Yasuhiro Tsuji (1), Chika Ogami (2), Daichi Yamaguchi (2), Hidefumi Kasai (3), Yoshiaki Matsumoto (1), Hideto To (2)
1) School of Pharmacy, Nihon University, Japan, (2) Faculty of Pharmacy and Pharmaceutical Sciences, University of Toyama, Japan, (3) School of Medicine, Keio University, Japan
Objectives: Lithium carbonate has been used for the treatment of the bipolar disorder for decades. Administered lithium carbonate immediately dissociates to lithium ions (Li+), and Li+ express antimanic effects [1]. The dose adjustment based on the blood concentrations of Li+ has been recommended because of its’ narrow therapeutic range. The adverse events such as tremor have been reported in patients with over doses of lithium carbonate [2]. Li+ is mainly excreted into urine, and approximately 70-80% of Li+ which are received glomerular filtration are reabsorbed in tubular [3]. The existing ratio of Li+ to lithium carbonate is variated by pH in urine, and this influence on tubular reabsorption of Li+.
- To develop the population pharmacokinetic (PK) model for lithium carbonate based on tubular reabsorption mechanism
- To evaluate the relationship between serum lithium concentration and occurring neurological symptoms including tremor which is a major adverse event of lithium
Methods: Clinical data including 389 serum lithium concentrations and pH in urine were collected from 214 subjects who received lithium carbonate for the treatment of psychiatric disorders. Basic PK model was described by 1-order distribution and elimination with 1-order absorption model. Additionally, a virtual renal compartment for considering the processes of glomerular filtration and tubular reabsorption of Li+ was assumed. The PK parameters were oral apparent clearance (CL/F), the apparent volume of distribution in central compartment (VC/F), inter-compartment clearance between central and virtual renal (QCR and QRC), and volume of the virtual renal compartment (VR).
Lithium carbonate has two dissociation constants (pKa, 6.2 and 9.7) and dissociates in two steps, as shown below:
2Li+ + CO32- ↔ Li+ + LiCO3- ↔ Li2CO3
2Li+ + CO32-, Li+ + LiCO3-, and Li2CO3 are denoted the ionic form (ION), the middle form (MID), and the molecular form (MOL) of lithium carbonate, respectively. The pH value in urine varies from 4.5 to 7.5, and few MOL exist in this pH (< 3%). In this study, it was assumed that only the ION and the MID were involved in the glomerular filtration and tubular reabsorption.
CL/F was defined as the sum of renal clearance (CLR) multiplied by renal function (RF) and non-renal clearance (CLNR). RF was the ratio of individual creatinine clearance (L/h) to standard clearance of 6 L/h calculated by Cockcroft-Gault formula [4]. CLR was assumed as the sum of the ION clearance (CLION) and the MID clearance (CLMID), each multiplied by their fractions FION and FMID.
CLR = CLION * FION + CLMID * FMID
FION and FMID were calculated by Henderson-Hasselbalch formula using urinary pH. QCR and QRC was assumed as the glomerular filtration and tubular reabsorption. The tubular reabsorption rate of lithium was fixed at 80% of glomerular filtrated lithium [3].
QCR = CLR * RF
QRC = QCR * 0.8
The relationship between serum lithium concentration and occurring neurological symptoms was assessed using exposure-response curves and a logit model.
Results: The final estimated parameters were:
CLR (L/h) = 0.549 * [1 / {10**(pH – 6.2) + 1}] + 0.584 * [10**(pH – 6.2) / {10**(pH – 6.2) + 1}]
CLNR (L/h) = 0.512
CL/F (L/h) = {(CLR * RF) + CLNR} * FSIZE
VC/F (L) = 58.4 * FSIZE
QCR (L/h) = CLR * RF * FSIZE
VR (L) = 23.5 * FSIZE
FSIZEs represent the factor of allometric size scaling [5]. The population mean values of CLION and CLMID were almost same. This indicated that lithium excretion was delayed under acidic concentrations with a high proportion of Li+. The final estimated values were well similar to the averages of the result from bootstrap samples, and were within the bootstrap 95 percentiles.
The serum lithium concentration was significantly related to occurring the neurological symptoms. The logit model demonstrated that the serum lithium concentration corresponding to 30% probabilities of developing neurological symptoms was calculated as 1.26 mEq/L.
Conclusions: We developed the population pharmacokinetic model for lithium carbonate considering the process of glomerular filtration and tubular reabsorption. This model suggested that lithium clearance was variable urinary pH. The neurological symptoms such as tremor correlated with serum lithium concentration, and the therapeutic index to prevent the development of such symptoms was presented.
References:
[1] Yildiz A, Vieta E, Leucht S, Baldessarini RJ. Efficacy of antimanic treatments: meta-analysis of randomized, controlled trials. Neuropsychopharmacology. 2011;36(2):375-89.
[2] Schou M, Amdisen A, Trap-Jensen J. Lithium poisoning. Am J Psychiatry. 1968;125(4):520-7.
[3] Thomsen K, Olesen O. Precipitating factors and renal mechanisms in lithium intoxication. Gen Pharmacol. 1978;9(2):85-9.
[4] Cockcroft DW, Gault MH. Prediction of Creatinine Clearance from Serum Creatinine. Nephron. 1976;16:31-41.
[5] Holford NHG, Anderson BJ. Allometric size: The scientific theory and extension to normal fat mass. Eur J Pharm Sci. 2017;109S:S59-S64.
Reference: PAGE () Abstr 9238 [www.page-meeting.org/?abstract=9238]
Poster: Drug/Disease Modelling - Other Topics