Pharmacokinetics of drugs in mutant Nagase analbuminemic rats and responses to select diuretics

Authors

  • Joo Hyun Lee,

    1. College of Pharmacy, Kyung Hee University, Seoul, South Korea
    Search for more papers by this author
  • Young-Joo Lee,

    1. College of Pharmacy, Kyung Hee University, Seoul, South Korea
    Search for more papers by this author
  • Euichaul Oh

    Corresponding author
    1. College of Pharmacy and Integrated Research Institute of Pharmaceutical Sciences, The Catholic University of Korea, Bucheon, South Korea
    • Correspondence

      Euichaul Oh, College of Pharmacy and Integrated Research Institute of Pharmaceutical Sciences, The Catholic University of Korea, 43 Jibong-ro, Wonmi-gu, Bucheon, Gyeonggi-do 420-743, South Korea.

      E-mail: eoh@catholic.ac.kr

    Search for more papers by this author

Abstract

Objectives

To report (1) the pharmacokinetics of drugs that are mainly metabolized via hepatic cytochrome P450s (CYPs) or mainly excreted via the urine and bile, (2) the mechanism for the urinary excretion of drugs (such as glomerular filtration or renal active secretion or re-absorption), and (3) the diuretic effect of some loop diuretics in mutant Nagase analbuminaemic rats (NARs), an animal model for human familial analbuminaemia based on the pharmacokinetics of drugs reported in the literatures.

Key findings

In NARs, the changes in the time-averaged non-renal clearances (CLNRs) of drugs that are mainly metabolized via CYPs were explained in terms of changes in the hepatic intrinsic clearance (mainly because of changes in CYPs), free (unbound) fractions of drugs in the plasma (fp) and hepatic blood-flow rate (QH) depending on the hepatic excretion ratios of drugs.

Summary

The CLNR changes of drugs mainly metabolized via hepatic CYPs can be sufficiently explained by the three earlier mentioned factors. The plasma albumin (furosemide) or globulin (azosemide, bumetanide and torasemide) binding affects their diuretic effects.

Introduction

Nagase et al.[1] derived mutant Nagase analbuminaemic rats (NARs) from Sprague–Dawley (SD) rats as an animal model for human familial analbuminaemia. NARs develop not only analbuminaemia but also hyperlipidaemia, which greatly increases their serum cholesterol and triglyceride concentrations.[1] In NARs, the serum concentrations of albumin and globulins are lower and higher, respectively, than those in control rats. For example, the serum concentrations of albumin, and α-, β- and γ-globulins were measured at 3.73%, 1.03%, 1.07% and 1.05%, respectively, in control SD rats;[2] the corresponding values in NARs were measured at approximately 0.0042%, 3.1%, 1.8% and 1.3%, respectively.[3] Interestingly, in NARs, the serum levels of albumin can be restored to control levels by hepatocyte transplantation.[4]

The pharmacokinetic processes of drugs may be altered in NARs. The factors affecting the gastrointestinal (GI) absorption of drugs in NARs have not been thoroughly studied. The distribution of drugs could be altered in NARs by changes in the plasma protein binding of drugs because of the changes in the plasma concentrations of albumin and globulins.[3] The plasma volume was larger in NARs than that in the control rats.[5] In NARs, the effects of hematocrit changes on the distribution of drugs that are mainly bound to the blood cells seemed to be almost negligible because the hematocrit level was comparable between the two types of rats.[5] The metabolism of drugs with low or intermediate hepatic extraction ratios (HERs), which are mainly metabolized via hepatic cytochrome P450s (CYPs), may be altered in NARs by changes in the protein expression of CYPs in the liver (Table 1). In NARs, liver functions are comparable with those in control rats.[1, 6, 9] Protein expression of P-glycoprotein, which could affect the GI absorption as well as hepatic uptake (resultant hepatic metabolism) of a drug, is comparable between the control rats and NARs.[10] The renal excretion of drugs could be altered in NARs by changes in the plasma-protein binding.[3] Urinary pH- and/or urine output (volume)-dependent time-averaged renal clearances (CLRs) of drugs have been reported.[11] In NARs, however, this could be almost negligible because the urinary pH or 24-h urine output (except furosemide[12] and azosemide[6]) appears to be comparable between the two types of rats. In NARs, kidney functions are also comparable with those in control rats.[6, 9]

Table 1. Changes in the protein expression of CYPs (using Western blot analysis) in the liver and intestine of NARs compared with those in control rats
CYP isozymeLiverCYP isozymeIntestine
  1. ↑, increased; ↓, decreased; ↔, not altered; CYP, cytochrome P450; NAR, Nagase analbuminemic rat.
CYP1A2↑ 250%[6]CYP1A subfamily↑ 100%[7]
CYP2B1/2↓ 75.5%[6]CYP3A subfamily↑ 110%[8]
CYP2C11↔[7]  
CYP2D subfamily↔[8]  
CYP2E1↔[6]  
CYP3A1↔[6]  

Hepatic clearance (CLH) is the efficiency of the liver to irreversibly remove a drug from the perfusing blood.[13] Based on the well-stirred model of the liver, under the assumptions of perfusion-limited clearance and apparent first-order conditions, the HER and CLH of a drug could be calculated by the following Equation (1):[13-16]

display math(1)

where CLint and math formula represent the hepatic intrinsic total drug and free drug clearances, respectively. The math formula is CLint.

