Drug Absorption Assessment Using In Vitro Data

When MAD is estimated with in vivo data, either the in vivo absorption rate constant, or the drug in vivo intestinal permeability is required for the calculation. However, during early stages of drug discovery and development, in vivo data are usually unavailable. The challenge is to optimize the process for selecting compounds to evaluate in vivo human studies based on in vitro data. Fortunately, drug permeability in Caco-2 cells and drug solubility are routinely screened in the pharmaceutical industry. These data can be utilized to predict fraction of drug absorbed and MAD in humans to identify the best candidates for further clinical development.

4.5.2.1 In Vitro Testing Conditions for Determining Drug Permeability in Caco-2 Cells and In Vitro/In Vivo Permeability Correlation

Many laboratories have established methods for measuring drug permeability in Caco-2 cells with different testing conditions (Chong et al., 1996; Yee, 1997;

O Furosemide

□ Hydrochlorothiazide X Atenolol

H Cimetidine X Mannitol A Terbutaline

* Lisinopril (C) O Metoprolol

■ Phenylalanine (C) O Desipramine — Antipyrine

Q Verapamil (C) A Ketoprofen + Naproxen E3D-Glucose (C)

Figure 4.10. In vitro/in vivo permeability correlation of 20 drugs at pH 6.5. Correlation coefficient (R2 = 0.7276) was calculated from the permeability of all 20 drugs. Correlation coefficient (R2 = 0.8492) was calculated from the permeability of the following drugs: furosemide, hydrochlorothiazide, atenolol, cimetidine, mannitol, terbutaline, metoprolol, propranolol, desipramine, antipyrine, piroxicam, ketoprofen, and naproxen. Correlation coefficient (R2 = 0.7854) was calculated from the permeability of the following drugs: cephalexin, enalapril, lisinopril, losartan, amoxicillin, phenylalanine, l-leucine, l-dopa, d-glucose, cyclosporin, and verapamil. Drugs are labeled with different symbols. Black symbols are drugs absorbed through carrier-mediated process, while gray and open symbols are drugs absorbed through passive diffusion (Sun et al., 2002)

Pade and Stavchansky, 1998; Yamashita etal., 2000). Some laboratories use buffer with pH 7.4 in both apical and basolateral sides of the Caco-2 cells, while others use pH 6.5 buffer at the apical side and pH 7.4 buffer at the basolateral side. When correlation analysis was performed between in vitro drug permeability in Caco-2 cells and in vivo drug permeability in humans, a better correlation was observed between human in vivo permeability and Caco-2 permeability measured at pH 6.5 than at pH 7.4 (Figs. 4.10 and 4.11). The correlation coefficient (R2) of in vitro and in vivo permeability of 24 drugs assayed at pH 7.4 was 0.5126 in (4.26), while the in vitro and in vivo permeability correlation coefficient ( R2) of the 20 drugs determined at pH 6.5 was 0.7276 in (4.27) (Sun et al., 2002).

Log Pf human = 0.4926 Log Pelf, Caco-2 - 0.1454, (4.26)

Log Peff, human = 0.6532 Log Peff, Caco-2 - 0.3036. (4.27)

However, if the drugs were absorbed through a carrier-mediated processes, such as cephalexin, enalapril, cyclosporin, amoxicillin, lisinopril, losartan, phenylalanine,

01 a

-logY = 0.6532 logX - 0.3036, R2 = 0.7276 (all drugs)

— logY = 0.7524 logX - 0.5441, R2 = 0.8492 (passively diffusive)

— ■ LogY = 0.542 LogX + 0.06, R2 = 0.7854 (Carrier-mediated)

-logY = 0.6532 logX - 0.3036, R2 = 0.7276 (all drugs)

— logY = 0.7524 logX - 0.5441, R2 = 0.8492 (passively diffusive)

— ■ LogY = 0.542 LogX + 0.06, R2 = 0.7854 (Carrier-mediated)

0.1 1 10 100 Caco-2 permeability ( x 10-6 cm/s) at pH 6.5

1000

A Frosemide O Hydrochlorothiazide X Atenolol ^Ranitidine X Cimetidine HMannitol + Terbutaline X Creatine ^Amoxicillin (C) ^Lisinopril (C) HLosartan (C) O Metoprolol O Cephalexin (C)

