Brief Article pubs.acs.org/molecularpharmaceutics
Saliva versus Plasma Pharmacokinetics: Theory and Application of a Salivary Excretion Classification System Nasir Idkaidek* and Tawfiq Arafat College of Pharmacy and Jordan Center for Pharmaceutical Research, University of Petra, Amman, Jordan ABSTRACT: The aims of this work were to study pharmacokinetics of randomly selected drugs in plasma and saliva samples in healthy human volunteers, and to introduce a Salivary Excretion Classification System. Saliva and plasma samples were collected for 3−5 half-life values of sitagliptin, cinacalcet, metformin, montelukast, tolterodine, hydrochlorothiazide (HCT), lornoxicam, azithromycin, diacerhein, rosuvastatin, cloxacillin, losartan and tamsulosin after oral dosing. Saliva and plasma pharmacokinetic parameters were calculated by noncompartmental analysis using the Kinetica program. Effective intestinal permeability (Peff) values were estimated by the Nelder−Mead algorithm of the Parameter Estimation module using the SimCYP program. Peff values were optimized to predict the actual average plasma profile of each drug. All other physicochemical factors were kept constant during the minimization processes. Sitagliptin, cinacalcet, metformin, tolterodine, HCT, azithromycin, rosuvastatin and cloxacillin had salivary excretion with correlation coefficients of 0.59−0.99 between saliva and plasma concentrations. On the other hand, montelukast, lornoxicam, diacerhein, losartan and tamsulosin showed no salivary excretion. Estimated Peff ranged 0.16−44.16 × 10−4 cm/s, while reported fraction unbound to plasma proteins (fu) ranged 0.01−0.99 for the drugs under investigation. Saliva/plasma concentrations ratios ranged 0.11−13.4, in agreement with drug protein binding and permeability. A Salivary Excretion Classification System (SECS) was suggested based on drug high (H)/low (L) permeability and high (H)/low (L) fraction unbound to plasma proteins, which classifies drugs into 4 classes. Drugs that fall into class I (H/H), II (L/H) or III (H/L) are subjected to salivary excretion, while those falling into class IV (L/ L) are not. Additional data from literature was also analyzed, and all results were in agreement with the suggested SECS. Moreover, a polynomial relationship with correlation coefficient of 0.99 is obtained between S* and C*, where S* and C* are saliva and concentration dimensionless numbers respectively. The proposed Salivary Excretion Classification System (SECS) can be used as a guide for drug salivary excretion. Future work is planned to test these initial findings, and demonstrate SECS robustness across a range of carefully selected (based on physicochemical properties) drugs that fall into classes I, II or III. KEYWORDS: Salivary Excretion Classification System, protein binding, effective permeability, bioequivalence, pharmacokinetics, SimCYP
1. INTRODUCTION Salivary excretion of some drugs has been reported previously as a good indicator for drug bioavailability, therapeutic drug monitoring, pharmacokinetics and also drug abuse. Saliva sampling offers a simple, noninvasive and cheap method as compared with plasma sampling with no contamination risk. Two structures in the oral cavity participate in the formation of saliva and the exchange of products between blood and whole saliva; these are the salivary glands and the oral mucosa. The salivary glands are responsible for the production of the largest portion of saliva with the formation of enzymes, mucin, and other components of saliva; the exchange of inorganic ions with blood; and the excretion of water which is obtained from the complex capillary system surrounding the salivary glands. The oral mucosa participates in the formation of whole saliva by the passage of some blood components from the capillaries of the oral mucosa into the saliva bathing these tissues. Emphasis has been shown on the rule of plasma protein binding on salivary © 2012 American Chemical Society