The hepatic CLint is used to indicate the maximal ability of the liver to irreversibly remove a drug in the absence of any flow limitation[13] and is calculated as the maximum velocity divided by the apparent Michaelis–Menten constant (Km) of the drug.

When a drug is mainly metabolized in the liver, the CLNR of the drug could represent its CLH. For a low-HER drug (math formula), the numerator, math formula, in Equation (1) is close to QH. Thus, for a drug with a low HER (<0.3), the CLH depends more on the CLint than the QH.[13, 16] For a drug with a high HER (math formula), the CLH of a drug depends more on the QH than the hepatic CLint of the drug.[13, 16] For a drug with an intermediate HER (0.3–0.7), the CLH depends on hepatic CLint, fp and QH.[13, 16] In this review, the hepatic CLint was calculated from an in vitro experiment. Because the fp was not considered in the in vitro experiment, it could affect the CLH of a drug target with a high (or an intermediate) HER. Changes in fp could also affect the CLH of a drug with a low HER, if the protein binding of the drug in control rats is extensive (thus, the fp in NARs is considerably greater than that in control rats). Thus, the changes in CYPs (CLint) could have the most profound influence on a drug target with a low (or an intermediate) HER.

In this review, the changes in time-averaged non-renal clearances (CLNRs; which could represent metabolic clearances) of drugs that are mainly metabolized via hepatic CYPs in NARs were explained, according to three factors: in vitro hepatic intrinsic clearances (CLints) for the disappearance of drugs primarily because of changes in the CYPs, free fractions of drugs in the plasma (fps) and hepatic blood-flow rates (QHs) depending on the HERs of drugs.[13-16]

The HER (the hepatic first-pass effect) of a drug can be directly measured by comparing the area under the plasma or serum concentration–time curve (AUC) values following intravenous (i.v.) and intraportal administration of a drug. Alternatively, the ratio may be indirectly estimated by dividing the CLNR (assuming that it equals CLH) of a drug after i.v. administration by the hepatic plasma flow rate.[17] The hepatic plasma flow rate can be estimated by multiplying QH of 55.2 ml/min/kg[18] by (1 – hematocrit), in which the hematocrit level was approximately 0.45[2] in rats. The HER could also be measured by an isolated perfused rat liver study.[15] If the CLNR of a drug was not available, the time-averaged total body clearance (CL) of a drug was compared, in which the CL could represent its CLNR.

After oral administration of both low- and high-HER drugs that are cleared by the liver, the oral AUC (AUCoral) or CL/F (where F represents the extent of absolute oral bioavailability) depends on various factors: the fraction of administered drug that is absorbed into the gut wall (fabs), the fraction that gets through the gut wall unchanged (fG), dose, fp and math formula, as shown in Equation (2):[14]

display math(2)

After i.v. administration of drugs, the fabs and fG do not need to be considered. Thus, the CLH of a drug depends on the math formula and QH, which are simpler and easier terms for explaining changes in the CLH than those for oral administration. Thus, in this review, the pharmacokinetic parameters of drugs in NARs after i.v. administration were compared with those for control rats.

This paper reports the following for the first time. (1) The changes in CLNR of a drug that is mainly metabolized by hepatic CYPs in NARs were compared with that in control rats. The reason for changes in CLNR was also explained. (2) Pharmacokinetic changes of drugs in NARs were also reviewed, even if their metabolisms were relatively unaffected by changes in hepatic CYPs, such as with respect to drugs primarily excreted via the urine or bile. (3) In NARs, the changes in the mechanism of the renal excretion of drugs, such as those that are mainly excreted via glomerular filtration or active tubular secretion or re-absorption, have not been thoroughly studied.[19] Therefore, the issue was also explored, if the plasma (serum) protein-binding values of drugs in control rats remain constant over a wide range of drug concentrations. (4) The effect of globulin binding of a diuretic on its diuretic effect in NARs and patients with nephrotic syndrome were compared.

Drugs that are altered pharmacokinetically in Nagase analbuminaemic rats

Only the CLNR (or CL), and urinary or biliary excretions of a drug primarily metabolized via hepatic CYPs and excreted in the urine or bile, respectively, were compared between the control rats and NARs.

Table 2 lists the hepatic CYPs involved in the metabolism of each drug primarily metabolized via hepatic CYPs and the corresponding pharmacokinetic observations (changes in CLNR) and the urinary or biliary excretion of drugs in NARs. Table 3 lists the plasma (serum) protein binding, urinary excretions up to time, t (Ae0–t), CL, CLNR, CLR, CLR based on the free fraction (CLR,fp) of drugs and the apparent volume of distribution at steady state (Vss) of some drugs.