□ Desipramine — Antipyrine ■ L-leucine (C) A Piroxicam X Verapamil (C) O Naproxen + Ketoprofen ▲ D-glucose (C)

Figure 4.11. In vitro/in vivo permeability correlation of 24 drugs at pH 7.4. Correlation coefficient (R2 = 0.5126) was calculated from the permeability of all 24 drugs. Correlation coefficient (R2 = 0.8376) was calculated from the permeability of the following drugs: furosemide, hydrochlorothiazide, atenolol, ranitidine, cimetidine, mannitol, terbu-taline, creatine, metoprolol, propranolol, desipramine, antipyrine, piroxicam, ketoprofen, and naproxen. Correlation coefficient (R2 = 0.6775) was calculated from the permeability of the following drugs: cephalexin, enalapril, lisinopril, losartan, amoxicillin, phenylala-nine, l-leucine, l-dopa, d-glucose, cyclosporin, and verapamil. Drugs are labeled with different symbols. Black symbols are drugs absorbed through carrier-mediated process, while gray and open symbols are drugs absorbed through passive diffusion (Sun et al., 2002)

verapamil, L-dopa, D-glucose, and L-leucine were excluded, the in vitro/in vivo permeability correlation improves at both pHs, such that the permeability correlation coefficient (R2) of 15 passively diffused drugs at pH 7.4 and 13 passively diffused drugs at pH 6.5 were 0.8376 in (4.28) and 0.8492 in (4.29), respectively.

Log Peff, human = 0.6836 Log Pelf, Caco-2 - 0.5579, (4.28)

Log Peff, human = 0.7524 Log Peff, Caco-2 - 0.5441. (4.29)

4.5.2.2 Estimation of Fraction of Drug Absorbed In Humans Using In Vitro Drug Permeability in Caco-2 Cells

When in vitro drug permeability in Caco-2 cells is plotted against drug fraction absorbed in humans, a relationship could also be established as shown in Figs. 4.12 and 4.13 (Sun et al., 2002). As these data clearly indicate, it might be difficult to predict the fraction of drug absorbed for the drugs with low Caco-2 permeability.

y

✓ ii ______ V

y

✓----

...••'"À

o

logY = 0.4926 logX -0.1454, R2 = 0.5126 (all drugs)

logY = 0.6836 logX -0.5579, R2 = 0.8376 (passively diffusive)

LogY = 0.4898 LogX + 0.3311, R2 = 0.6775 (Carrier-mediated)

1000

100-

100-

Figure 4.12. Prediction of the fraction of drug absorbed using Caco-2 permeability at pH 6.5. Drugs are labeled with different symbols. Closed symbols are drugs absorbed through carrier-mediated process, while open symbols are drugs absorbed through passive diffusion (Sun et al., 2002)

M

D-glucose

m

Verapamil

Phenylalanine

Piroxicam

Enalapril

Cephalexin

A

Lisinopril

Amoxicillin

o

Ketoprofen

Antipyrine

A

Desipramine

V

Propanolol

o

Metoprolol

0

Terbutaline

®

Mannitol

s

Cimetidine

A

Atenolol

<î>

hydrochlorothiazide

Furosemide

1

Trend line

Figure 4.12. Prediction of the fraction of drug absorbed using Caco-2 permeability at pH 6.5. Drugs are labeled with different symbols. Closed symbols are drugs absorbed through carrier-mediated process, while open symbols are drugs absorbed through passive diffusion (Sun et al., 2002)

10090807060504030

10090807060504030

r>

A

V

A

V

A

O

D-glucose

Verapamil

Piroxicam

Y

L-dopa

Cyclosporin

Cephalexin

O

Losartan

Lisinopril

A

Amoxicillin

O

Naproxen

Desipramine

A

Propanolol

V

Metoprolol

O

Terbutaline

0

Mannitol

©

Cimetidine

0

Ranitidine

A

Atenolol

W

Hydrochlorothiazide

<î>

Furosemide

Trend line

Figure 4.13. Prediction of the fraction of drug absorbed using Caco-2 permeability at pH 7.4. Drugs are labeled with different symbols. Closed symbols are drugs absorbed through carrier-mediated process, while open symbols are drugs absorbed through passive diffusion (Sun et al., 2002)

More discrepancy was also observed for the drugs with carrier-mediated absorption routes especially when drug permeability in Caco-2 cells was obtained at pH 7.4 in the apical side.

