Hi, I have been trying to model nonlinear kinetics behavior with enzyme induction and repeated oral dosing regimen. I was having a lot of trouble mapping the repeated dosing to the parameter “Dose” in closed form equations. I succeeded in using the dosepoint() statement with differential equation coding, so that I could map my repeated dosing under the dosing tab under “setup”. But how do you assign repeated dosing to the “Dose” parameter in closed form equations? I then thought about using the sequence() and sleep() statement for repeated dosing in closed form. But it did not work. I am guessing the program has trouble recognizing my covariate(Time) as the t required in sequence and sleep statement. I am including my codes below. Any ideas would be appreciated! test(){ double(D) double(SW1,SW2,SW3) covariate(Time) C = ((DDKorKa)/((Ka-Ke)V))(exp(-KeTime) - exp(-KaTime)) Ke = SW1Kpre + SW2Kperi + SW3*Kpost Kperi = Kss- (Kss-Kpre)exp(-kout(Time-216)) Kpost = Kpre - (Kpre - Ass)exp(-kout(Time-512)) Ass = Kss- (Kss-Kpre)exp(-kout(512-216)) Kpre = Clpre / V Kss = Clss / V DKor = 3796.94 sequence{ D = 10 while(Time < 697){ sleep(8) D = D + 10 } } sequence{ SW1 = 1 SW2 = 0 SW3 = 0 sleep(216) SW1 = 0 SW2 = 1 SW3 = 0 sleep(512-216) SW1 = 0 SW2 = 0 SW3 = 1 } error(CEps = 1) observe(CObs = C + CEps) stparm(V = tvV) stparm(Ke = tvKe) stparm(Ka = tvKa) stparm(kout = tvkout) stparm(Clpre = tvClpre) stparm(Clss = tvClss) fixef(tvV = c(, 1260, )) fixef(tvKe = c(, 3.5, )) fixef(tvKa = c(, 3, )) fixef(tvkout = c(, 0.005, )) fixef(tvClpre = c(, 40, )) fixef(tvClss = c(, 140, )) }
I am also working on nonlinear kinetics with enzyme induction, only difference is my compound is administered through repeated subcutaneous dosing. I therefore would be equally interested in hearing if anyone has some input to jdeng415 's problem. Best regards
To used a closed form expression with multiple dosing you have to find a way to repeat the expression for each dose and dose time. The means storing the dosing information in a vector in PML and looping over the dose events at each observation time (not possible in current PML). The built in functions “cfmicro” and “cfmacro” do this work for you, though. They can even handle time varying covariates, so you can add a column of flags to the data to indicate which CL to use for the various observations. The most general solution to nonlinear or time-invariant systems is, of course, to use differential equations. The sequence block provides a way to interrupt time-stepping in an ODE integrator routine and thus only works with differential equations.