Table 2. Pharmacokinetic observations of drugs in NARs compared with control rats
No.Drug (i.v. dose)Hepatic CYPs for drugs mainly metabolized via CYPsPharmacokinetic observation
  1. 5-FU, 5-fluorouracil; CL, time-averaged total body clearance;;CLNR, time-averaged non-renal clearance; CYP, cytochrome P450; i.v., intravenous; MTX, methotrexate; NAR, Nagase analbuminemic rat; PSP, phenolsulfonphthalein.
Drugs altered pharmacokinetically in NARs  
 1Furosemide (2 mg/kg[12])CYP2C11, 2E1, 3A1 and 3A2[20]Faster CLNR (by 226%)[12]
 2Azosemide (10 mg/kg[6])CYP1A1/2[21]Significantly faster CLNR (by 307%)[6]
 3Bumetanide (10 mg/kg[22])CYP2B1/2[23]Significantly faster CLNR (by 124%)[22]
 4Torasemide (10 mg/kg[24])CYP2C11[25, 26]Significantly faster CLNR (by 750%)[24]
 5Theophylline (5 mg/kg[19])CYP1A1/2, 2B1/2 and 3A1/2[24, 27]Significantly faster CLNR (by 49.3%)[19]
 6Omeprazole (20 mg/kg[8])CYP1A1/2, 2D1 and 3A1/2[28]Significantly faster CLNR (by 136%)[8]
 7Oltipraz (10 mg/kg[29])CYP1A1/2, 2B1/2, 2C11, 2D1 and 3A1/2[30]Significantly faster CL (by 125%)[29]
 8Phenytoin (10 mg/kg[31])CYP2C6 and less via 2C11 for both metabolism of phenytoin and formation of 4'-HPPH[32]Significantly faster plasma CL (by 99.4%) but comparable blood CL[31]
 9Warfarin (1 and 40 mg/kg racemic [14C]warfarin[33])CYP1A1, 2B1, 2C6, 2C11 and 3A2[34]Significantly faster CL (by 5980 and 2070% for 1- and 40-mg/kg, respectively)[33]
105-FU (30 mg/kg[35])CYP1A1/2[36]Significantly faster CLNR (by 78.1%)[35]
11Gliclazide (30 mg/kg in NARs[37])CYP2C and 3A less extent subfamily[38]Significantly faster CL (by 659%)[37]
12

Bilirubin

Trace amount, 3.6 nmol/kg [3H]bilirubin[39]

Loading amount, 0.5 μmol/kg [3H]bilirubin[39]

 

Comparable 1-h biliary excretion of total radioactivity (primarily monoglucuronide and diglucuronide[39]

Smaller (by 37.0%) 1-h biliary excretion of total radioactivity (primarily monoglucuronide and diglucuronide[39]

13

PSP

(10 μmol/kg[40])

(10 μmol/kg[41])

 

Smaller Ae0–6 h (by 54.5%)[40]

Greater 1-h biliary excretion (by 48.2%) and smaller Ae0–6 h (by 55.1%)[41]

14Mercapturic acid (10 μmol/kg S-benzyl-N-acetyl-L-[U-[14C]cysteine[42]) Smaller Ae0–30 min of total radioactivity (by 54.8%)[42]
15BSP (5 μmol/kg[43]) Significantly smaller 1-h biliary excretions of both free and glutathione conjugate (by 29.3%)[43]
16Taurocholic acid (8 μmol/kg of [3H]taurocholic acid[44]) Significantly greater 1-h biliary excretion (by 21.5%)[44]
Drugs not altered pharmacokinetically in NARs  
 1Ipriflavone (20 mg/kg[7])CYP1A1/2 and 2C11[45]Comparable CLNR[7]
 2Micafungin (1 mg/kg in NARs[10])No dataComparable CL[10]
 3MTX (100 mg/kg[46])Not metabolized via CYPs[46]Comparable CLNR[46]
Table 3. The mean plasma (serum) protein-binding values (PPB), time-averaged total body clearance (CL), percentage of the dose excreted in the urine up to time t (Ae0–t), time-averaged renal clearances from total (CLR) and free (CLR,fp) fractions and apparent volume of distribution at steady state (Vss) of drugs in control rats (C) and Nagase analbuminemic rat (N)
No.Drug (intravenous dose)PPB (%)CL (ml/min per kg)Ae0–t (% of the dose)CLNR (ml/min per kg)CLR (ml/min per kg)CLR,fp (ml/min per kg)Vss (ml/kg)Ref.
CNCNCNCNCNCNCN
  1. 5-FU, 5-fluorouracil; MTX, methotrexate; PSP, phenolsulfonphthalein. *Significantly different (P < 0.05) from controls.
1Furosemide98.912.09.424.4*267*6.9622.72.441.712221.9459.6545.5*[12]
(2 mg/kg)(at 20 μg/ml) (0–60 min)        
2Azosemide97.984.6*3.5919.8*21.037.7*3.0512.4*0.7727.36*36.847.8234496*[6]
(10 mg/kg)(at 1 μg/ml) (0–8 h)(0–8 h)     
3Bumetanide97.636.8*4.059.88*7.9216.0*3.718.30*0.2751.49*11.52.3623.333.9[22]
(10 mg/kg)(at 10 μg/ml) (0–6 h)     
4Torasemide94.123.3*3.7433.2*10.714.9*3.3328.3*0.3864.81*6.546.27225176[24]
(10 mg/kg)(at 10 μg/ml) (0–8 h)      
5Theophylline26.912.0*2.864.81*22.130.1*2.253.36*0.5711.39*0.7811.58523860*[19]
(5 mg/kg)(at 5 μg/ml) (0–24 h)      
6Omeprazole81.560.7*46.8110*0.3850.265*46.6110*0.1830.291*0.9890.740406817[8]
(20 mg/kg)(at 10 μg/ml)  (0–24 h)        
7Oltipraz80.341.4*45.7103*    0.01170.02790.05940.0476a37607860*[29]
(10 mg/kg)(at 5 μg/ml)            
8Phenytoin92.860.3*15.731.3*          [31]
(plasma data) 
(10 mg/kg)(at 7.48 μg/ml)(at 2.21 μg/ml)18.315.0           
(blood data) 
9Warfarin               
(1 mg/kg)98.963.6*0.21212.9*1.64.6*  0.20335.5    [33]
 (at 5.8 μg/ml)(at 0.41 μg/ml)  (0–24 h)          
(40 mg/kg)92.738.1*0.46510.1*          [33]
 (at 43.5 μg/ml)(at 236 μg/ml)             
105-FU32.226.0*32.855.5*12.77.55*28.851.3*3.793.895.595.26191352*[35]
(30 mg/kg)(at 5 μg/ml)  (0–24 h)          
11Gliclazide89.610.5*1.057.97*        154514*[37]
(30 mg/kg)               
12PSP               
(10 μmol/kg)94.432.04.753.9*49.222.4      49.1179.1[40]
(at 0.1 mmol/l)  (0–6 h)          
(10 μmol/kg)84.212.51.37.0*30.713.8*      33.0112*[41]
 (at 0.5 mmol/l)        
13Mercapturic acid80.118.425.245.7*60.427.3     [42]
(10 μmol/kg)(at 0.5 mm) (0–30 min)      
14Ipriflavone98.798.725.122.70.04260.059025.122.70.01000.01360.7691.0536502970[7]
(20 mg/kg)(at 5 μg/ml) (0–24 h)      
15MTX32.041.4*10.59.1736.026.7*6.686.633.712.37*5.464.04161110[46]
(100 mg/kg)(at 10 μg/ml)  (0–24 h)        

