TABLE 4.

Tacrolimus population pharmacokinetic parameter estimates and bootstrap results for the final model ^a .

Parameter	Units	Final model parameter estimate (RSE%)	Median (95% CI) bootstrap result**
Disposition parameters
CL/F $=\left(\mathrm{\theta }1·\mathrm{\theta }2-\mathrm{\theta }3·\text{AGE}·\text{agecat}\right)·{\left(BW/70\right)}^{0.75}$
θ1	L/h/70 kg	26.5 (7.8%)	26.4 (22.7–30.7)
θ2	—	0.666 (8.1%)	0.663 (0.579–0.759)
θ3	L/h/70 kg/y	0.0562(26.0%)	0.0560 (0.0266–0.0863)
Vc/F	L/70 kg	327(14.0%))	326 (259–479)
Vp/F	L	298 (17.4%)	304 (164–468)
CLD/F	L/h/70 kg	51.9 (27.0%)	53.5 (29.9–98.5)
Absorption parameters
Ka	h−1	2.0 FIX	—
Lag-time	h	0.341 (9.4%)	0.346 (0.255–0.413)
Between-patient variability
ω2 CL	%	32.1(15.1%)	31.5 (26.9–37.2)
ω2 Vc	%	48.2(75.4%)	46.4 (19.2–99.9)
Residual variability	%	28.2(3.8%)	27.7 (24.8–30.1)

^a CL/F, apparent blood elimination clearance; CL_D/F, apparent inter-compartmental clearance between central and peripheral compartments; Vc/F and Vp/F, apparent distribution volumes of the central and peripheral compartments; Ka, first-order absorption rate constant; F, bioavailability; σ2, proportional residual variability expressed as the coefficient of variation; ω2, variance of between patient variabilities associated with the PK parameters, expressed as the coefficient of variation. θ2, change of CL/F values in CYP3A5*1 non-carriers with respect to CL/F values in CYP3A5*1 carriers; BW, bodyweight in kg; AGE, age in years; agecat, categorical variable that equals 0 for patients 60 years old and younger and equals 1 for patients older than 60 years; CL/F, CL_D/F, and Vc/F were allometrically scaled to 70 kg of bodyweight; allometric exponents were fixed to 0.75 for flow pharmacokinetic parameters and to 1 for central compartment distribution volume as follows, CL/F $=\left(\mathrm{\theta }1·\mathrm{\theta }2-\mathrm{\theta }3·\text{AGE}·\text{agecat}\right)·{\left(BW/70\right)}^{0.75}$ ; CL_D/F $=\mathrm{\theta }5·{\left(BW/70\right)}^{0.75}$ ; Vc/F $=\mathrm{\theta }4·\left(BW/70\right)$ .

**Estimated from 500 re-samplings with the non-parametric bootstrap method.