Page 37 - 2021_05-Haematologica-web

Individualized PEGasparaginase dosing guideline
Table 2. Pharmacokinetic models. Parameter
OFV
CL (L/day/m2)
Vd (L/m2)
Slope CLind (L/day/m2/day) Split point (days) Fractional increase IIV-Vd
Structural model Mean (RSE)
-1140.4
0.075 (5.2%) 0.92 (4.7%) 0.079 (31.5%) 12.9 (1.1%) 1.11 (13.4%)
Final model Mean (RSE)
-1203.77
0.084 (4.4%) 0.94 (4.5%) 0.082 (20.5%) 12.7 (0.2%) 1.26 (10.2%)
Bootstrap of the final model
Estimate
0.084 0.94 0.080 12.7 1.28
95% CI
0.078 – 0.090 0.87 – 1.01 0.052 – 0.115 11.8 – 13.1 0.98 – 1.51
Covariates Treatment phase
1B
M
MRG intens. MRG maint. SRG protocol IV
Infection
IIV
CL (%)
Vd
IOV
CL (%)
Residual variability
Additional (IU/L)
Shrinkage
-
- 1(fix) - -
1A
ref
0.87 (5.2%) 0.87 0.89 (5.2%) 0.88 0.81 (3.9%) 0.81 0.81 (6.5%) 0.81 1.38 (10.5%) 1.38
19.7 (13.2%) 19.5
-
0.80 – 0.95 0.82 – 0.98 0.75 – 0.86 0.73 – 0.90 1.15 – 1.67
14.6 – 24.0
NA 18.5 – 27.5
12.7 – 298
- - - - -
24.1 (12.8%)
ref
NA NA NA
25.7 (10.1%)
19.1 (26.2%)
23.6 (10.3%) 23.1
20.2 (18.2%) 20.2
Proportional (%)
14.2 – 19.5
17.3 (9.2%)
IIVCL 22.0% 24.0% - -
17.0 (9.2%) 16.8
IOV 5.0% 15.0% - - Final model:
TAD <12.7 days:
CL = 0.084 * BSA * 1.38INFECTION * 0.87M * 0.89MRG INTENS. * 0.81MRG MAINT. * 0.81SRG PROTOCOL IV * emηCL-IIV +ηCL-IOV
TAD >12.7 days:
CL = 0.084 * BSA * (1+0.082 * (TAD – 12.7)) * 1.38INFECTION * 0.87M * 0.89MRG INTENS. * 0.81MRG MAINT. * 0.81SRG PROTOCOL IV * e IIV + IOV Vd = 0.94 * e1.26 * CL-IIV)
Values for INFECTION, M, MG INTENS., MRG MAINT. and SRG PROTOCOL IV: 0 (no) or 1 (yes). RSE: relative standard error; CI: Confidence Interval; OFV: objective function value; CL: clearance,Vd: volume of distribution, Cind: induced clearance; MRG intens.: medium risk group intensification; MRG maint.: medium risk group maintenance; IIV: inter-indi- vidual variability; IOV: interoccasion variability; NA: not applicable; TAD: time after dose.
Goodness of fit and model validation
The parameter estimates of the final model were pre- cise with regards to the relative standard errors, and the shrinkage values were acceptable (Table 2). The good- ness of fit plots showed an even distribution of the pop- ulation predictions and individual predictions around the line of unity (Online Supplementary Figure S1). The condi- tional weighted residuals, plotted against the time after dose, were also evenly distributed with a mean of zero, and no trend was found. The bootstrap analysis showed that the bootstrap estimates were consistent with the parameter estimates of the final model and the 95% Confidence Intervals (CI) were accurate (Table 2). The visual predictive check of the index dataset is shown in Figure 3. Both the median estimates and the 95% CI are within the simulated predicted values. Thus, the final model is accurate in predicting the PEGasparaginase popPK.
As an external validation, the final model derived from the index dataset was used to predict the asparaginase activity levels of the independent validation dataset. The
goodness of fit plots and VPC of the validation dataset also show that the final model is adequate to describe the PEGasparaginase PK (Online Supplementary Figures S1 and S3).
Clinical implications
Table 4 shows several dosing regimens for PEGasparaginase (loading dose and maintenance dose) during induction and MRG intensification. Targeting a trough PEGasparaginase activity level of 100 IU/L, the loading dose recommended during induction is 800 IU/m2 followed by a biweekly maintenance dose of 600 IU/m2. Lower doses should be administered due to decreased CL for PEGasparaginase treatment during intensification, with doses being 600 IU/m2 and 400 IU/m2, respectively, to target 100 IU/L.
Based on the half-life during the first 13 days, the steady state is estimated to be reached after two doses, after which the dose can be adjusted based on trough or week asparaginase activity levels targeting trough levels between 100-250 IU/L or 250-400 IU/L (Table 5). If an
haematologica | 2021; 106(5)
1257