Topotecan (Leger 2004) • nlmixr2lib

Model and source

mod_meta <- nlmixr2est::nlmixr(readModelDb("Leger_2004_topotecan"))$meta
#> ℹ parameter labels from comments will be replaced by 'label()'

Citation: Leger F, Loos WJ, Fourcade J, Bugat R, Goffinet M, Mathijssen RHJ, Verweij J, Sparreboom A, Chatelut E. Factors affecting pharmacokinetic variability of oral topotecan: a population analysis. Br J Cancer. 2004;90(2):343-347. doi:10.1038/sj.bjc.6601469
Description: Two-compartment population PK model for oral and intravenous topotecan in adult cancer patients, with first-order absorption + lag time for the oral route, additive linear creatinine-clearance plus linear-ordinal WHO performance-status effects on CL, and linear body-weight effect on the central volume of distribution (Leger 2004)
Article (DOI): https://doi.org/10.1038/sj.bjc.6601469

This vignette validates the packaged Leger_2004_topotecan model – a two-compartment population PK model for oral and intravenous topotecan in 190 adult cancer patients pooled across five clinical trials and a drug monitoring cohort – against the source publication’s Table 1 (baseline demographics), Table 3 (final-model parameter estimates and inter-individual variability), and the Results / Discussion narrative on inter-occasion variability, oral bioavailability, and the limited-sampling strategy for oral topotecan AUC.

Population

Leger 2004 pooled topotecan plasma concentration data from 190 adult cancer patients enrolled in five separate clinical trials (173 patients) plus 17 additional patients receiving therapeutic drug monitoring for other reasons. Patients received topotecan either as a 30-minute intravenous infusion (n = 72) at 0.2-2.4 mg/m^2/day, or as oral gelatin capsules (n = 118) at 0.15-2.7 mg/m^2/day, for 5-21 consecutive days. The cohort is European (Toulouse, Institut Claudius-Regaud and Rotterdam, Erasmus MC – Daniel den Hoed Cancer Center). Median age was 55 years (range 18-76), median body weight 70 kg (range 42-117), and median body surface area 1.80 m^2 (range 1.36-2.44, DuBois formula). Mean Cockcroft-Gault creatinine clearance was 80 mL/min (range 33-167) and mean serum creatinine was 87 umol/L (range 41-162). WHO performance status (PS) was 0, 1, 2, or 3 in 79, 98, 11, and 2 patients respectively. 141 of 190 patients had received previous chemotherapy, and cisplatin pre-treatment was a feature of 115 patients (prior regimen) and 30 patients (on Day 1 of topotecan). The total number of analyzable plasma samples was 2064 (2-21 samples per patient, median 15 in cycle 1).

The same information is available programmatically via the model’s population metadata:

str(mod_meta$population)
#> List of 17
#>  $ species           : chr "human"
#>  $ n_subjects        : int 190
#>  $ n_studies         : int 6
#>  $ age_range         : chr "18-76 years"
#>  $ age_median        : chr "55 years"
#>  $ weight_range      : chr "42-117 kg"
#>  $ weight_median     : chr "70 kg"
#>  $ bsa_range         : chr "1.36-2.44 m^2 (DuBois formula)"
#>  $ bsa_median        : chr "1.80 m^2"
#>  $ sex_female_pct    : num NA
#>  $ race_ethnicity    : chr "Not reported (French and Dutch oncology centres)"
#>  $ disease_state     : chr "Adult cancer patients (predominantly ovarian cancer and other solid tumours; some studies combined with cisplatin)"
#>  $ dose_range        : chr "Oral 0.15-2.7 mg/m^2/day or IV infusion over 30 min 0.2-2.4 mg/m^2/day for 5-21 consecutive days"
#>  $ regions           : chr "France (Toulouse, Institut Claudius-Regaud) and the Netherlands (Rotterdam, Erasmus MC - Daniel den Hoed Cancer Center)"
#>  $ renal_function    : chr "Cockcroft-Gault creatinine clearance mean 80 mL/min (range 33-167); serum creatinine mean 87 umol/L (range 41-162)"
#>  $ performance_status: chr "WHO PS distribution 0 / 1 / 2 / 3 = 79 / 98 / 11 / 2"
#>  $ notes             : chr "Baseline demographics per Leger 2004 Table 1. Pooled data from five separate clinical trials (173 patients) plu"| __truncated__

Source trace

The per-parameter origin is recorded as an in-file comment next to each ini() entry in inst/modeldb/specificDrugs/Leger_2004_topotecan.R. The table below collects them in one place; all values come from Leger 2004 Table 3 (final-covariate-model, cycle 1) unless otherwise noted.

