Inverse Gaussian absorption

[rosamartae]

Dear All,

I am relatively new to PK modeling and now I am trying to test the inverse gaussian absorption. My problem is that when I do multiple dose simulations, I notice an accumulation in the dosing compartment and as a result, the PK becomes non-linear in time (single dose AUCinf is not equal to multiple dose AUCtau). I am sure this is not a normal behavior, so my model must be wrong. Is this not how you implement the inverse gaussian absorption?

Thank you,
Marta

[CMT] @annotated
DEPOT : Dosing compartment (gut)
CENT : Central compartment
PERI : Peripheral compartment

[PARAM]
TVVc = 700000,
TVK12 = 0.1,
TVK21 = 0.2,
TVK10 = 0.03

TVCV = 0.8, // skewness
TVMAT = 4 // mean absorption time

[OMEGA] @Annotated
nVc : 0.03 : ETA on Vc
nK10 : 0.05 : ETA on K10

[SIGMA]
0

[MAIN]
double Vc = TVVc * exp(nVc);
double K12 = TVK12;
double K21 = TVK21;
double K10 = TVK10 * exp(nK10);
double CV = TVCV;
double MAT = TVMAT;

double TAD = self.tad();
double Inp = 0;

if(TAD == 0) {
Inp = 0;
} else {
Inp = sqrt(MAT /(2 * 3.14 * pow(CV, 2) * pow(TAD, 3))) * exp( -1 * (pow(TAD - MAT, 2)/(2 * pow(CV,2) * MAT * TAD)));
}

double Input = Inp;

[ODE]
dxdt_DEPOT = - (DEPOT * Input);
dxdt_CENT = (DEPOT * Input) - (CENT * K12 - PERI * K21) - (CENT * K10);
dxdt_PERI = (CENT * K12)-(PERI * K21);

[TABLE]
capture IPRED = CENT/Vc;

[CAPTURE]
Input

[curtisKJ]

Did you ever get this to work? Was looking to try this recently myself, I could review your code above if no solution found. Thanks.