4.5.2.3 Estimation of MAD in Human Based on In Vitro Data

Since an in vitro and in vivo drug permeability correlation has been established in (4.26), in vivo drug permeability in human could be easily estimated by in vitro drug permeability in Caco-2 cells. Although some of the transporter substrates showed high discrepancy from the in vitro/in vivo permeability correlation when Caco-2 permeability was obtained at pH 7.4, the overall correlation has shown reasonable prediction when Caco-2 permeability was obtained at pH 6.5 (Sun et al., 2002). Since MAD could be estimated using in vivo drug permeability in human with (4.25), the MAD could be estimated with in vitro drug permeability in Caco-2 cells in the following (4.30)

MAD = Peff, humanSAeff T = 10(a6532Log Pf Caco-0-3036) ¿AfT, (4.30)

where Peff, human is the in vivo drug permeability in human, Peff, Caco is the in vitro drug permeability in Caco-2 cells, S is the drug solubility, Aeff is the effective absorption surface area without considering villi and microvilli, and T is transit time in small intestine (3 h). As discussed earlier, the error associated when not considering surface area of villi and microvilli is cancelled in the MAD calculation using permeability in (4.25) and (4.30). In addition, the surface area of microvilli in Caco-2 cells is also irrelevant in calculation of MAD, since the human permeability is calculated with Caco-2 permeability by the correlation analysis.

Alternatively, the MAD can be estimated using (4.22) and (4.23), where Ka can be estimated with human permeability in vivo or Caco-2 permeability in vitro with the following (4.31)

Ka = Peff, human(2/R) = (2/R) 10(0 6532Log Peff,Caco-0.3036), (4.31)

where Peff, human is drug in vivo permeability in human, Peff, Caco is drug in vitro permeability in Caco-2 cells, and R is the radius of small intestine (2 cm). The examples of MAD estimation using in vitro drug permeability in Caco-2 cells are summarized in Table 4.3.

Table 4.3. Estimation of MAD using drug permeability in Caco-2 cells with following equation: MAD = 10(0 6532LogPeff,caco-0 3036)SAeffT, or with calculated absorption rate constant (Ka) with following equations: Ka = (2/R) 10(a6532LogPeff,caco-0-3036) and MAD = SKaVT

Peff, Caco-2

Solubility

MAD (mg)

Calculated K a from

MAD (mg) from

(x10-6 cm s-1 )

Peff, Caco-2

Peff, Caco-2 (mm-1)

calculated K a

0.1

0.001

0.0955

0.000663

0.0298

0.1

0.01

0.955

0.000663

0.298

0.1

0.1

9.545

0.000663

2.98

0.1

1

95.45

0.000663

29.8

1

0.001

0.429

0.002982

0.134

1

0.01

4.294

0.002982

1.34

1

0.1

42.94

0.002982

13.4

1

1

429.4

0.002982

134

10

0.001

1 . 932

0.01342

0.603

10

0.01

19 . 32

0.01342

6.03

10

0.1

193. 2

0.01342

60.3

10

1

1,932

0.01342

603

100

0.001

8.696

0.060388

2.17

100

0.01

86.96

0.060388

27.17

100

0.1

869.6

0.060388

271.7

100

1

8,696

0.060388

2,717

Was this article helpful?

0 0
Natural Treatments For Dandruff

Natural Treatments For Dandruff

Stop Your Scalp From Snowing! Finally... A Real Cure for Dandruff! No medication! No weird cures! No messy creams, smelly solutions or strange diets! Are you sick to death of being sold products that could make the problem worse? Are you sick of the shampoos and conditioners that make your head smell like sulphur? If so then try a permanent revolutionary treatment for dandruff!

Get My Free Ebook


Post a comment