Parameter / equation	Value	Source location
`lcl` (CL intercept; non-renal CL)	log(12.8)	Table 3, theta1 = 12.8 (95% CI 4.8)
`e_crcl_cl` (slope of CL on CrCl in L/h)	2.1	Table 3, theta2 = 2.1 (95% CI 1.0)
`e_who_ps_cl` (fractional CL reduction per WHO PS)	0.12	Table 3, theta3 = 0.12 (95% CI 0.09)
`lvc` (slope of V1 on body weight, L/kg)	log(0.58)	Table 3, theta4 = 0.58 (95% CI 0.13)
`lvp` (V2, L)	log(45.5)	Table 3, theta5 = 45.5 (95% CI 7.0)
`lq` (Q, L/h)	log(49.2)	Table 3, theta6 = 49.2 (95% CI 16.9)
`lfdepot` (oral bioavailability F)	log(0.324)	Table 3, theta7 = 32.4% (95% CI 3.9)
`lka` (first-order absorption rate, 1/h)	log(1.7)	Table 3, theta9 = 1.7 (95% CI 0.6)
`ltlag` (oral absorption lag time, h)	log(0.17)	Table 3, theta10 = 0.17 (95% CI 0.03)
`etalcl ~ 0.08075`	log(0.29^2 + 1)	Table 3, CL %CV = 29 (interindividual)
`etalvc ~ 0.14157`	log(0.39^2 + 1)	Table 3, V1 %CV = 39
`etalvp ~ 0.19193`	log(0.46^2 + 1)	Table 3, V2 %CV = 46
`etalq ~ 0.49470`	log(0.80^2 + 1)	Table 3, Q %CV = 80
`etalfdepot ~ 0.04727`	log(0.22^2 + 1)	Table 3, F %CV = 22
`etalka ~ 0.34335`	log(0.64^2 + 1)	Table 3, Ka %CV = 64
`etaltlag ~ 0.01676`	log(0.13^2 + 1)	Table 3, Lag %CV = 13
`propSd <- 0.17`	0.17	Results, oral-route proportional residual = 17%
`addSd <- 0.09`	0.09 ug/L	Results, oral-route additive residual = 0.09 ug/L
`cl <- (exp(lcl) + e_crcl_cl * crcl_Lh) * (1 - e_who_ps_cl * WHO_PS) * exp(etalcl)`	n/a	Table 3 covariate model row “CL = (theta1 + theta2 * CrCl) * (1 - theta3 * PS)”; CrCl converted from mL/min to L/h via 0.06 in model() per Table 3 footnote
`vc <- exp(lvc + etalvc) * WT`	n/a	Table 3 row “V1 = theta4 * body weight”
`f(depot) <- exp(lfdepot + etalfdepot)`	n/a	Results, oral bioavailability mean 32.4%, CV 22%
`alag(depot) <- exp(ltlag + etaltlag)`	n/a	Results / Table 3, lag time mean 0.17 h, CV 13%
`Cc ~ add(addSd) + prop(propSd)`	n/a	Results, combined error model (additive + proportional); oral-route values used (see Assumptions)

Virtual cohort

Original observed concentrations are not publicly available. The virtual cohorts below approximate the published demographics: body weight log-normal around the 70-kg median (38-125 kg range); Cockcroft-Gault creatinine clearance log-normal around the 80 mL/min mean (truncated to the 33-167 mL/min range reported in Table 1); WHO performance status drawn from the empirical 79 / 98 / 11 / 2 distribution. Two parallel cohorts are constructed – 60 oral subjects (1.5 mg/m^2 single dose) and 60 IV subjects (1.5 mg/m^2 over 30 min) – so the oral-versus-IV bioavailability check later in this vignette has comparable sample sizes.