Furosemide (frusemide; a low-hepatic extraction ratio drug, direct[47] and indirect[17] hepatic extraction ratios ∼0.1)

In male SD rats, furosemide, an inhibitor of the Na+–K+–2Cl symport (loop or high-ceiling diuretic), is metabolized via hepatic CYP2C11, 2E1, 3A1 and 3A2 (but not CYP1A1/2, 2B1/2, 2C6 or 2D1).[20] After i.v. administration of furosemide to NARs, its plasma CLNR (which could represent its metabolic clearance[48]) can be roughly estimated by subtracting CLR from CL.[12] The CLR of furosemide can also be roughly estimated from the CL and Ae0–60 min of furosemide.[12] In NARs, the estimated CLNR was more rapid (by 226%) than that in the control rats (Table 3). This could be due to a greater (by 7900%) fp than that in the control rats (Table 3) because the protein expression of hepatic CYP2C11, 2E1 and 3A1 is not altered in NARs (hepatic CLint for the disappearance of furosemide is not available) (Table 1).

The protein-binding values of furosemide to human serum remain constant (97.4–97.6%) at furosemide concentrations ranging from 1 to 36 μg/ml.[49] Thus, the CLR,fps of furosemide can be roughly estimated from CL, Ae0–60 min and plasma protein-binding value of furosemide (Table 3). The estimated CLR,fp values of furosemide in control rats and NARs (Table 3) were faster and slower, respectively, than the creatinine clearances (CLcrs) of 4.35 and 4.38 ml/min/kg for control rats and NARs, respectively.[6] This suggests that furosemide is mainly excreted in the urine via active renal tubular secretion and re-absorption for control rats and NARs, respectively. The plasma protein-binding value of furosemide in NARs is only 12% (Table 3), suggesting that plasma globulin binding of furosemide in NARs seemed to be almost negligible.

Azosemide (a low-hepatic extraction ratio drug, direct hepatic extraction ratio ∼0.2[50])

In male SD rats, azosemide, a loop diuretic, is metabolized via hepatic CYP1A1/2[21] but not CYP2E1.[51] After i.v. administration of azosemide to male NARs, its plasma CLNR (which could represent its metabolic clearance[50]) was more rapid (by 307%) than that in control rats[6] (Table 3). This could be due to a greater (by 633%) fp than that in the control rats (Table 3). The CLint for the disappearance of azosemide was not available (the protein expression of hepatic CYP1A2 is increased in NARs; Table 1).

Based on the CLR,fp (Table 3), azosemide is primarily excreted in the urine via active renal tubular secretion for both types of rats. The protein-binding values of azosemide to 4% human serum albumin (HSA), similar to the ratio of albumin in rat plasma,[2] remain constant (95.5%) at azosemide concentrations ranging from 5 to 100 μg/ml.[52] Plasma protein-binding value of azosemide in NARs (84.6%) could be due to the binding to 3.1% α-globulin (82.6%), 1.8% β- plus 0.63% γ-globulin (68.9%), and 3.1% α- and 1.8% β- plus 0.63% γ-globulins (83.1%) because the value for 0.0042% HSA was only 10.2%.[6]

Bumetanide (a low-hepatic extraction ratio drug, direct hepatic extraction ratio almost negligible[53])

In male SD rats, bumetanide, a loop diuretic, is metabolized via hepatic CYP2B1/2[23] but not CYP2E1.[51] After i.v. administration of bumetanide to male NARs, its plasma CLNR (which could represent its metabolic clearance[54]) was more rapid (by 124%) than that in control rats[22] (Table 3). This could be due to the greater (by 2530%) fp than that in control rats (Table 3) because the protein expression of hepatic CYP2B1/2 is decreased in NARs (hepatic CLint for the disappearance of bumetanide is not available) (Table 1).

