Complete responder in mouse xenograft model

Phoenix Certara Forums Modelling and Simulation
1

Hello all modelers,

I am doing a mouse xenograft model as below.
The model itself ran well, however, I just wonder how you guys handle the complete responder where the tumor measurement is “0” after treatment for long time and not grow back.
In my current model, I just assuming the “0” as missing values after the treatment and only use those mouse does have regrow tumor volume to handle the tumor regrow estimation.

Is there any way that I can assume or incorporate the “real percentage of complete responder” in the model??

Thanks,
Shu-Pei

deriv(T = Kng * T (1-T/Tss)- K4T)
deriv(K1 = ((kmax* C/(KC50+C))-K1)/tau)
deriv(K2 = (K1 - K2)/tau)
deriv(K3 = (K2 - K3)/tau)
deriv(K4 = (K3 - K4)/tau)

2

I think the idea would be to use mixture models where one of the model is the non respondent one.

The LL statement would then incorporating the 2 models.

LL(iObs,log(p1*N(model1_mean,sigma1)+()1-p1)*N(model2_mean,sigma2)))

N is the normal distribution.

The program should estimate p1 which will be the estimated proportion of patients belonging to the first model(non respondent for example).

Best would be to get data and template project. Then I can help you.

best

Serge