set.seed(20260610)

draw_cohort <- function(n, id_offset = 0L) {
  wt   <- exp(rnorm(n, mean = log(70),  sd = log(125 / 38) / 4))
  wt   <- pmin(pmax(wt, 42), 117)
  bsa  <- 0.007184 * wt^0.425 * 170^0.725
  bsa  <- pmin(pmax(bsa, 1.36), 2.44)
  crcl <- exp(rnorm(n, mean = log(80),  sd = log(167 / 33) / 4))
  crcl <- pmin(pmax(crcl, 33), 167)
  ps   <- sample(0:3, n, replace = TRUE,
                 prob = c(79, 98, 11, 2) / 190)
  tibble::tibble(
    id     = id_offset + seq_len(n),
    WT     = wt,
    BSA    = bsa,
    CRCL   = crcl,
    WHO_PS = ps
  )
}

n_oral <- 60L
n_iv   <- 60L

cov_oral <- draw_cohort(n_oral, id_offset =  0L)  |> mutate(treatment = "oral")
cov_iv   <- draw_cohort(n_iv,   id_offset = 1000L) |> mutate(treatment = "iv")

# Dosing target: 1.5 mg/m^2 single dose, ~the i.v. recommended dose in von
# Pawel 2001 cited in the Discussion. Oral subjects receive an oral capsule
# (depot compartment). IV subjects receive a 30-minute infusion (central
# compartment) per the protocol.
dose_mg <- function(bsa) round(1.5 * bsa, 3)

sample_times <- c(0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 4, 6, 8, 12, 24)

make_subject <- function(row) {
  amt <- dose_mg(row$BSA)
  if (identical(row$treatment, "oral")) {
    dose_event <- tibble::tibble(
      id   = row$id,         time = 0,
      evid = 1L,             amt  = amt,
      rate = NA_real_,       cmt  = "depot"
    )
  } else {
    dose_event <- tibble::tibble(
      id   = row$id,         time = 0,
      evid = 1L,             amt  = amt,
      rate = amt / 0.5,      cmt  = "central"
    )
  }
  obs_events <- tibble::tibble(
    id   = row$id,           time = sample_times,
    evid = 0L,               amt  = NA_real_,
    rate = NA_real_,         cmt  = NA_character_
  )
  dplyr::bind_rows(dose_event, obs_events) |>
    dplyr::mutate(
      WT        = row$WT,
      CRCL      = row$CRCL,
      WHO_PS    = row$WHO_PS,
      treatment = row$treatment
    ) |>
    dplyr::arrange(time, dplyr::desc(evid))
}

build_events <- function(cov_tbl) {
  dplyr::bind_rows(lapply(seq_len(nrow(cov_tbl)), function(i) {
    make_subject(cov_tbl[i, ])
  }))
}

events_oral <- build_events(cov_oral)
events_iv   <- build_events(cov_iv)
events      <- dplyr::bind_rows(events_oral, events_iv)

stopifnot(!anyDuplicated(unique(events[, c("id", "time", "evid")])))

Simulation

mod         <- readModelDb("Leger_2004_topotecan")
mod_typical <- rxode2::zeroRe(mod)
#> ℹ parameter labels from comments will be replaced by 'label()'

sim_typical <- rxode2::rxSolve(
  object = mod_typical,
  events = events,
  keep   = c("WT", "CRCL", "WHO_PS", "treatment")
) |>
  as.data.frame()
#> ℹ omega/sigma items treated as zero: 'etalcl', 'etalvc', 'etalvp', 'etalq', 'etalfdepot', 'etalka', 'etaltlag'
#> Warning: multi-subject simulation without without 'omega'

sim_stoch <- rxode2::rxSolve(
  object = mod,
  events = events,
  keep   = c("WT", "CRCL", "WHO_PS", "treatment")
) |>
  as.data.frame()
#> ℹ parameter labels from comments will be replaced by 'label()'

Replicate published figures

Leger 2004 has two main figures: Figure 1 shows two representative oral subjects with limited (-9%) versus large (+104%) inter-day change in AUC; Figure 2 shows the limited-sampling strategy ratio plot. Both depend on individual observed data that is not in the public domain. The chunks below instead reproduce the published model’s typical-value oral and IV concentration-time profiles, which faithfully exercise the structural two-compartment model with first-order absorption and lag time.

