In Michorlab/ACESO: A Cancer Evolution Simulation Optimizer

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  warnings= FALSE
)

We are going to simulate complex dosing schedules using mrgsolve and ACESO packages.

#' Load libraries
library(ACESO)
library(mrgsolve)
#install mrgsolve from github: devtools::install_github("metrumresearchgroup/mrgsolve")

Pharmacokinetic model

We are going to simulate a one compartment model with intravenous administration.

Read the model from the model libraries included in ACESO:

model_library(list=T)

cmt1_iv <- mread("1cmt_iv", model_library()) %>% Req(CP)
#See the code of the model
see(cmt1_iv)

See the parameters of the model with the default values:

param(cmt1_iv)

We are going to change the default parameter values for TVCL (typical value for the CLearance) and TVV (Typical value for the Volume of distribution):

newpar <-  list('TVCL' = 171.4, #L/h
          'TVV'  = 153.5) 
cmt1_iv<-param(cmt1_iv,newpar)

param(cmt1_iv)

Event objects in mrgsolve

First dosing schedule: 150mg dose at time 0, once every day in a week

e1 <- ev(amt = 150, ii = 1, addl = 6, time=0)
e1

• amt: amount of administered dose
• ii: time interval between doses
• addl: additional doses
• time: time of first dose
• cmt: compartment number (in which compartment a dosing is ocurring)
• evid: type of record (1: defines the record as a dose event)

Simulate

Simulate the model with easy.mrgsim function (extended version of mrgsim function in mrgsolve package)

• model: model name
• dosing_schedule: event object where the dosing schedule is specified
• tend: final time
• delta: the increment of time to simulate a sequence from 0 to tend.

easy.mrgsim(model=cmt1_iv,dosing_schedule=e1,delta=0.1,tend=7) %>% plot

1 week of holiday between each treatment period:

e2 <- seq(e1, wait = 7, e1, wait = 7)
e2
easy.mrgsim(model=cmt1_iv,dosing_schedule=e2,delta=0.1,tend=7*4) %>% plot

A alternating dosing sequence: 150 mg daily for a week, then 50 mg daily for a week etc.

e3 <- ev(amt = 50, ii = 1, addl = 6)
e4 <- seq(e1, e3, e1)
e4
easy.mrgsim(model=cmt1_iv,dosing_schedule=e4,delta=0.1,tend=7*3) %>% plot

A cycle of 3 weeks on, one week off for two months

e5 <- ev(amt = 150, ii = 1, addl = (3*7)-1)
e6 <- seq(e5, wait = 7, e5)
easy.mrgsim(model=cmt1_iv,dosing_schedule=e6,delta=0.1,tend=7*4*2) %>% plot

Load + maintenance doses

#High dose: 1600mg once every week
load<- ev(amt = 1600, ii = 7, addl = 0, time=0)
#Low dose: 50mg once every day, after the high dose
maintenance<- ev(amt = 50, ii = 1, addl = 5,time=1)
easy.mrgsim(model=cmt1_iv,dosing_schedule=load+maintenance,delta=0.1,tend=7) %>% plot

Repeat this for one month period

load_maint <- ev_rep(load+maintenance, n=4)
head(load_maint)
easy.mrgsim(model=cmt1_iv,data=load_maint,delta=0.1,tend=7*4) %>% plot

Simulate several individuals

To simulate different individuals we need to introduce variability into the pk parameters. To this end, a variance-covariance matrix needs to be defined. Let's simulate 5 individuals, with 0.1 variance for the clearance (CL) and 0.2 variance for the volume of distribution (V).

Nindividuals=5
easy.mrgsim(model=cmt1_iv,dosing_schedule=load+maintenance,delta=0.1,tend=7,
            variability = dmat(0.1, 0.2),nid=Nindividuals) %>% plot

Plot with ggplot2

You prefer plotting with ggplot2? Just save the results as a data.frame

pk<-as.data.frame(easy.mrgsim(model=cmt1_iv,dosing_schedule=load+maintenance,delta=0.1,tend=7,
                              variability = dmat(0.1, 0.2),nid=Nindividuals) )
head(pk)

library(ggplot2)
ggplot(data=pk,aes(x=time,y=CP,col=as.factor(ID)))+geom_line(size=1.3)

IV infusion

For drug administration by infusion, or any zero-order input models, the rate of drug administration may be defined using the rate argument.

inf <- ev(amt = 150, time=0, rate=48)
easy.mrgsim(model=cmt1_iv,dosing_schedule =inf,delta=0.1,tend=7*2) %>% plot

Oral administration

Now we are going to simulate a one compartment model with extravascular administration. We first select a model from the model libraries already included in ACESO instead of writing all the code by ourselves:

model_library(list=T)

cmt1_oral <- mread("1cmt_ev", model_library()) %>% Req(CP) 

oral1 <- ev(amt = 150, time=0)

easy.mrgsim(model=cmt1_oral,dosing_schedule=oral1,delta=0.1,tend=7) %>% plot

oral2 <- ev(amt = 150, ii=1, addl=6, time=0)

easy.mrgsim(model=cmt1_oral,dosing_schedule=oral2,delta=0.1,tend=7*3) %>% plot

Introduce a lag time of 2 days

#See the parameters of the model
param(cmt1_oral)
#The parameter associated with the lag time is ALAG
easy.mrgsim(model=cmt1_oral,dosing_schedule=oral1,delta=0.1,tend=7*3,parameters=list(ALAG=2)) %>% plot

Michorlab/ACESO documentation built on June 4, 2021, 4:57 p.m.