excretion, and that only unbound drug is available for diffusion into saliva. Polar drugs' saliva/plasma concentration ratios were suggested to depend on saliva pH, yet some deviations from pH partition theory existed.1−8,22 In this research, rules of drug protein binding and membrane permeability were investigated. Studied drugs with different pharmacokinetic, physicochemical and pharmacodynamic properties were selected randomly from ongoing bioequivalence studies in humans. Salivary excretion investigated in this study was for sitagliptin, cinacalcet, metformin, montelukast, tolterodine, hydrochlorothiazide (HCT), lornoxicam, azithromycin, diacerhein, rosuvastatin, cloxacillin, losartan and tamsulosin. Paracetamol, phenytoin, carbamazepine, fluconaReceived: Revised: Accepted: Published: 2358
February 18, 2012 July 1, 2012 July 11, 2012 July 11, 2012 dx.doi.org/10.1021/mp300250r | Mol. Pharmaceutics 2012, 9, 2358−2363
Molecular Pharmaceutics
Brief Article
Table 1. Summary of the Analytical Methods Used in Saliva and Plasma Samples
a
drug
% intraday CV
LOQ
linear range
% interday accuracy range
% interday precision range
cinacalcet sitagliptin diacerheina tolterodine losartan HCT tamsulosin montelukast lornoxicam metformin cloxacillin azithromycin rosuvastatin
3.85 4.70 6.11 6.12 11.18 9.56 2.98 8.19 5.83 2.23 10.96 10.17 14.26
0.2 ng/mL 5 ng/mL 50 ng/mL 75 pg/mL 2.5 ng/mL 2 ng/mL 0.25 ng/mL 5 ng/mL 20 ng/mL 0.06 ng/mL 0.25 μg/mL 1 ng/mL 0.5 ng/mL
0.2−50 ng/mL 5−500 ng/mL 50−8000 ng/mL 75−7500 pg/mL 2.5−500 ng/mL 2−200 ng/mL 0.25−15 ng/mL 5−800 ng/mL 20−1200 ng/m 0.06−3.00 ng/mL 0.25−20.00 μg/mL 1−200 ng/mL 0.5−40 ng/mL
101.96−107.28 95.02−97.84 98.14−102.58 95.78−103.00 96.08−98.53 95.22−99.13 95.15−96.91 95.11−104.27 93.93−103.65 96.15−100.75 96.97−99.25 92.80−100.94 91.76−102.80
2.24−3.32 2.21−4.19 1.01−4.88 0.93−3.04 3.23−11.37 2.01−5.67 3.13−3.78 3.14−7.51 1.70−6.47 1.52−2.53 4.54−8.79 6.97−10.03 2.21−13.41
Rhein was assayed.
Table 2. Saliva and Plasma Mean (% CV) Pharmacokinetic Parameters after Oral Doses to Healthy Volunteers saliva
a
plasma
drug
AUC0→t (ng/mL h)
Cmax (ng/mL)
Tmax (h)
half-life (h)
AUC0→t (ng/mL h)
Cmax (ng/mL)
Tmax (h)
half-life (h)
cinacalcet sitagliptin diacerhin tolterodine losartan HCT tamsulosin montelukast lornoxicam metformin cloxacillin azithromycinb rosuvastatin
23.2(123) 592(57) NSEa 2.47(48) NSE 265(16) NSE NSE NSE 990(38) 31518(111) 5624(49) 23.4(95)
10.3(145) 111(75) NSE 1.11(69) NSE 89.6(70) NSE NSE NSE 166(46) 36620(76) 2263(82) 4.19(94)
2.3(137) 6(0) NSE 1.22(69) NSE 2.39(28) NSE NSE NSE 4.1(68) 0.58(25) 4.33(66) 4.5(40)
2.59(51) 2.98(4) NSE 1.88(27) NSE 5.2(8) NSE NSE NSE 4.2(40) 0.66(79)
97.4(50) 3717(56) 22724(53) 11.75(95) 1405(49) 646(32) 176(46) 2690(44) 2063(35) 8711(14) 17859(24) 1003(36) 279(60)
15.0(29) 585(72) 5748(20) 3.60(50) 610(79) 113(34) 6.3(38) 395(37) 587(24) 1392(18) 14051(27) 134(42) 24.9(56)
1.3(88) 1.5(0) 2.3(53) 0.8(33) 1.75(2) 2.14(34) 5.7(61) 3.2(38) 2.0(37) 1.8(55) 0.58(49) 4.2(28) 4.3(24)
4.65(23) 6.19(23) 4.1(27) 1.98(40) 2.38(18) 3.9(25) 38(16) 4.7(28) 3.9(27) 2.7(6) 1.22(3)
4.3(31)
5(20)
b
NSE stands for no salivary excretion. Azithromycin half-life was not calculated since design was truncated at 72 h.
