Multiple (not repetitive) dosing

[PascaleRietveld]

---

Hi Bill,

Thank you for your useful tutorial! I am trying to calculate the AUC of my population but because of multiple dosing (not repetitive) I cannot make it seem to work.
I have a dataset where each patient receives per cycle (2 cycles per ID) an IV bolus and then a continuous infusion of 22 hours or 44 hours. Then some patients receive a second bolus and continuous infusion after about 22 hours. I use time after first dose (TAFD) because of multiple dosing.
My dataset kindoff looks like this:

Data <- data.frame(
  cycle = c(1, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1),
  ID = c(1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2),
  Conc = c(0, 0, 2.8, 2.5, 0, 0, 0, 0, 3, 3.2, 0, 0, 5),
  TAFD = c(0, 0.4, 5, 22, 0, 0.5, 0, 0.5, 10, 22, 23, 23.5, 45),
  duration = c(0.25, 22, NA, NA, 0.25, 22, 0.25, 22, NA, NA, 0.25, 22, NA),
  Dose = c(700, 1000, NA, NA, 500, 1100, 700, 1300, NA, NA, 700, 1100, NA))
 
data_conc <- Data %>% filter(is.na(Dose)) %>% select(-Dose,-duration)
 
knitr::kable(Data)
 
 
 
my.conc <- PKNCAconc(data_conc, Conc~TAFD|ID)
 
Dosdata <- Data %>% select(ID, Dose, TAFD, duration) %>% filter(!is.na(Dose))
my.dose <- PKNCAdose(Dosdata, Dose~TAFD|ID)

I am stuck because of this error: Rows that are not unique per group and time (column names: TAFD, ID) found within dosing data.

How can I calculate the PK parameters for this multiple dosing dataset?
Thanks a lot in advance!

Best,
Pascale

[billdenney]

@PascaleRietveld, Thanks for using PKNCA. I've moved this to a discussion since it's a question about use rather than a problem with PKNCA.

Here is your original data/code as a reprex showing the specific error and with some minor modifications to highlight the issue. From your code, I added the cycle column to Dosdata, and I added a line to show the Dosdata at the bottom:

library(tidyverse)
#> Warning: package 'ggplot2' was built under R version 4.3.1
#> Warning: package 'purrr' was built under R version 4.3.1
library(PKNCA)
#> 
#> Attaching package: 'PKNCA'
#> The following object is masked from 'package:stats':
#> 
#>     filter

Data <- data.frame(
  cycle = c(1, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1),
  ID = c(1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2),
  Conc = c(0, 0, 2.8, 2.5, 0, 0, 0, 0, 3, 3.2, 0, 0, 5),
  TAFD = c(0, 0.4, 5, 22, 0, 0.5, 0, 0.5, 10, 22, 23, 23.5, 45),
  duration = c(0.25, 22, NA, NA, 0.25, 22, 0.25, 22, NA, NA, 0.25, 22, NA),
  Dose = c(700, 1000, NA, NA, 500, 1100, 700, 1300, NA, NA, 700, 1100, NA))

data_conc <- Data %>% filter(is.na(Dose)) %>% select(-Dose,-duration)

my.conc <- PKNCAconc(data_conc, Conc~TAFD|ID)

Dosdata <- Data %>% select(ID, Dose, TAFD, duration, cycle) %>% filter(!is.na(Dose))
my.dose <- PKNCAdose(Dosdata, Dose~TAFD|ID)
#> Error in PKNCAdose.data.frame(Dosdata, Dose ~ TAFD | ID): Rows that are not unique per group and time (column names: TAFD, ID) found within dosing data.  Row numbers: 3

Dosdata
#>   ID Dose TAFD duration cycle
#> 1  1  700  0.0     0.25     1
#> 2  1 1000  0.4    22.00     1
#> 3  1  500  0.0     0.25     2
#> 4  1 1100  0.5    22.00     2
#> 5  2  700  0.0     0.25     1
#> 6  2 1300  0.5    22.00     1
#> 7  2  700 23.0     0.25     1
#> 8  2 1100 23.5    22.00     1

^{Created on 2023-09-25 with reprex v2.0.2}

The error is saying that row 3 of your data is duplicated for the dosing data. The reason it's duplicated is because TAFD and ID are the same on that row compared to another (in this case, the duplicated row is row 1).

