Population pharmacokinetics and pharmacodynamics of intravenous immunoglobulin treatment in patients with Guillain-Barré syndrome
S.J. van Tilburg (1), S.D.T Sassen (2), W.J.R Fokkink (1,3), C. Walgaard (3), B.C.M. de Winter (2), P.A. van Doorn (3), B.C.P. Koch (2), B.C. Jacobs (1,3)
(1) Department of Immunology, Erasmus MC, University Medical Center Rotterdam, The Netherlands. (2) Department of Hospital Pharmacy, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands (3) Department of Neurology, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands
Introduction: The Guillain-Barré syndrome (GBS) is an acute immune-mediated polyneuropathy frequently triggered by a preceding infection. Clinically GBS is characterized by a rapidly progressive weakness of the limbs and in 25% of respiratory muscles. GBS has a typical monophasic course followed by a slow and frequently incomplete clinical recovery. Treatment of choice for GBS is intravenous immunoglobulin (IVIg) at a standard dose of 0.4 g/kg/day for five consecutive days, yet response to treatment is highly variable. Previous studies showed that a higher increase in total IgG after two weeks is related to clinical outcome, suggesting a relation between IVIg exposure and PD [1]. A second dose of IVIg, exclusively administered to patients with a poor predicted outcome one week after the initial course, was examined in a recent double-blind, randomized placebo-controlled phase 3 trial (SID-GBS) [2]. Surprisingly, the trial did not demonstrate improvement in the clinical outcomes in these patients receiving a second IVIg course. Parallel to this trial we recently developed the first two compartment population PK model to predict the pharmacokinetics of a standard dosage (one course) of IVIg in patients with GBS [3]. The model was previously externally evaluated using the non-randomized (single IVIg course) patients (those with a predicted good prognosis) from the SID-GBS cohort, but not the patients receiving a second course.
Objectives:
- To evaluate the previously described population PK model in patients randomized to receive either a second IVIg course or placebo (SID-GBS trial).
- To describe the relation between the PK and PD in all patients enrolled in the SID-GBS trial.
Methods: In this study, non-linear mixed effect modelling (NONMEM) was employed to perform an external evaluation of the model using patients randomized for either a second IVIg course or placebo (poor prognosis) included in the SID-GBS trial. Model performance was evaluated using diagnostic plots and visual predictive checks. The precision and bias of the model were assessed through the computation of the mean prediction error (MPE) and root mean square error (RMSE).
Total IgG exposure (AUC) at two and four weeks after the initiation of treatment was determined via post-hoc analysis for all individual patients included in the SID-GBS trial (including non-randomized). Regression analysis and Spearman's rank correlation were employed to assess associations with clinical outcomes, including the Medical Research Council (MRC) sum score, a measure for weakness of 12 limb muscle groups ranging from 0 (quadriplegic) to 60 (normal strength). Kaplan–Meier analysis was used to estimate the time to regain the ability to walk unaided stratified by the exposure in quartiles.
Results: For the external evaluation of the model 90 randomized patients were included, with 339 serial serum IgG levels. Of these patients, 49 received a second IVIg course, and 41 received placebo. Stratification of the visual predictive checks by treatment arm demonstrated that the median and variability of the observations were consistent with the model-predicted boundaries, indicating the accuracy of the model in predicting the pharmacokinetics after multiple IVIg courses. The individual predicted values exhibited no bias (MPE = -0.28), and good precision was observed (RMSE = 10.4%).
Pharmacodynamic analysis was performed using the complete SID cohort consisting of 267 GBS patients including 1067 IgG measurements, stratified for treatment group (non-randomized: one IVIg course; randomized: IVIg + placebo, or 2 IVIg courses). Significant correlation was observed between the total IgG exposure at two weeks in patients receiving one IVIg course and the MRC sum score at 12 weeks (rho: 0.2, p = 0.003). Survival analysis of patients receiving a single IVIg course revealed that the group of patients with the lowest total IgG exposure at both two (log-rank p = 0.018) and four weeks (log-rank p = 0.025) were less likely to regain the ability to walk. No significant relation between PK and PD was found in the patients receiving two IVIg courses.
Conclusions: This study shows that the PK model for IVIg accurately predicts IgG levels after two courses of IVIg. Importantly, we show that a lower IgG exposure is still related to a worse clinical outcome in patients with GBS. This suggests that a subset of patients may benefit from higher dosing at treatment initiation.
References:
[1] Kuitwaard, K., de Gelder, J., Tio-Gillen, A.P., Hop, W.C.J., van Gelder, T., van Toorenenbergen, A.W., van Doorn, P.A. and Jacobs, B.C. (2009), Pharmacokinetics of intravenous immunoglobulin and outcome in Guillain-Barré syndrome. Ann Neurol., 66: 597-603.
[2] Walgaard, C., Jacobs, B. C., Lingsma, H. F., Steyerberg, E. W., van den Berg, B., Doets, A. Y., ... & Trip, R. (2021). Second intravenous immunoglobulin dose in patients with Guillain-Barré syndrome with poor prognosis (SID-GBS): a double-blind, randomised, placebo-controlled trial. The Lancet Neurology, 20(4), 275-283.
[3] Fokkink, W. J. R., van Tilburg, S. J., de Winter, B. C., Sassen, S. D., van Doorn, P. A., Koch, B. C., & Jacobs, B. C. (2022). Population Pharmacokinetic Modelling of Intravenous Immunoglobulin Treatment in Patients with Guillain–Barré Syndrome. Clinical Pharmacokinetics, 61(9), 1285-1296.