Table 2.
Parameter estimates of the final pharmacokinetic-pharmacodynamic model
| Parameter | Estimates (% RSE) | 95% CI | %CV BSV (% RSE) | 95% CI |
|---|---|---|---|---|
| Pharmacokinetics | ||||
| Artesunate | ||||
| F (%) | 100 fix | – | 31.2 (29.4) | 19.3–50.8 |
| MTT (h) | 1.34 (18.8) | 1.04–1.96 | 85.3 (24.9) | 65.7–133.0 |
| CLARS/F (l/h) | 1750 (8.55) | 1570–2090 | 26.8 (44.3) | 11.9–39.1 |
| VARS/F (l) | 1300 (12.6) | 1110–1660 | 74.7 (27.3) | 57.8–129 |
| RUV (%) | 73.2 (3.95) | 69.3–78.7 | – | – |
| Dihydroartemisinin | ||||
| CLDHA/F (l/h) | 76.7 (6.99) | 69.9–87.8 | 21.3 (30.3) | 13.3–88.1 |
| VDHA/F (l) | 102.0 (8.95) | 89.5–119.0 | 31.6 (40.5) | 21.3–131.0 |
| RUV (%) | 58.5 (3.34) | 56.6–63.4 | – | – |
| Covariate effects | ||||
| aPARAMTT (Log10 parasitaemia) | 0.115 (8.88) | 0.121–0.156 | – | – |
| aPARAmaxF | 1.51 (11.9) | 1.35–2.02 | – | – |
| aPARA50F (Log10 parasitaemia) | 8.32 (3.58) | 8.19–9.21 | – | – |
| Pharmacodynamics | ||||
| KGROWTH (48 h−1) | 10 fix | – | ||
| BASEPARA (Log10) | 11.0 (0.704) | 10.8–11.1 | 4.4 (19.6) | 3.13–5.78 |
| ke0 (h−1) | 0.123 (33.1) | 0.0584–0.188 | – | |
| aEC50 (nM) | 30.4 (34.2) | 13.5–46.1 | – | |
| aEmaxS (h−1) | 0.268 (5.89) | 0.242–0.295 | b49.0 (22.4) | 34.3–70.1 |
| EmaxR (h−1) | 0.155 (6.08) | 0.142–0.172 | 12.2 (45.5) | 6.54–35.8 |
| PMIX, resistant (%) | 56.1 (20.9) | 39.1–73.8 | – | |
| RUV (%) | 33.3 (5.91) | 30.5–37.1 | – | |
Coefficient of variation (%CV) of between subject variability (BSV) was calculated as 100 × (variance-1)1/2. Relative standard errors (% RSE) were calculated as 100 × (standard deviation/mean). The 95% confidence intervals (95% CI) of parameter estimates were obtained with the Sampling Importance Resampling (SIR) approach
ARS artesunate, BASEPARA baseline parasitaemia, CL clearance, DHA dihydroartemisinin, F bioavailability, KGROWTH parasite multiplication per 48 h parasite cycle, MTT mean transit time, PARAMTT estimated linear effect of parasite density on MTT, PARAmaxF maximum effect of parasite density on F, PARA50F parasite density which produces 50% of the maximum covariate response, PMIX, resistant probability of having an artemisinin-resistant infection, V volume of distribution, EC50 the DHA concentration which produces 50% of maximum parasite killing effect, EmaxR maximum parasite killing effect of a resistant parasite population, EmaxS maximum parasite killing effect of a sensitive parasite population, ke0 effect compartment rate constant governing the delayed drug effect, RUV unexplained residual variability
aEstimation of these parameters were obtained by applying a frequentist prior approach using a previously published PK/PD model developed on data from Thailand and Cambodia (37)
bBSV (%CV) of EmaxS was calculated based on simulations (10,000 patients) with an estimated variance of 0.430 and the applied transformation presented in Eq. 7