Skip to contents

Miridesap / CPHPC (Sahota 2015)

Source: vignettes/articles/Sahota_2015_miridesap.Rmd
Sahota_2015_miridesap.Rmd

Model and source

  • Citation: Sahota T, Berges A, Barton S, Cookson L, Zamuner S, Richards D. Target Mediated Drug Disposition Model of CPHPC in Patients With Systemic Amyloidosis. CPT Pharmacometrics Syst Pharmacol. 2015;4(2):e15. doi:10.1002/psp4.15.
  • Description: Target-mediated drug disposition (TMDD) PK/PD model for CPHPC (miridesap, GSK2315698, Ro 63-8695) and serum amyloid P (SAP) in healthy volunteers (study CPH113776) and patients with systemic amyloidosis (study CPH114527). Two-compartment PK for CPHPC (IV plus first-order subcutaneous depot); two-compartment turnover model for SAP with first-order endogenous production and elimination; bimolecular CPHPC + free SAP -> complex binding treated as effectively irreversible (KOFF set to zero because the complex internalisation rate is much faster than the dissociation rate). Final-model covariates (Sahota 2015 Eq. 1 and Eq. 2): creatinine clearance modifies CPHPC clearance below an 80 mL/min threshold; hepatic amyloid involvement multiplies SAP intercompartmental clearance Q4; whole-body amyloid load (categorical 0-3) multiplies SAP peripheral volume V4 in two cumulative steps; biological sex multiplies baseline plasma SAP.
  • Article: https://doi.org/10.1002/psp4.15

The model captures the SAP-depletion pharmacology of CPHPC (also referred to by its INN miridesap, or the GSK code GSK2315698 / Ro 63-8695): a small molecule that binds plasma serum amyloid P component (SAP) and rapidly depletes it from the circulation. The TMDD formulation treats the SAP-CPHPC complex as cleared by the liver and the dissociation rate as negligible relative to internalisation, consistent with Sahota 2015 Discussion (“the model assumes that the elimination rate constant of the complex from the central compartment, KINT, is much faster than its dissociation”).

Population

Sahota 2015 Table 1 summarises two pooled studies:

  • CPH113776 (NCT01323985): 21 male healthy volunteers with normal renal function and no amyloid, receiving short (1 h) and extended (24 h) IV infusions of CPHPC at doses spanning 5-70 mg (single dose) and 86-960 mg (24 h infusion regimens).
  • CPH114527 (NCT01406314): patients with systemic amyloidosis in four cohorts, spanning small-to-large whole-body amyloid load and normal-to-moderate-severe renal impairment. Each subject received a 48 h IV infusion (124.8-1440 mg) followed by one or three SC doses (10-60 mg).

A total of 38 subjects across the two studies contributed PK (total plasma CPHPC) and PD (total plasma SAP) measurements. Finer demographics (precise age range, weight range, female fraction) are reported only in the paper’s Supplementary Materials S1 and S2, which were not on disk for this extraction; the model’s population metadata records the fields that are derivable from the main paper.

The same information is available programmatically via rxode2::rxode(readModelDb("Sahota_2015_miridesap"))$meta$population.

Source trace

Each parameter line in inst/modeldb/specificDrugs/Sahota_2015_miridesap.R carries an in-file comment pointing to the matching Table 2 row, Eq. 1 / Eq. 2 term, or paper-narrative source. The table below consolidates the trace for review.

Equation / parameter Value Source location
lcl – CPHPC CL at CRCL >= 80 mL/min log(6.85) Table 2 (clearance row; footnote a)
e_crcl_cl – CRCL slope below threshold 0.015 per mL/min Table 2 (CRCL x clearance); Eq. 1
lvc – CPHPC V1 log(16.15) Table 2 (central volume V1)
lq – CPHPC Q2 log(1.72) Table 2 (intercompartmental clearance Q2)
lvp – CPHPC V2 log(17.57) Table 2 (peripheral volume V2)
lka – SC absorption rate KSC log(1.5), FIX Table 2 (KSC: 1.5 FIX); Methods
lfdepot – SC bioavailability log(1), FIX Methods: “bioavailability assumed complete for SC doses”
lkout – SAP elimination KOUT log(0.046) Table 2 (KOUT)
lsap0 – baseline SAP (males) log(31.10) Table 2 (SAP_BASE)
V3 = V1 (structural constraint) n/a Methods (Binding and elimination): “common central compartment (V1 = V3)”
lvp_sap – SAP V4 reference log(12.15) Table 2 (peripheral volume V4)
lq_sap – SAP Q4 reference log(2.84) Table 2 (intercompartmental clearance Q4)
e_amliver_q4 4.01 Table 2 (Amyloid liver x Q4); Eq. 2
e_sexf_sap0 -0.30 Table 2 (Gender x SAP baseline); Eq. 2
e_amload2_vp_sap 6.39 Table 2 (Amyloid load x V4, moderate); Eq. 2
e_amload3_vp_sap 26.39 Table 2 (Amyloid load x V4, large); Eq. 2
lkon – binding association rate log(1.94e6) Table 2 (association rate KON)
lkint – complex internalisation log(5.78) Table 2 (complex elimination KINT)
KOFF 0 (set, not estimated) Methods (Binding and elimination): “Mathematically, this assumption is equivalent to setting the dissociation rate constant, KOFF, to zero.”
propSd – residual PK SD 0.2862 Table 2 (Residual error, PK: 28.62%)
propSd_sap – residual SAP SD 0.2710 Table 2 (Residual error, SAP: 27.10%)
Two-compartment CPHPC PK structural Methods (CPHPC PK) and Figure 3 schematic
Two-compartment SAP turnover structural Methods (SAP turnover and distribution) and Figure 3
Bimolecular 1:1 binding with KOFF=0 structural Methods (Binding and elimination); 1:1 ratio adopted because it “yielded greater numerical stability”