Based on the CLR,fp (Table 3), bumetanide is excreted in the urine via active renal tubular secretion and re-absorption for the control rats and NARs, respectively. The protein-binding values of bumetanide to 4% HSA remain constant (87.5%) at bumetanide concentrations ranging from 0.1 to 100 μg/ml.[55] The plasma protein-binding value of bumetanide in NARs (36.8%) could be due to binding to the 3.1% α-globulin (38.8%) and 1.8% β- plus 0.63% γ-globulins (34.6%) because the value for 0.0042% HSA was only 4.94%.[22]

Torasemide (torsemide; a low-hepatic extraction ratio drug, direct hepatic extraction ratio 0.03–0.04[56])

Torasemide, a loop diuretic, is primarily metabolized via hepatic CYP2C11 (not via CYP1A1/2, 2B1/2 or 2E1) in male SD rats.[25, 26] After i.v. administration of torasemide to male NARs, its plasma CLNR (which could represent its metabolic clearance[57]) was more rapid (by 750%) than that in control rats[24] (Table 3). This could be due to the greater (by 1200%) fp than that in control rats (Table 3) because the protein expression of hepatic CYP2C11 is comparable between the two types of rats (hepatic CLint for the disappearance of torasemide is not available) (Table 1).

Based on CLR,fp (Table 3), torasemide is primarily excreted in the urine via glomerular filtration, and active renal tubular secretion of the diuretic could contribute somewhat to its renal excretion for both types of rats. The protein-binding values of torasemide to 4% HSA remain constant (95.2%) at torasemide concentrations ranging from 1 to 50 μg/ml.[56] The plasma protein-binding value of torasemide in NARs (23.3%) could be due to binding to 3.1% α-globulin (26.2%) and 1.8% β- plus 0.63% γ-globulins (12.3%) because the value for 0.0042% HSA was only 9.69%.[24]

Theophylline (a low-hepatic extraction ratio drug, indirect hepatic extraction ratios 0.0743[19] and 0.0441[58])

Theophylline, a bronchodilator, is metabolized via hepatic CYP1A1/2, 2B1/2 and 3A1/2 (not CYP2C11, 2D1 or 2E1) in male SD rats.[24, 27] After i.v. administration of theophylline to male NARs, its plasma CLNR was faster (by 49.3%) than that in control rats[19] (Table 3). This could be due to a more rapid hepatic CLint for the disappearance of theophylline because of an increased level of protein expression of hepatic CYP1A2 in NARs (Table 1). Although the CLint was not available, the more rapid CLNR of theophylline could be supported by the faster (by 48.3%) hepatic CLint for the formation of 1,3 dimethyluric acid, a metabolite of theophylline than that in control rats.[19] The fps of control rats and NARs were 73.1% and 88.0%, respectively (Table 3). Thus, the contribution by the increase in the fp in NARs (by 20.4%) to the more rapid CLNR of theophylline seemed to be almost negligible because theophylline is a drug with a low HER.

Based on the CLR,fp (Table 3), theophylline is primarily excreted in the urine via active renal tubular re-absorption for both types of rats. The protein-binding values of theophylline to human serum remain constant (44.4%) at theophylline concentrations ranging from 15 to 150 μg/ml.[59]

Omeprazole (an intermediate-hepatic extraction ratio drug, direct hepatic extraction ratio 0.59[60])

Omeprazole, an irreversible proton pump inhibitor in gastric parietal cells, is metabolized via hepatic CYP1A1/2, 2D1 and 3A1/2 (not CYP2B1/2, 2C11 or 2E1) in male SD rats.[28] After i.v. administration of omeprazole to male NARs, its plasma CLNR (which could represent its CLH[60, 61]) was more rapid (by 136%) than that in control rats[8] (Table 3). This could be due to the faster hepatic CLint for the disappearance of omeprazole (by 37.0% and 75.6%, based on ml/min per mg protein and ml/min per g liver, respectively) because of an increased protein expression of hepatic CYP1A2 (Table 1), and greater (by 112%) fp than those in control rats (Table 3) because QH did not seem to be altered in the NARs.[1, 6, 9]

Based on the CLR,fp (Table 3), omeprazole is mainly excreted in the urine via active renal tubular re-absorption for both types of rats. The protein-binding values of omeprazole to 4% HSA remain constant (91.7%) at omeprazole concentrations ranging from 1 to 200 μg/ml.[8] The plasma protein-binding value of omeprazole in NARs (60.7%) could be due to binding to 1.8% β- plus 0.63% γ-globulins (46.0%), 0.63% γ-globulin (36.0%) and 1.3% γ-globulin (40.5%) because the value for 0.0042% HSA was only 8.14%.[8]

Oltipraz (an intermediate-hepatic extraction ratio drug, direct hepatic extraction ratio 0.4[62])

Oltipraz, an anthelmintic (schistosomiasis), is metabolized via hepatic CYP1A1/2, 2B1/2, 2C11, 2D1 and 3A1/2 in male SD rats.[30] After i.v. administration of oltipraz to male NARs, its plasma CL (the Ae0–24 h of oltipraz was below the detection limit; therefore, the CL of oltipraz could represent its CLNR) was more rapid (by 125%) than that in control rats[29] (Table 3). This could be due to a more rapid (by 11.4%) hepatic CLint for the disappearance of oltipraz (based on ml/min per mg protein) than that in control rats[29] because of an increase in the protein expression of hepatic CYP1A2 (Table 1) and greater (by 197%) fp than that in control rats (Table 3), as QH did not seem to be altered in NARs.[1, 6, 9]

