Time-varying input/flexibility to change input rate on the fly

Dear all, I was wondering if any progress has been made on the topic raised originally by Bill Denney in 2018: https://www.mail-archive.com/[email protected]/msg06990.html Are there any simpler ways in NM 7.5 to adapt input (e.g. infusion rates) in $DES during the integration step without adapting the dataset itself? I.e. to model the malfunctioning of an infusion pump (at random), the loss of part of a tablet, or the detachment of a patch? Thank you! I could not answer to the original topic which is why I just linked to it. -- Dr. ir. Robin Michelet Senior scientist Freie Universitaet Berlin Institute of Pharmacy Dept. of Clinical Pharmacy & Biochemistry Kelchstr. 31 12169 Berlin Germany Phone: + 49 30 838 50659 Fax: + 49 30 838 4 50656 Email: [email protected] www.clinical-pharmacy.eu https://fair-flagellin.eu/

Hi Robin, I don't think that I've seen an update. That said, the need I had then was for a very specific need for an unusual drug. I've only seen this type of issue once where it seemed to need time-dependent effects. Generally, effects similar-- but not identical-- to what I was experiencing at the time are better-modeled with simpler systems. For example, adsorption to infusion sets can almost always be modeled as a decrease in bioavailability and/or a lag time (it's not typically time-dependent behavior). I would assume that loss of part of a tablet or detachment of a patch could be simply modeled as random variability (or a fixed effect) on bioavailability. Random pump malfunction would depend on how it malfunctioned, but I would be wary of trying to model random effects as this more complex time-dependent bioavailability unless you had data on the malfunction method-- in which case I would suggest putting it into the dataset as a different dosing record. Thanks, Bill

one can do it by hands, like set F1=1 and then use DA1/dt = -KA*A(1) DA2/dt = FF1(any function of time)*A(1) .. will it do the trick? Leonid

Hi Robin, I have used the DECLARE option in the $ABBREVIATED block to initialize DOSE as a global variable. Then I used MTIME()/MPAST() to flexibly deliver dose into a DEPOT compartment from a patch, assuming different infusion durations after detachment and reattachment of the patch. See the code below. Best Eliford ``` $PROBLEM One compartment model with erratic absorption from a patch $INPUT ID TIME DV AMT CMT EVID MDV $DATA ../data/erratic_patch.csv IGNORE=@ $SUBROUTINE ADVAN6 TRANS1 TOL=6 $MODEL COMP = (DEPOT) COMP = (CENTRAL) COMP = (ADMINDOSE) ;; Accumulate total dose delivered to depot compartment $ABBR DECLARE DOSE(10) ;; Declare DOSE as a global variable with 10 dimensions as place holder $THETA (0, 0.3) ; KA (0, 3) ; CL (0, 15) ; Vc (0, 3) ; DT1 ;; Detach and re-attach time can be estimated (0, 6) ; DT2 ;; (0, 15) ; DD1 (0, 12) ; DD2 ;; Variable durations of zero order delivery to depot compartment can be estimated (0, 4) ; DD3 (0, 1) ; ADD (0, 0.1) ; PROP $OMEGA BLOCK(1) 0.1 ; EKA $OMEGA BLOCK(1) 0.1 ; ECL $OMEGA BLOCK(1) 0.1 ; EVC $OMEGA BLOCK(1) 0.1 ; EDT1 $OMEGA BLOCK(1) 0.1 ; EDT2 $OMEGA BLOCK(1) 0.1 ; EDD1 $OMEGA BLOCK(1) 0.1 ; EDD2 $OMEGA BLOCK(1) 0.1 ; EDD3 $SIGMA 1 FIX $PK TVKA = THETA(1) ; Absorption rate constant from depot TVCL = THETA(2) ; Clearance from central cmpt TVVC = THETA(3) ; Distribution volume of central cmpt TVDT1 = THETA(4) ; Detach time 1 TVDT2 = THETA(5) ; Detach time 2 TVDUR1 = THETA(6) ; Delivery duration 1 TVDUR2 = THETA(7) ; Delivery duration 2 TVDUR3 = THETA(8) ; Delivery duration 2 KA = TVKA * EXP(ETA(1)) ; CL = TVCL * EXP(ETA(2)) ; VC = TVVC * EXP(ETA(3)) ; DT1= TVDT1 * EXP(ETA(4)) ; DT2= TVDT2 * EXP(ETA(5)) ; DUR1= TVDUR1 * EXP(ETA(6)) ; DUR2= TVDUR2 * EXP(ETA(7)) ; DUR3= TVDUR3 * EXP(ETA(8)) ; K = CL/VC ; F1 = 0 ; Prevent NONMEM from adding dose into depot ADMDOSE = A(3) ; Total Dose delivered to depot compartment at any time after dose MTIME(1) = DT1 ; Model event time for detachment MTIME(2) = DT2 ; MOdel event time for reattachment IF(EVID.EQ.1) THEN DOSE(1) = AMT ; Give value to the global dose variable created with $ABBR DECLARE ENDIF $DES DOSE1 = DOSE(1) ; Make the dose global RATE1 = DOSE1/DUR1 ; Initial rate of drug delivery into depot cmpt IF(T.GE.DT1) THEN DOSE2 = DOSE(2) ; Remaining dose in the patch just before detachment RATE2 = DOSE2/DUR2 ; rate of drug delivery into depot cmpt after detach 1 ENDIF IF(T.GE.DT2) THEN DOSE3 = DOSE(3) ; Remaining dose in the patch just before re-attachment RATE3 = DOSE3/DUR3 ; rate of drug delivery into depot after reattachment ENDIF RATE = RATE1 * (1 - MPAST(1)) ; FLEXIBLE RATE BASED ON MODEL EVENT TIMES RATE = RATE + RATE2 * (MPAST(1) - MPAST(2)) ; RATE = RATE + RATE3 * MPAST(2) ; IF(ADMDOSE.GT.DOSE1) RATE= 0 ; Switch off delivery to depot when all dose has been consumed ADMDOSE1 = RATE1*DT1 ; Calculate remaining dose in the patch ADMDOSE2 = RATE2*(DT2-DT1) ; DOSE(2) = DOSE1 - ADMDOSE1 ; DOSE(3) = DOSE1 - (ADMDOSE1+ADMDOSE2) ; DADT(1) = RATE - KA*A(1) ; Differential equations DADT(2) = KA*A(1) - K*A(2) ; DADT(3) = RATE ; $ERROR IPRED = F ; CENTRAL is default observation compartment ADD = THETA(9) ; Additive residual PROP = IPRED * THETA(10) ; Prop residual W = SQRT(ADD**2 + PROP**2) ; Weighting factor Y = IPRED + W*EPS(1) ; IRES = DV-IPRED IWRES = IRES/W DEPOTA = A(1) NREPLICATE = NREP IREPLICATE = IPRED $SIMULATION (82536690) SUBPROBLEM=1 ONLYSIMULATION $TABLE ID TIME DV IPRED EVID KA CL VC DT1 DT2 DUR1 DUR2 DUR3 RATE1 RATE2 RATE3 RATE DOSE1 DOSE2 DOSE3 DEPOTA ADMDOSE NREPLICATE IREPLICATE NOPRINT ONEHEADER APPEND FILE=MYTAB1 ``` Simulation dataset patchdata <- expand.grid(ID=1, TIME=seq(0, 48, 1), DV=".") %>% mutate(AMT=ifelse(TIME==0, 10, NA), CMT= ifelse(TIME==0, 1, 2), EVID= ifelse(TIME==0, 1, 2), MDV=1)