BSV magnitudes for each parameter come from Table 2’s “BSV (%RSE)” column. The paper reports CV-percent; the model converts via the log-normal relation omega^2 = log(1 + CV^2). Parameters with the label “15% FIX” in Table 2 are wrapped in fixed() in ini(); identifiability of the binding kinetics required the authors to hold KON, KINT, KSC, Q2, and V2 BSVs constant at 15 percent.

Virtual cohort

Original observed data are not publicly available. The figures below use virtual subjects whose covariate distributions reflect the patient strata described in Sahota 2015 Table 1.

set.seed(20260517)

make_cohort <- function(n, crcl, sexf, amload, amliver,
                        dose_mg, infusion_dur_h, sample_grid,
                        cohort_label, id_offset = 0L) {
  ids <- id_offset + seq_len(n)
  dose_rows <- data.frame(
    id   = ids,
    time = 0,
    amt  = dose_mg,
    rate = dose_mg / infusion_dur_h,
    evid = 1L,
    cmt  = "central"
  )
  obs_rows <- expand.grid(id = ids, time = sample_grid)
  obs_rows$amt  <- 0
  obs_rows$rate <- 0
  obs_rows$evid <- 0L
  obs_rows$cmt  <- "Cc"
  out <- dplyr::bind_rows(dose_rows, obs_rows)
  out$CRCL    <- crcl
  out$SEXF    <- sexf
  out$AMLOAD  <- amload
  out$AMLIVER <- amliver
  out$cohort  <- cohort_label
  out[order(out$id, out$time, -out$evid), ]
}

# Replicate the Figure 5 left-column scenarios: a typical male, 20 mg/h
# IV infusion for 48 h followed by three 60 mg SC doses every 8 h, with
# AMLOAD swept across 0 (no amyloid), 1 (small), 2 (moderate), and 3
# (large) and AMLIVER fixed at 0 for these AMLOAD-sweep panels.
sample_grid <- c(seq(0, 80, by = 0.5), seq(81, 720, by = 4))

events_amload <- dplyr::bind_rows(
  make_cohort(1L, crcl = 90, sexf = 0, amload = 0, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid, cohort_label = "AMLOAD = 0 (healthy)",
              id_offset = 0L),
  make_cohort(1L, crcl = 90, sexf = 0, amload = 1, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid, cohort_label = "AMLOAD = 1 (small)",
              id_offset = 1L),
  make_cohort(1L, crcl = 90, sexf = 0, amload = 2, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid, cohort_label = "AMLOAD = 2 (moderate)",
              id_offset = 2L),
  make_cohort(1L, crcl = 90, sexf = 0, amload = 3, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid, cohort_label = "AMLOAD = 3 (large)",
              id_offset = 3L)
)
stopifnot(!anyDuplicated(unique(events_amload[, c("id", "time", "evid")])))

# Replicate the Figure 5 right-column scenarios: AMLOAD fixed at 2
# (moderate), AMLIVER fixed at 0, CRCL swept from severe impairment (15)
# to normal (90) so the renal effect on CPHPC PK is visible.
events_crcl <- dplyr::bind_rows(
  make_cohort(1L, crcl = 15, sexf = 0, amload = 2, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid,
              cohort_label = "CRCL = 15 (severe)", id_offset = 100L),
  make_cohort(1L, crcl = 45, sexf = 0, amload = 2, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid,
              cohort_label = "CRCL = 45 (moderate)", id_offset = 101L),
  make_cohort(1L, crcl = 90, sexf = 0, amload = 2, amliver = 0,
              dose_mg = 20 * 48, infusion_dur_h = 48,
              sample_grid = sample_grid,
              cohort_label = "CRCL = 90 (normal)", id_offset = 102L)
)
stopifnot(!anyDuplicated(unique(events_crcl[, c("id", "time", "evid")])))