Based on the CLR,fp (Table 3), oltipraz is mainly excreted in the urine via active renal tubular re-absorption for both types of rats. The protein-binding values of oltipraz to 4% HSA remain constant (94.0–95.9%) at oltipraz concentrations ranging from 1 to 100 μg/ml.[63]

Phenytoin (an intermediate-hepatic extraction ratio drug, hepatic extraction ratio by isolated perfused rat liver 0.53[15])

In male SD rats, metabolism of phenytoin to form (S)-5-(4-hydroxyphenyl)-5-phenylhydantoin is mediated via hepatic CYP2C6 and, to a lesser extent, via CYP2C11.[32] After i.v. administration of phenytoin to male NARs, its plasma CL (because the Ae of phenytoin was <1% of the dose, the CL of phenytoin could represent its CLNR) was faster (by 99.4%) than that of the controls, but its blood CL was comparable with control rats[31] (Table 3). This difference in the plasma and blood CLs of phenytoin could be due to the several times greater red blood cell to plasma concentration ratio of the drug in NARs; the distribution of phenytoin into the red blood cells was greater in NARs.[31] In NARs, the more rapid plasma CL of phenytoin could possibly be due to the greater fp (by 451%) (Table 3) because the protein expression of hepatic CYP2C11 (hepatic CLint for the disappearance of phenytoin is not available; see Table 1) and QH did not seem to be altered in NARs.[1, 6, 9]

Warfarin (a low-hepatic extraction ratio drug, indirect hepatic extraction ratios 0.00687 and 0.0146 for 1 and 40 mg, respectively[33])

In rats, racemic warfarin, a coumarin anticoagulant, is metabolized via hepatic CYP1A1, 2B1, 2C6, 2C11 and 3A2.[34] After i.v. administration of racemic [14C]warfarin to male NARs (unchanged warfarin measured), its plasma CLs (Ae0–24 h was 1.6% of the dose at 1 mg/kg; therefore, the CL of warfarin could represent its CLNR) were more rapid (by 5980% and 2070% for 1 and 40 mg/kg, respectively) than those in control rats[33] (Table 3). This could be due to greater the fps (by 3210% and 748% for 1 and 40 mg/kg, respectively) (Table 3). The hepatic CLint for the disappearance of warfarin was not available.

5-Fluorouracil (a high-hepatic extraction ratio drug, indirect hepatic extraction ratio 0.945[36])

In male SD rats, 5-fluorouracil (5-FU), a pyrimidine analogue anti-neoplastic agent, is metabolized via hepatic CYP1A1/2.[36] After i.v. administration of 5-FU to male NARs, its plasma CLNR was faster (by 78.1%) than that in control rats[35] (Table 3). This could be partly due to the greater fp (by 9.14%) than that in control rats (Table 3) because the QH did not seem to be altered in NARs.[1, 6, 9]

Based on the CLR,fp (Table 3), 5-FU is mainly excreted in the urine via glomerular filtration for both types of rats. The protein-binding values of 5-FU to 4% HSA remain constant (34.7%) at 5-FU concentrations ranging from 1 to 100 μg/ml.[35] The plasma protein-binding value of 5-FU in NARs (26.0%) could be due to binding to 3.1% α-globulin (6.48%), 1.8% β- plus 0.63% γ-globulins (24.8%), and 1.3% γ-globulin (3.48%) because the value for 0.0042% HSA was only 8.79%.[35]

Gliclazide (a low-hepatic extraction ratio drug, hepatic extraction ratio by the perfused rat liver 0.1[64])

Gliclazide, an oral antidiabetic drug, is primarily metabolized via hepatic CYP2C subfamily and, to a lesser extent, the CYP3A subfamily in rats.[38] After i.v. administration of gliclazide to male NARs, its plasma CL (<30% of the i.v. dose was excreted in Ae0–150 min; therefore, the CL of gliclazide may represent its CLNR) was more rapid (by 659%) than that in control rats.[37] This could be due to the greater fp (by 761%)[37] because the protein expression of hepatic CYP2C11 and 3A1/2 was comparable between the two types of rats (Table 1). Although the CLint for the disappearance of gliclazide was not available, the metabolism of gliclazide in the 10 000 g supernatant fraction of liver homogenates was comparable between the two groups of rats.[37]

Bilirubin

Bilirubin, a cholephilic organic anion, is rapidly taken up by the liver and forms monoglucuronide and diglucuronide conjugates mainly in the rat intestinal mucosa and is excreted into the bile.[65] Interestingly, the 1-h biliary excretion (total radioactivity of primarily the monoglucuronide and diglucuronide conjugates were measured) is dependent on bilirubin doses. After an i.v. bolus administration of trace amounts of 3.6 nmol/kg [3H]bilirubin to male NARs, the 1-h biliary excretion was comparable with control rats (16.4% and 18.3% of the dose for NARs and control rats, respectively).[39] However, after a loading amount of 0.5 μmol/kg [3H]bilirubin was injected, the 1-h biliary excretion was smaller (by 37.0%) in NARs (34% of dose) than that in control rats (54% of dose), which could be due to the low binding capacity of serum proteins in NARs for ligandin (glutathione S-transferase) (approximately 7% of control rats).[39] When bilirubin was administered with equimolar albumin or ligandin to NARs, the 1-h biliary excretion (∼60% of the dose) was greater than that without albumin or ligandin.[39]

