Table 3.
Random effect (variance) parameter estimates for the models
| Interindividual random effects, ω, as %CV (%RSE) [Shrinkage*] | ||||
|---|---|---|---|---|
| Omalizumab or IgE parameter | Control, constant | Exponential | Feedback interim | Feedback final |
| Clearance omalizumab, CLX/F | 33% | 32% (11%) | 32% (11%) | 37% (11%) [17%] |
| Clearance IgE, CLE/F | 16% | 13% (45%) | 13% (46%) | 45% (31%) [58%] |
| Clearance complex, CLC/F | 23% | 19% (24%) | 19% (25%) | 36% (38%) [55%] |
| Volume omalizumab and IgE, VX/F and VE/F | 24% | 24% (22%) | 24% (22%) | 26% (23%) [28%] |
| Volume complex, VC/F | 21%† (13%) [76%] | |||
| Absorption rate, ka | 57%† (7.9%) [51%] | |||
| Binding dissociation constant, Kd | 12% | 17% (21%) | 17% (21%) | 23% (18%) [38%] |
| IgE production prior to omalizumab, RB | 32% | 29% (15%) | 30% (15%) | 56% (20%) [38%] |
| IgE production at new equilibrium, RN | – | 180% (54%) | 93% (27%) | 73% (21%) [63%] |
| Rate of change in IgE production with omalizumab‡ | – | 835 (31%) | No random effect | |
| Background rate of change in IgE production, kE‡ | – | 1050§ (12%) [44%] | ||
| Residual random effects, σ, as %CV (%RSE) [Shrinkage¶] | ||||
|---|---|---|---|---|
| Omalizumab | 36% | 37% (17%) | 37% (17%) | 36% (18%) [13%] |
| Total IgE | 34% | 27% (6.8%) | 28% (6.7%) | 27% (6.7%) [16%] |
| Free IgE | 37% | 34% (5.8%) | 35% (5.8%) | 35% (5.9%) [13%] |
| Objective function value | −18 178 | −22 847 | −22 927 | −23 446 |
| Terminated with rounding errors | −4 669 from control P < 0.001 for 4 d.f. | −4 748 from control P < 0.001 for 4 d.f. | −512 from feedback interim P < 0.001 for 14 d.f. | |
Note that minimization for the control model terminated with rounding errors and should be regarded cautiously. RSE, relative standard errors; these are missing from the control model as it did not run the $COV procedure. Note that for the final model a covariate search was conducted for patient factors that may affect the change in IgE production with time and that covariance was allowed between the parameters CLX, CLE, CLC, VX, RB and Kd.
Shrinkage in the post hoc ETA estimates (ηph) was calculated as 1 − SD(ηph)/√ω.
Values for VC and ka were fixed from a separate first order conditional with interaction estimation using data from single-dose bioequivalence studies.
Variance reported in original units rather than as CV =√(exp(ω) − 1) as an additive error model was specified.
Estimated independently from placebo data and fixed in the models.
Shrinkage in the residual error ε was calculated as 1 − SD (residual)/√σ. Subjects receiving placebo were discarded for PK-related parameters.