INTRODUCTION:
Description of levothyroxine
Levothyroxine sodium is synthetic tetraiodothyronine sodium salt [levothyroxine Sodium (T4)]. The T4 produced in the human thyroid gland and Synthetic T4 is identical.2 Significance of Thyroid Hormone synthesis.
Thyroid hormones are vitals hormones required by the human body, deficiency of these hormones in the body requires specific treatment. In practice, levothyroxine is used to treat thyroid hormone deficiency, anddose optimisation or individualisation of Levothyroxine will be beneficial for the clinician for optimal treatment. Many metabolic processes are regulated by these hormones and are important forgrowth and development.2
Clinical pharmacology:
Synthetically available Levothyroxine (T4) is a replacement therapy for thyroid hormone that is normally produced by the thyroid gland to regulate the energy and metabolism of the body.
Pharmacokinetics:
Absorption and Distribution:
Orally administered Levothyroxine has absorption of 40% to 80%which is from the gastrointestinal tract. The absorption of levothyroxine is maximum at jejunam and upper ileum.Very minimal absorption in stomach. Hence higher doses are required for patients with shorter small intestine as the absorption will be poor. Fasting condition increases the Levothyroxine absorption, and decreased in malabsorption condition. Absorption is also reduced with age.1,2
The thyroid hormones to a great extent (approx. 99%)are protein bound. These plasma proteins includesalbumin (TBA), thyroxine-binding globulin (TBG), thyroxine-binding prealbumin (TBPA)and has different capacities and affinities to these plasma protein. There is a reverse equilibrium between Protein-bound thyroid hormones with small amounts of free hormone.1,2,7
Metabolism and Elimination:
The elimination of Levothyroxine is followed slowly. The Levothyroxine is majorly metabolized through sequential Deiodination process. Through the Monodeiodination process the peripheral levothyroxine is converted to T3 which amounts to almost 80 percent of the circulating T3. Liver is the major site of metabolism for T4 and T3, with also at additional site such as kidney and other tissues through the Deiodination process. Thyroid hormones also undergo glucoronide and sulfate conjugation and also undergoes enterohepatic recirculation when they are released in bile and intestine.1,2,7
The thyroid hormones are primarily eliminated through kidneys. Nearly a small part of conjugated thyroid hormone released unchanged into the intestine is eliminated through the feces. Levothyroxine excretion through urinary system decreases with age.1,2,7
Population Pharmacokinetics:
Population pharmacokinetics is the study of the sources and correlates of variability in drug concentrations among individuals who are the target patient population receiving clinically relevant doses of a drug of interest. Population modeling seeks to evaluate the inter individual and intraindividuality based on raw subject data and the assay error, and to describe findings in terms that useful both for research and for optimal patient care.
Certain patient demographical, pathophysiological, and therapeutical features, such as body weight, excretory and metabolic functions, and the presence of other therapies, can regularly alter dose-concentration relationships.
MATERIALS AND METHODS:
Population Pharmacokinetic modelling with Non Linear mixed effects modelling approach using Phoenix® NLMETM (Version 7.0, Certara USA Inc, Princeton, NJ)
Data:
The levothyroxine population pharmacokinetic (PPK) model was developed using data from 90 patients who had differentiated thyroid cancer who underwent complete thyroidectomy followed by radio Iodine [131I] ablation treatment and in patients with thyrotoxicosis in treatment with radio Iodine [131I]. Institutional ethics approval was taken and informed consent was taken from each patient prior to their enrollment into the study. The sampling time points were 0 h, 1h, 2h, 3h, 4h, 5h, 6h, 8h, 10h, 12h, 14h and 24h. A sparse sampling strategy was used with each patient having contributed three samples (a predose plus any two other time points) taken between a dosage interval at steady state.
Population pharmacokinetic model development:
The PPK model was developed in three stages, consisting of a base, full and final model. First, base model was developed to describe the PK of Levothyroxine without consideration of covariate effects. Second, a full covariate model was developed by incorporating the effect of all prespecified covariate parameter relationships in a step-wise manner. In the third stage, the final PPK model was developed by retaining covariates that improved a goodness-of-fit statistic (log likelihood).
The first step of PPK analysis was base model development, which consisted of the development of structural model, inter-individual variability (IIV) model and a residual error model. A one compartment extravascular model with first order elimination was used. Model was parameterized in terms of Clearance (CL), volume of distribution (V) and absorption rate constant (Ka). Interindividual variability (IIV) in CL and Ka was modeled assuming a log-normal distribution (Eq. 1 and 2). Volume of distribution (V) was fixed at 14 L.
Ka = tvKa * exp(ηKa)……1
Cl= tvCl * exp(ηCl)……….2
Where Ka and V is the estimate for a PK parameter in
an individual as predicted in the model without covariate effects and tvKa and
tvCl is the typical value of the population PK parameter. ηKa and ηCl
represents the random variable (for IIV) for Ka and CL respectively, with mean
0 and variance 
. Residual variability was modeled using a
proportional error model (Eq. 3)
Cobs= C * (1+CEps)
Where Ceps is a random variable (for residual error)
with mean 0 and variance of
.
The full model was developed to obtain unbiased estimates of the magnitude of covariate effects on the model parameters. Covariates were included if the change in objective function value (dOFV) was larger than 3.84, which corresponds to a statistical significance of p < 0.05, based on a Chi-squared distribution (df= 1). For each addition of a covariate to the base model, the improvement in fit was assessed. Covariates such as gender, age and weight, TSH values, T3 concentration were evaluated for their effect on CL and V graphically. Based on this, only Weight on CL was formally added to the base model subsequently in a step wise forward additive manner (Figure 1).
Figure 1. Population covariate scatter plot of weight on clearance
The first-order conditional estimation Extended Least Squares (FOCE ELS) algorithm was used throughout the modeling process implemented in Phoenix® NLMETM (Version 7.0, Certara USA Inc, Princeton, NJ).
Model evaluation:
Model evaluation was performed using standard goodness-of-fit plots including population predicted concentrations (PRED) versus observed concentrations (DV), individual predicted concentrations (IPRED) versus DV, and time after dosing versus weighted residuals (WRES). Visual predictive check (VPC) was done to provide an evaluation of model assumptions and population parameter estimates. A total of 500 replicates were simulated using the final model to simulate expected concentrations. The check was performed by plotting the 5th, 50th and 95% percentiles of observed plasma concentration-time data with corresponding 5th, 50th and 95th prediction intervals. The observed data were overlaid on the prediction intervals and compared visually.
Bootstrapping was also performed on 500 re-sampled datasets from the original dataset.
RESULTS:
Pharmacokinetic analysis of Levothyroxine:
Average plasma concentration- time profile is shown in Figure 2. The one compartment model described the data well. Weight was the only covariate that affected the Clearance of Levothyroxine.. The population estimates along with the bootstrap estimates are given in Table 1.
The plots of PRED vs DV and IPRED vs DV demonstrate goodness-of-fit. The random distribution in the time after dosing vs WRES plot reiterates good fitting. The model thus developed without any covariate effects was able to adequately describe the observed concentrations. Further, the VPC plot demonstrates that most of the observed data lie within 95 % of the predicted quantiles which lends credence to the model.