Phenolsulfonphthalein

Phenolsulfonphthalein (PSP), a nephrophilic (an amphipathic organic anion) dye, is mainly excreted via the urine and bile. Serum concentrations of bile acid and the bile flow rate are comparable between the NARs and control rats.[66] After i.v. administration of PSP to male NARs, its Ae0–6 h was smaller (by 54.5%) than that in control rats[40] (Table 3). Interestingly, injection of PSP with equimolar albumin to NARs resulted in smaller (by 30.5%) and greater (by 52.7%) Ae0–6 hs of PSP than that in control rats and NARs without albumin, respectively.[40] In other studies, after i.v. administration of PSP to male NARs, its Ae0–1 h and 1-h biliary excretion were smaller (by 55.1%) and greater (by 48.2%), respectively, than those in control rats[41] (Table 3). Interestingly, when PSP was injected with equimolar albumin into NARs, its Ae0–1 h was smaller (by 31.6%) and greater (by 52.2%) than that in control rats and NARs without albumin, respectively. The corresponding values of 1-h biliary excretion were comparable with and smaller than (by 32.0%) those in control rats and NARs without albumin, respectively.[41] Together, these observations indicate that the interaction of PSP with circulating albumin is important for its renal excretion.[40, 41]

Mercapturic acid (N-acetylcysteine S-conjugates of xenobiotics)

After i.v. administration of S-benzyl-N-acetyl-L-[U-14C]cysteine, the final metabolite derived from glutathione S-conjugation, which is mainly excreted via the urine, to NARs, its Ae0–30 min (total radioactivity measured) was smaller (by 54.8%) than that in control rats[42] (Table 3). Interestingly, when mercapturic acid was injected with equimolar albumin into NARs, its Ae0–30 min was greater (by 50.2%) and smaller (by 32.1%) than that in control rats and NARs without albumin, respectively.[42] Together, these observations indicate that albumin is important for the urinary excretion of mercapturic acid.[42]

Sulfobromophthalein

After i.v. administration of sulfobromophthalein (BSP), an organic anion, which is mainly excreted via the bile, to male NARs, its 1-h biliary excretion of both free and glutathione conjugates were slower (by 29.3%; 58 vs. 41%) than that in control rats.[43] Although in NARs, the 1-h biliary excretion of BSP was smaller, the plasma CL of BSP was more rapid (by 68.2%) than those in control rats.[43] The exact reason for this is not clear because the CLR and CLNR (except biliary clearance) of BSP were not available. Interestingly, when 5 μmol/kg BSP bound to equimolar albumin was injected into NARs, its 1-h biliary excretion was greater than that in control rats and NARs without albumin (by 20.7% and 70.9%, respectively).[43]

Taurocholic acid

After i.v. administration of [3H]taurocholic acid, a cholephilic organic anion, which is mainly excreted via the bile, to male NARs, its 1-h biliary excretion (total radioactivity measured) was greater (by 21.5%) than that in control rats.[44] These results are in contrast with those for BSP[43] and bilirubin.[39]

Drugs that are not altered pharmacokinetically in Nagase analbuminaemic rats

Ipriflavone (a low-hepatic extraction ratio drug, direct hepatic extraction ratio 0.294[67])

Ipriflavone, an agent for the treatment of osteoporosis, is primarily metabolized via hepatic CYP1A1/2 and 2C11 (not CYP2B1/2, 2D1, 2E1, or 3A1/2) in male SD rats.[45] Ipriflavone is primarily metabolized in the rat liver, so the plasma CLNR of ipriflavone could represent its hepatic metabolic clearance in rats.[45] After i.v. administration of ipriflavone to male NARs, its plasma CLNR was comparable with control rats[7] (Table 3). This could be supported by the comparable hepatic CLint for the disappearance of ipriflavone and the comparable fp for both types of rats.[7]

Based on the CLR,fp (Table 3), ipriflavone is excreted in the urine via active renal tubular re-absorption for both types of rats. The protein-binding values of ipriflavone to 4% HSA remain constant (96.6%) at ipriflavone concentrations ranging from 1 to 200 μg/ml.[68] The plasma protein-binding value of ipriflavone for 0.5% HSA was 85.5%,[68] so the value of 98.7% in NARs suggests that ipriflavone also binds to other proteins (globulin) in the plasma of NARs.

Micafungin (a low-hepatic extraction ratio drug, indirect hepatic extraction ratio <0.0379[10])

The hepatic CYPs responsible for the metabolism of micafungin, an antifungal agent, do not seem to have been reported in rats. After i.v. administration of micafungin to male NARs, its plasma CL (the renal excretion of micafungin was low; therefore, the CL of micafungin could represent its CLNR) was comparable with control rats.[10] This could be due to comparable hepatic CLint for the disappearance of micafungin (CLint not available) because fp was comparable between the two types of rats.

Methotrexate

In male SD rats, methotrexate (MTX), an anticancer agent, is not metabolized via hepatic CYP isozymes.[46] After i.v. administration MTX to male NARs (Ae0–24 h was 36.0% in control rats), its plasma CLNR was comparable with control rats[46] (Table 3).