Simulation 

mod <- readModelDb("Sahota_2015_miridesap")
mod_typical <- rxode2::zeroRe(mod)
#> ℹ parameter labels from comments will be replaced by 'label()'

sim_amload <- rxode2::rxSolve(mod_typical, events = events_amload,
                              keep = c("cohort")) |>
  as.data.frame()
#> ℹ omega/sigma items treated as zero: 'etalcl', 'etalvc', 'etalq', 'etalvp', 'etalka', 'etalkout', 'etalsap0', 'etalq_sap', 'etalvp_sap', 'etalkon', 'etalkint'
#> Warning: multi-subject simulation without without 'omega'

sim_crcl <- rxode2::rxSolve(mod_typical, events = events_crcl,
                            keep = c("cohort")) |>
  as.data.frame()
#> ℹ omega/sigma items treated as zero: 'etalcl', 'etalvc', 'etalq', 'etalvp', 'etalka', 'etalkout', 'etalsap0', 'etalq_sap', 'etalvp_sap', 'etalkon', 'etalkint'
#> Warning: multi-subject simulation without without 'omega'

Replicate published figures 

# Replicates the top-left panel of Sahota 2015 Figure 5: typical-value
# CPHPC plasma profile across amyloid loads. AMLOAD does not enter the
# CPHPC PK pathway, so the four traces overlay -- consistent with the
# paper's Discussion ("the amyloid load, however, had no effect on the
# PK profile of CPHPC").
sim_amload |>
  dplyr::filter(time <= 80) |>
  ggplot(aes(time, Cc, colour = cohort)) +
  geom_line() +
  labs(x = "Time (h)", y = "CPHPC plasma (ng/mL)",
       title = "Figure 5 top-left -- CPHPC PK across amyloid load",
       caption = "Replicates Sahota 2015 Figure 5 top-left.") +
  theme(legend.position = "bottom")

# Replicates the middle-left panel of Sahota 2015 Figure 5: SAP profile
# over the first 80 h. The depletion nadir deepens to ~0.5 mg/L in
# AMLOAD = 0/1 subjects and is shallower in AMLOAD = 2/3 because the
# amyloid load increases the peripheral SAP volume V4, releasing
# additional SAP into circulation.
sim_amload |>
  dplyr::filter(time <= 80) |>
  ggplot(aes(time, sap / 1000, colour = cohort)) +
  geom_line() +
  labs(x = "Time (h)", y = "Plasma SAP (mg/L)",
       title = "Figure 5 middle-left -- SAP depletion across amyloid load (0-80 h)",
       caption = "Replicates Sahota 2015 Figure 5 middle-left.") +
  theme(legend.position = "bottom")

# Replicates the bottom-left panel of Sahota 2015 Figure 5: SAP profile
# across the full 0-720 h horizon. Recovery to baseline takes ~200 h in
# AMLOAD = 0/1 and extends well past 600 h in AMLOAD = 3, mirroring the
# paper's narrative ("the plasma SAP concentration in this patient
# returned to its baseline value after more than 600 hours compared
# with 200 hours in a patient with no/small amyloid load").
sim_amload |>
  ggplot(aes(time, sap / 1000, colour = cohort)) +
  geom_line() +
  labs(x = "Time (h)", y = "Plasma SAP (mg/L)",
       title = "Figure 5 bottom-left -- SAP recovery across amyloid load (0-720 h)",
       caption = "Replicates Sahota 2015 Figure 5 bottom-left.") +
  theme(legend.position = "bottom")

# Replicates the top-right panel of Sahota 2015 Figure 5: CPHPC PK
# across CRCL levels. Lower CRCL increases CPHPC exposure roughly
# 4-fold between normal renal function and severe impairment, in line
# with the paper's narrative ("CPHPC plasma concentrations were
# fourfold higher in a typical subject with severe renal impairment").
sim_crcl |>
  dplyr::filter(time <= 80) |>
  ggplot(aes(time, Cc, colour = cohort)) +
  geom_line() +
  labs(x = "Time (h)", y = "CPHPC plasma (ng/mL)",
       title = "Figure 5 top-right -- CPHPC PK across renal function",
       caption = "Replicates Sahota 2015 Figure 5 top-right.") +
  theme(legend.position = "bottom")