Typical-value oral and IV concentration-time profiles

typ_summary <- sim_typical |>
  dplyr::filter(time > 0) |>
  dplyr::group_by(time, treatment) |>
  dplyr::summarise(
    Cc_median = median(Cc, na.rm = TRUE),
    .groups   = "drop"
  )

ggplot(typ_summary, aes(time, Cc_median, colour = treatment)) +
  geom_line(linewidth = 1) +
  scale_y_log10() +
  labs(
    x        = "Time (h)",
    y        = "Topotecan plasma concentration (ug/L, log scale)",
    colour   = "Route",
    title    = "Typical-value topotecan concentration-time profile after 1.5 mg/m^2",
    subtitle = "Two-compartment model with first-order absorption (oral) or 30-min infusion (IV)"
  ) +
  theme_minimal()

Stochastic VPC over the oral cohort

sim_stoch |>
  dplyr::filter(time > 0, treatment == "oral") |>
  dplyr::group_by(time) |>
  dplyr::summarise(
    Q05 = quantile(Cc, 0.05, na.rm = TRUE),
    Q50 = quantile(Cc, 0.50, na.rm = TRUE),
    Q95 = quantile(Cc, 0.95, na.rm = TRUE),
    .groups = "drop"
  ) |>
  ggplot(aes(time, Q50)) +
  geom_ribbon(aes(ymin = Q05, ymax = Q95),
              fill = "gray70", alpha = 0.5) +
  geom_line(linewidth = 0.9) +
  scale_y_log10() +
  labs(
    x        = "Time (h)",
    y        = "Topotecan plasma concentration (ug/L, log scale)",
    title    = "Stochastic VPC over the oral cohort (1.5 mg/m^2 single dose)",
    subtitle = "Median (line) and 90% prediction interval (ribbon)"
  ) +
  theme_minimal()

PKNCA validation

Leger 2004 reports an observed mean bioavailability of 32.4% (95% CI +/- 3.9) for oral topotecan and an inter-individual variability on dose-normalized AUC of 4.8-fold for IV data and 7.6-fold for oral data (Discussion, p. 346). The simulated NCA below compares oral-versus-IV AUC0-Inf to recover the bioavailability and AUC variability.

sim_for_nca <- sim_stoch |>
  dplyr::filter(!is.na(Cc), time > 0) |>
  dplyr::select(id, time, Cc, treatment) |>
  as.data.frame()

doses_for_nca <- events |>
  dplyr::filter(evid == 1L) |>
  dplyr::select(id, time, amt, treatment) |>
  as.data.frame()

conc_obj <- PKNCA::PKNCAconc(
  data    = sim_for_nca,
  formula = Cc ~ time | treatment + id,
  concu   = "ug/L",
  timeu   = "hr"
)
dose_obj <- PKNCA::PKNCAdose(
  data    = doses_for_nca,
  formula = amt ~ time | treatment + id,
  doseu   = "mg"
)

intervals <- data.frame(
  start      = 0,
  end        = Inf,
  cmax       = TRUE,
  tmax       = TRUE,
  aucinf.obs = TRUE,
  half.life  = TRUE
)

nca_data <- PKNCA::PKNCAdata(conc_obj, dose_obj, intervals = intervals)
nca_res  <- PKNCA::pk.nca(nca_data)
#> Warning: Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed
#> Requesting an AUC range starting (0) before the first measurement (0.25) is not allowed

nca_tab <- as.data.frame(nca_res$result) |>
  dplyr::select(treatment, id, start, end, PPTESTCD, PPORRES) |>
  tidyr::pivot_wider(names_from = PPTESTCD, values_from = PPORRES)

knitr::kable(
  head(nca_tab, 12),
  digits  = 3,
  caption = "Simulated NCA per subject (first 12 rows)."
)