Based on the CLR,fp (Table 3), MTX is mainly excreted in the urine via glomerular filtration in both types of rats. The protein-binding values of MTX to human serum remain constant (41.5–51%) at MTX concentrations ranging from 45.4 pg/ml to 454 μg/ml.[69] The plasma-protein binding value of MTX in NARs (41.4%) could be due to the binding to 1.8% β- plus 0.63% γ-globulins (34.2%) because the value for 0.0042% HSA was only 5.41%.[46]

Conclusions

Although the factors influencing the GI absorption of drugs in NARs have not been thoroughly studied, the GI absorption of drugs does not seem to be altered in NARs. The fraction of the oral dose unabsorbed from the GI tract up to 24 h (funabs) could be estimated based on reported Equation (3), based on the previous work, assuming linear pharmacokinetics:[17]

display math(3)

where GI24 h represents the fraction of the dose remaining in the GI tract at 24 h (including its contents and faeces).

The fractions of the oral dose absorbed (fabs; 1 − funabs) of omeprazole (0.982 and 0.990 for control rats and NARs, respectively[8]), oltipraz (0.933 and 0.890, respectively[29]) and ipriflavone (0.843 and 0.763, respectively[7]) were found to be similar between the control rats and NARs. The GI absorption of gliclazide was also reported to be comparable between the two types of rats.[37]

The distribution of drugs could be altered in NARs. The Vss values of furosemide, azosemide, theophylline, oltipraz, 5-FU, gliclazide and PSP in NARs were larger, and this could be due to the greater fps than those in control rats (Table 3). In NARs, the Vss values of ipriflavone (Table 3) and micafungin[5] were not altered because the fps were also not altered compared with control rats. In NARs, however, Vss values of bumetanide, torasemide and omeprazole were not altered, although the fps were greater than those in control rats, and the Vss value of MTX was not altered although fp was smaller than that in control rats (Table 3). The exact reason for this is not clear.

The changes in the CLNRs (or CLs) of drugs that are mainly metabolized via hepatic CYPs could be explained well by the three factors: changes in hepatic in vitro CLint (because of CYP changes), fp and QH depending on the HERs of the drugs.

Patients with nephrotic syndrome (hypoalbuminaemia, <3 g/100 ml; proteinuria, >3 g/day; and oedema formation) also exhibit hypercholesterolaemia, hyperlipidaemia and low serum albumin concentration.[70] In patients with nephrotic syndrome, the plasma concentration of α2-macroglobulin was also higher (by 197%) than that in control subjects.[70] Although hepatic CYP changes in patients with nephrotic syndrome have not been reported, NARs could serve as an animal model for the patients. In NARs and patients with nephrotic syndrome, effects of plasma albumin binding in circulating blood in the diuretic effect of furosemide have been reported.[12] After i.v. administration of furosemide, urine output (rate of urine formation) is increased in control rats, but diuresis is not enhanced in NARs.[12] This could be due to the smaller Ae0–60 min of furosemide in NARs than that in control rats (Table 3) because both types of rats have identical renal sensitivities to furosemide.[12] The smaller Ae0–60 min and failure to enhance diuresis of furosemide in NARs could be due to their reduced plasma protein (albumin) binding than that in control rats.[12] Interestingly, in NARs, injected furosemide bound to rat albumin markedly promotes the diuresis compared with administration of furosemide alone.[12] Furthermore, injected furosemide bound to human albumin rapidly increased urine output in 14 out of 16 patients with hypoalbuminaemia who were resistant to furosemide.[12] After i.v. administration of combined furosemide and human albumin (or albumin infusion before furosemide administration) to patients with nephrotic syndrome[71-75] (not compared with furosemide alone[73-75]), liver cirrhosis and ascites,[76] and refractory nephrotic syndrome and chronic renal failure,[77] urine output increases (or resolution of ascites improves, oedema is reduced, or body weight is reduced) compared with administration of furosemide alone. However, contrasting results have also been reported: the effect of combined furosemide and human albumin is not beneficial compared with furosemide alone in patients with hypoalbuminaemia[78] and nephrotic syndrome.[79] The effects of combined furosemide and HSA compared with those of furosemide alone have been reviewed in patients with nephrotic oedema[80] or diuretic-resistant oedema.[81] Some possible mechanisms for the diminished diuretic effect of furosemide in patients with nephrotic syndrome have been reported: decreased urinary excretion because of decreased plasma protein binding of furosemide,[12] binding of furosemide to urinary albumin,[82-84] and tubular resistance in nephrotic rats[84] and in humans.[82, 84, 85] Unlike furosemide,[12] however, in NARs, the diuretic effects of azosemide,[6] bumetanide[22] and torasemide[24] were larger in the case of azosemide and comparable in the cases of bumetanide and torasemide compared with the respective control rats. This could be due to the significantly greater Ae0−t of each loop diuretic (Table 3) because of the binding to globulin. The earlier data suggest the role of plasma globulin binding in the circulating blood for the diuretic effects of the three diuretics. If the earlier rat data could be extrapolated to humans, azosemide, bumetanide or torasemide alone could have a diuretic effect in patients with nephrotic syndrome. It has been reported that azosemide alone[86] or bumetanide alone[87-89] showed diuretic effects (alleviation of oedema and ascites or body weight loss) in patients with nephrotic syndrome. More studies on patients are required to prove the earlier hypothesis.

Declarations

Conflict of interest

The author(s) declare(s) that they have no conflicts of interest to disclose.

Funding

This research was supported by the Research Fund of the Catholic University of Korea.