PKNCA validation

Sahota 2015 does not tabulate NCA parameters in the main paper, so the PKNCA block below computes them on the simulated CPHPC profiles for the AMLOAD-sweep cohort only (the SAP output is a PD response of total plasma SAP rather than a classic concentration-time profile, and the paper’s clinical objective is SAP depletion, not exposure). The treatment grouping uses the AMLOAD-derived cohort label so the single-row reference summary matches the figure replicates above.

sim_nca <- sim_amload |>
  dplyr::filter(!is.na(Cc), time > 0, time <= 200) |>
  dplyr::select(id, time, Cc, cohort)

dose_nca <- events_amload |>
  dplyr::filter(evid == 1L) |>
  dplyr::select(id, time, amt, cohort)

conc_obj <- PKNCA::PKNCAconc(sim_nca, Cc ~ time | cohort + id)
dose_obj <- PKNCA::PKNCAdose(dose_nca, amt ~ time | cohort + id)

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  <- suppressWarnings(PKNCA::pk.nca(nca_data))
nca_summary <- as.data.frame(summary(nca_res))
knitr::kable(nca_summary,
             caption = "Simulated CPHPC NCA parameters by AMLOAD-sweep cohort (typical-value subject; 20 mg/h IV infusion x 48 h).")
Simulated CPHPC NCA parameters by AMLOAD-sweep cohort (typical-value subject; 20 mg/h IV infusion x 48 h).
start end cohort N cmax tmax half.life aucinf.obs
0 Inf AMLOAD = 0 (healthy) 1 2890 48.0 7.96 NC
0 Inf AMLOAD = 1 (small) 1 2890 48.0 7.96 NC
0 Inf AMLOAD = 2 (moderate) 1 2880 48.0 8.31 NC
0 Inf AMLOAD = 3 (large) 1 2870 48.0 8.02 NC

Assumptions and deviations

  • Source breadth. This extraction is paper-only. Sahota 2015’s Supplementary Materials S1 (participant demographics) and S2 (adverse-effect tabulations) and Supplementary Material S3 (NONMEM control stream) were not on disk for this extraction; finer demographics and any control-stream-only conventions are therefore absent from population and from the in-file parameter comments. A parallel extraction from the DDMORE Foundation Model Repository entry for the same paper is available as rxode2::rxode(readModelDb("NA_NA_miridesap"))$meta; that companion model carries control-stream-derived final estimates (slightly different rounding, e.g. CL = 6.82 vs paper 6.85) and additional structural detail. The two models are clinically equivalent for typical-value simulation.
  • Molecular-weight assumptions. The model converts CPHPC plasma mass to molar concentration with MW_CPHPC = 340 g/mol and SAP pentamer mass to molar concentration with MW_SAP = 127,310 g/mol, both drawn from Sahota 2015 Discussion (“relative molecular mass Mr 340” and “SAP, Mr 127,310”). The DDMORE control stream uses 5 x 25 kDa = 125,000 g/mol for the SAP pentamer; the two values differ by ~2% and produce indistinguishable typical-value simulations because the KON estimate (1.94e6 L/(mol h)) absorbs the choice.
  • target_peripheral compartment name. The model carries the SAP peripheral pool in a state named target_peripheral rather than the canonical peripheral2. The skill’s compartment register (R/conventions.R::targetLocationRegex) currently recognises target_csf, target_isf, complex_csf, complex_isf but not target_peripheral; nonetheless the name is the natural extension used by two existing TMDD-extension models (NA_NA_miridesap, Aguiar_2021_ustekinumab). Renaming to peripheral2 would falsely imply a second peripheral of the parent drug (CPHPC); the paper-faithful semantic is “peripheral compartment of the target species (SAP).” This raises a single compartment-naming checkModelConventions() warning that we accept rather than rename away from the established TMDD pattern.
  • Units header. units$dosing = "mg" and units$concentration = "ng/mL (CPHPC plasma); ng/mL-equivalent for SAP" differ by a factor of 1000 in magnitude; the model model() block explicitly applies the conversion (Cc <- central / vc * 1000 and `sap <- total_target
    • MW_SAP * 1000) so the observation columns are in ng/mL while the internal volume is in L.checkModelConventions()` emits a units-magnitude info note for this combination; the conversion is intentional and verified.
  • Population scope. Sahota 2015 enrolled 21 healthy male volunteers in CPH113776; CPH114527 enrolled patients with systemic amyloidosis with a sex balance reported in the (off-disk) Supplementary Material S1. population$sex_female_pct is left as NA_real_ because the paper’s main text does not enumerate this figure.
  • Errata. No erratum or corrigendum is associated with the paper on the CPT Pharmacometrics & Systems Pharmacology landing page as of this extraction (last checked 2026-05-17).