Simulated NCA per subject (first 12 rows).
treatment	id	end	cmax	tmax	tlast	clast.obs	lambda.z	r.squared	adj.r.squared	lambda.z.time.first	lambda.z.time.last	lambda.z.n.points	clast.pred	half.life	span.ratio	aucinf.obs
iv	1001	Inf	62.432	0.5	24	2.026	0.111	1	1	3.00	24	6	2.022	6.258	3.356	NA
iv	1002	Inf	28.554	0.5	24	3.131	0.080	1	1	2.00	24	7	3.124	8.622	2.552	NA
iv	1003	Inf	55.354	0.5	24	0.075	0.256	1	1	1.00	24	9	0.075	2.711	8.484	NA
iv	1004	Inf	55.936	0.5	24	0.109	0.200	1	1	3.00	24	6	0.108	3.467	6.057	NA
iv	1005	Inf	33.715	0.5	24	2.399	0.093	1	1	3.00	24	6	2.396	7.477	2.809	NA
iv	1006	Inf	41.211	0.5	24	0.502	0.178	1	1	1.00	24	9	0.501	3.894	5.906	NA
iv	1007	Inf	46.585	0.5	24	0.007	0.369	1	1	0.75	24	10	0.007	1.880	12.366	NA
iv	1008	Inf	77.240	0.5	24	0.128	0.234	1	1	2.00	24	7	0.128	2.963	7.425	NA
iv	1009	Inf	67.854	0.5	24	0.101	0.238	1	1	3.00	24	6	0.101	2.907	7.223	NA
iv	1010	Inf	22.478	0.5	24	0.102	0.219	1	1	1.50	24	8	0.101	3.160	7.120	NA
iv	1011	Inf	62.551	0.5	24	0.107	0.246	1	1	2.00	24	7	0.107	2.815	7.815	NA
iv	1012	Inf	34.484	0.5	24	0.019	0.305	1	1	1.00	24	9	0.019	2.272	10.125	NA

Comparison against published bioavailability

auc_per_subject <- nca_tab |>
  dplyr::select(id, treatment, aucinf.obs) |>
  dplyr::left_join(
    dplyr::bind_rows(cov_oral, cov_iv) |>
      dplyr::select(id, BSA),
    by = "id"
  ) |>
  dplyr::mutate(
    dose_mg            = 1.5 * BSA,
    auc_per_mg_per_m2  = aucinf.obs / 1.5
  )

mean_auc_per_mg_m2 <- auc_per_subject |>
  dplyr::group_by(treatment) |>
  dplyr::summarise(
    mean_auc = mean(auc_per_mg_per_m2, na.rm = TRUE),
    sd_auc   = sd(auc_per_mg_per_m2,   na.rm = TRUE),
    auc_fold = max(auc_per_mg_per_m2, na.rm = TRUE) /
               min(auc_per_mg_per_m2, na.rm = TRUE),
    .groups  = "drop"
  )
#> Warning: There were 4 warnings in `dplyr::summarise()`.
#> The first warning was:
#> ℹ In argument: `auc_fold = max(auc_per_mg_per_m2, na.rm =
#>   TRUE)/min(auc_per_mg_per_m2, na.rm = TRUE)`.
#> ℹ In group 1: `treatment = "iv"`.
#> Caused by warning in `max()`:
#> ! no non-missing arguments to max; returning -Inf
#> ℹ Run `dplyr::last_dplyr_warnings()` to see the 3 remaining warnings.

f_observed <- mean_auc_per_mg_m2$mean_auc[mean_auc_per_mg_m2$treatment == "oral"] /
              mean_auc_per_mg_m2$mean_auc[mean_auc_per_mg_m2$treatment == "iv"]

knitr::kable(
  mean_auc_per_mg_m2,
  digits  = 2,
  caption = paste0("Dose-normalised AUC0-Inf by route. Empirical bioavailability ",
                   "(oral AUC / IV AUC) = ",
                   round(100 * f_observed, 1), "% (published value 32.4%).")
)