signing informed consent. Medical history, vital signs, physical examination, and laboratory safety test results showed no evidence of clinically significant deviation from normal medical condition as evaluated by the clinical investigator. 3.2. Assay Methodology. All samples were deep frozen until assayed by validated assay methods as summarized in Table 1. These methods were used for the assay of saliva and plasma samples. 3.3. Data Analysis. 3.3.1. Calculated Pharmacokinetic Parameters. Individual pharmacokinetic parameters for drug concentration in both saliva and plasma samples were calculated by noncompartmental analysis (NCA) using the Kinetica program. Pharmacokinetic parameters were area under the concentration curves to last collection time (AUC0→t), maximum measured concentration (Cmax), time to maximum concentration (Tmax) and half-life. 3.3.2. Optimized Effective Intestinal Permeability. Effective intestinal permeability (Peff) values were estimated by Nelder− Mead algorithm of the Parameter Estimation module using SimCYP program.19 The Nelder−Mead method, which is also called downhill simplex, is a commonly used nonlinear optimization algorithm. This was done by searching for the best parameter values that produce plasma concentration that matches the actual plasma concentration at the same time. The objective function is the weighted sum of squared differences of observed and model predicted values. Polar surface area (PSA)
zole, norfloxacin, erythromycin and ibuprofen salivary excretion has been investigated previously.8,11−18
2. OBJECTIVES The objectives are to investigate the scientific rationale of the salivary excretion of randomly selected drugs based on a suggested Salivary Excretion Classification System (SECS), to compare pharmacokinetic parameters calculated from saliva samples with those from plasma samples, and to correlate saliva and plasma drug concentrations. 3. EXPERIMENTAL SECTION 3.1. Study Designs. Saliva pharmacokinetics under fasted state, in healthy human volunteers was compared to plasma pharmacokinetics in 2 way crossover design studies. Thirteen bioequivalence studies were conducted as per the ICH, GCP, and Helsinki declaration guidelines after IRB of Al-Muwasah Hospital and Jordan FDA approvals. Oral dosing with 240 mL of water was given after 10 h overnight fasting without dietary restrictions. Then resting saliva (without stimulation) and plasma samples were collected up to five half-life values during each study phase. Thorough rinsing of the mouth was done prior to saliva sampling to avoid contamination of saliva samples by any drug residues. Washout time between phases was up to ten half-life values. 2−3 humans were randomly enrolled in each study for both saliva and plasma sampling after 2359
dx.doi.org/10.1021/mp300250r | Mol. Pharmaceutics 2012, 9, 2358−2363
Molecular Pharmaceutics
Brief Article
was used first, using SimCYP, to predict an initial estimate of Peff. All other physicochemical factors used in calculations such as Log P, MW and fu were obtained from the literature and were kept constant during the minimization processes. In vitro dissolution rate, in vivo clearance and volume of distribution were input from actual dissolution and plasma profiles. Fraction absorption (Fa) was calculated according to equations below:
variability values (% CV) were high in both saliva and plasma due to small sample size used in each drug.
Fa = 1 − e−2An An = Peff tres/R
where An is the absorption number, R and tres are radius, set at 1.75 cm, and mean residence time, set at 3 h, in the human small intestine respectively.20 3.3.3. Dimensional and Statistical Analysis. Saliva/plasma correlation, dimensional analysis and descriptive statistics of concentrations and pharmacokinetic parameters were done using Microsoft Excel program for comparison purposes. The following dimensionless ratios were used for comparisons: AUC* = saliva AUCt/plasma AUCt Tmax* = saliva Tmax/plasma Tmax Cmax* = saliva Cmax/plasma Cmax t0.5* = saliva t0.5/plasma t0.5 C* = saliva concentration/plasma concentration = Cs/Cp Peff* = dimensionless effective permeability = RPeff/D, where D is drug diffusivity as predicted by SimCYP S* = saliva dimensionless number = (fu)Peff*
Figure 1. Plasma and saliva of cloxacillin mean profiles.