The way to clarify the difference between rows 1 and 3 in your data is to include cycle in the grouping (see cycle+ID for both PKNCAconc() and PKNCAdose() below):

library(tidyverse)
#> Warning: package 'ggplot2' was built under R version 4.3.1
#> Warning: package 'purrr' was built under R version 4.3.1
library(PKNCA)
#> 
#> Attaching package: 'PKNCA'
#> The following object is masked from 'package:stats':
#> 
#>     filter

Data <- data.frame(
  cycle = c(1, 1, 1, 1, 2, 2, 1, 1, 1, 1, 1, 1, 1),
  ID = c(1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2),
  Conc = c(0, 0, 2.8, 2.5, 0, 0, 0, 0, 3, 3.2, 0, 0, 5),
  TAFD = c(0, 0.4, 5, 22, 0, 0.5, 0, 0.5, 10, 22, 23, 23.5, 45),
  duration = c(0.25, 22, NA, NA, 0.25, 22, 0.25, 22, NA, NA, 0.25, 22, NA),
  Dose = c(700, 1000, NA, NA, 500, 1100, 700, 1300, NA, NA, 700, 1100, NA))

data_conc <- Data %>% filter(is.na(Dose)) %>% select(-Dose,-duration)

my.conc <- PKNCAconc(data_conc, Conc~TAFD|cycle+ID)

Dosdata <- Data %>% select(ID, Dose, TAFD, duration, cycle) %>% filter(!is.na(Dose))
my.dose <- PKNCAdose(Dosdata, Dose~TAFD|cycle+ID)
#> Found column named duration, using it for the attribute of the same name.

^{Created on 2023-09-25 with reprex v2.0.2}

Another option would be to make the TAFD column represent the time since the first dose in the first cycle; then there would be no duplication.

[PascaleRietveld]

Thank you Bill! I am clearly new to this:)
Later down the code I get the message to report a bug.
The PKNCAdata function also gives me trouble that I don't understand.
What I see is this:

use the default settings for AUC ranges, whether to get cmax, tmax, kel.pred, etc.)
my.data.auto <- PKNCAdata(my.conc,my.dose)
Error in vectbl_assign(x[[j]], i, recycled_value[[j]]) :
DLL requires the use of native symbols

PKNCA.options("single.dose.aucs")
my.data.auto$intervals

manually set what items to calculate and what intervals
my.intervals <- data.frame(start = 0,
end=50,
cmax=T,
tmax=T,
aucinf=F,
auclast=T,
kel.obs=T,
cl.obs=T)

my.data.manual <- PKNCAdata(my.conc,my.dose, intervals=my.intervals)
my.data.manual$intervals

=============== Results ======================
auto

my.results.auto <- pk.nca(my.data.auto)
Error in purrr::pmap():
ℹ In index: 1.
Caused by error in vectbl_assign():
! Please report a bug.
: Error with interval start=0.416666666666667, end=Inf: DLL requires the use of native symbols

Should I change column names or types to use native symbols?

Best,
Pascale

[billdenney]

The fact that the error is with vectbl_assign suggests that the underlying issue is with the vctrs library (see tidyverse/dplyr#6793).

Can you please provide the output of sessionInfo()?

[PascaleRietveld]

Hi Bill,

Somehow the error went away (I think because I work on a different computer today).
Everything worked well in the end, using the data imputation method start_conc0.
Thank you!

[billdenney]

Thanks for following up. I think that the underlying issue you were seeing the second time was with the vctrs library and that's why a different computer (possibly with a different version of some other libraries) worked.