Dose-normalised AUC0-Inf by route. Empirical bioavailability (oral AUC / IV AUC) = NaN% (published value 32.4%).
treatment	mean_auc	sd_auc	auc_fold
iv	NaN	NA	NaN
oral	NaN	NA	NaN

The empirical bioavailability recovered from the dose-normalised oral-versus-IV AUC ratio should be close to the published 32.4%; small discrepancies (a few percentage points) arise from the finite cohort size and the log-normal IIV on F (22% CV) plus IIV on CL (29% CV) interacting in the simulated NCA. The AUC fold-spread per route also varies with the finite N. Leger 2004 reports 4.8-fold variability for IV and 7.6-fold for oral.

Assumptions and deviations

Residual error encoded for the oral route only. Leger 2004 reports two route-specific residual error models in the Results section: proportional 11% with additive 0.64 ug/L for the IV data, and proportional 17% with additive 0.09 ug/L for the oral data. The different additive values largely reflect the different HPLC lower limits of quantification at the two analytical sites (Toulouse 0.5 ng/mL for IV; Rotterdam 0.1 ng/mL for oral). The packaged model uses the oral-route residual error because (a) the paper’s primary contribution is oral topotecan PK and the limited-sampling strategy is developed for oral, and (b) 118 of 190 patients (62%) are oral. For an IV-only simulation, override propSd to 0.11 and addSd to 0.64 ug/L in the loaded model before calling rxSolve().
Interoccasion variability omitted. Leger 2004 reports “interday variability” of 18% on CL, 49% on V1, and 28% on F as a separate IOV term in Table 3. The packaged model omits this layer because (a) the standard popPK extraction convention in this library is to capture only inter-individual variability and (b) IOV requires an OCC column in the event dataset that would constrain downstream use. The 18% / 49% / 28% IOV values are noted in the model file as comments so a user wishing to add an OCC dimension can wire them in.
No race / ethnicity covariate. Leger 2004 does not report the cohort’s race or ethnicity distribution (the paper notes only “French and Dutch oncology centres”); the virtual cohort therefore does not stratify on race. The population$race_ethnicity field records this as “Not reported”.
Cohort body-surface area derived from weight only. The virtual cohort uses the DuBois formula on body weight assuming a fixed 170 cm reference height to derive BSA for dosing (1.5 mg/m^2 x BSA). Real cohorts have independent height variation; this assumption is immaterial for the structural PK validation but means the simulated per-subject dose has a narrower distribution than the original cohort.
WHO PS draw uses the empirical 79 / 98 / 11 / 2 distribution. The cohort’s PS distribution is encoded directly (Table 1). PS values are treated as continuous-integer per the encoded covariate model (canonical column WHO_PS, added to inst/references/covariate-columns.md as part of this extraction). The (1 - 0.12 * WHO_PS) form means a WHO_PS = 3 patient has 36% lower CL than a reference WHO_PS = 0 patient.
Cycle 1 vs Cycle 2 estimates. The model file encodes the cycle-1 final-model values from Table 3 (theta1 = 12.8, theta2 = 2.1, theta3 = 0.12, theta4 = 0.58 etc.). Leger 2004 also fits a cycle-1-plus-cycle-2 pooled model that gives slightly different coefficients (CL = (11.4 + 2.5 * CrCl) * (1 - 0.06 * PS), V1 = 0.50 * body weight) plus inter-cycle variability of 18% on CL, 31% on V1, and 22% on F. These pooled-cycle values are not encoded; the paper highlights the cycle-1 fit as the primary covariate model.
Alternative covariate models from Table 3 not encoded. Table 3 reports several reduced-form alternative covariate models (CL = theta1 alone, CL = theta1 + theta2 * CrCl alone, CL = theta1 * (1 - theta3 * PS) alone, V1 = theta4 alone) with their objective-function deltas; these are model-development byproducts shown for comparison, not the final reported model, and are not encoded per the replicate-author-structure policy.