Figure 2. Correlation of saliva (Cs) and plasma (Cp) concentrations of cloxacillin.
Dimensional analysis results are shown in Table 3. They indicated lower bioavailability in saliva for most drugs. This can be related to the effect of protein binding, which was suggested before with salicylates where higher protein binding led to lower saliva/plasma ratios.3 Azithromycin had markedly higher saliva profiles than plasma profiles, suggesting high salivary distribution. This is in agreement with azithromycin monograph, where the drug concentrations in tissues such as tonsils can reach 60-fold that of plasma. On the other hand, Tmax was longer in saliva for most drugs suggesting a lag time between plasma and saliva compartments due to drug distribution/ redistribution processes in the body. Moreover, AUC* and Cmax* values were in close agreement with C* values. However, t0.5* were either low or high, indicating variable elimination rates from plasma and saliva. Plasma mean concentration profiles were used to estimate the effective intestinal permeability values as shown in Table 5.
4. RESULTS AND DISCUSSION Pharmacokinetic parameters in saliva and plasma for all of the 13 drugs are summarized in Table 2. Montelukast, tamsulusin, losartan, diacerhein and lornoxicam were not excreted in saliva. However, sitagliptin, cinacalcet, metformin, tolterodine, HCT, azithromycin, rosuvastatin and cloxacillin had salivary excretion with correlations of 0.59−0.99 between saliva and plasma concentrations up to median Tmax values of plasma profiles as shown in Table 3. Assuming one compartment linear model, Table 3. Dimensional Analysis Values of Pharmacokinetic Parameters
a
drug
AUC*
Cmax*
Tmax*
t0.5*
C*
Ra
cinacalcet sitagliptin tolterodine HCT metformin cloxacillin azithromycin rosuvastatin
0.24 0.16 0.21 0.41 0.11 1.76 5.61 0.08
0.69 0.19 0.31 0.79 0.12 2.61 16.89 0.17
1.56 4.00 1.53 1.12 2.23 1.00 1.03 1.05
0.56 0.48 0.95 1.35 1.58 0.54
0.32 0.14 0.42 0.20 0.11 1.24 13.4 0.13
0.59 0.99 0.99 0.83 0.87 0.99 0.99 0.89
0.86
Table 4. Salivary Excretion Classification System (SECS) According to Drug Permeability (Peff) and Fraction Unbound to Plasma Proteins (fu) class class class class
Correlation coefficient of Cs versus Cp up to median Tmax of Cp.
I II III IV
Peff
fu
salivary excretion
high low high low
high high low low
yes yes yes no
Figures 3 and 4 show observed versus SimCYP-predicted plasma concentration profiles for cloxacillin with correlation coefficient of 0.93 indicating good fitting of observed plasma concentrations. This suggests that permeability and protein binding are major key factors in salivary excretion, assuming that intestinal permeability is similar to salivary mucosal permeability. Similar
salivary excretion rate is dependent on plasma drug concentration. This explains the close behavior of the saliva and plasma profiles. Salivary concentrations showed good correlation with plasma concentrations up to median Tmax values of plasma data. This result is in agreement with previous studies done on paracetamol and caffeine.11 Representative profiles of cloxacillin are shown in Figures 1 and 2. Intersubject 2360
dx.doi.org/10.1021/mp300250r | Mol. Pharmaceutics 2012, 9, 2358−2363
Molecular Pharmaceutics
Brief Article
intestinal permeability corresponds to fraction absorption Fa > 0.9, while high protein binding corresponds to low fraction unbound fu of
