracellular fluid volume (ECV) is undesirable for carboplatin dosimetry. To explore this, we examined the change in CL values for a wide range of observed versus age related expected ECV for three CL fit models—adaptively Tikhonov regularized gamma variate (Tk-GV)—numerical integration (NI), and—biexponential (E2).
Methods 412 Tc-99m-DTPA studies of 0.6- to 56- y/o’s with suspected fluid disturbance were processed using each method to construct ratios of CL-values from 4 plasma samples over 4-h to their 9 samples over 8-h controls. Next, relative ECV (rECV) for each method was calculated as ECV at 8-h expressed as a percent of body weight (ECV%) by dividing by the expectation that ECV% = 26.93-2.979[asterisk]ln[A (y/o)] from prior thiosulfate normal data. Intravenous thiosulfate levels may underestimate ECV by 5% or 10%. Regressions were used to examine how the ratios of early to late time CL-values varied with rECV volumes.
Results From each method, the rECV expected values were from 1.09 to 1.14 times larger than the age adjusted ECV estimates. For Tk-GV, the ECV% volumes ranged from 12.9% to 67.0% of body weight, and the corresponding rECV-values ranged from 0.63 to 4.48 times expected. Over the full range of rECVs from each method’s ECV%, the regressed 4-h/8-h CL-ratios varied least for Tk-GV, 0.997 to 1.039, and more for both NI (0.966 to 1.115) and E2 (0.984 to 1.094). Both the NI and E2 trend lines had overlapping 95% confidence intervals for the full rECV range. However, at rECV > 0.97, relative CL(Tk-GV) increased significantly less than for NI and E2.
Conclusions Despite the derivative nature of the rECV estimates, the Tk-GV method had a significantly lesser inflation of 4-h/8-h CL values for normal and increased rECV than the NI and E2 methods, suggesting that use of the Tk-GV method would reduce the impact of increased ECV on 4-h CL-values for better dosimetry.