Supplementary material – Gertz et al. 2013: Reduced physiologically-based pharmacokinetic model of repaglinide: impact of OATP1B1 and CYP2C8 genotype and source of in vitro data on the prediction of drug-drug interaction risk(PharmRes7865)
1. Source of repaglinide clinical data
Table I: Summary of all repaglinide clinical studies used in the current analysis and corresponding repaglinide dose, SLCO1B1 genotype and demographic data
ID / Dose (mg) / SLCO1B1 c.521 T>C / N (individuals) / Age (years) / Weight (kg) / Reference1 / 0.5 / TT / 16 / 22 / 73 / (1)
2 / 0.5 / TC / 12 / 22 / 68 / (1)
3 / 0.5 / CC / 4 / 22 / 78 / (1)
4 / 0.25 / TT / 12 / 24 / 75 / (2)
5 / 0.25 / TC / 6 / 23 / 74 / (2)
6 / 0.25 / CC / 6 / 23 / 75 / (2)
7 / 0.5 / TT / 16 / 22 / 73 / (3)
8 / 0.5 / TT / 8 / 24 / 74 / (3)
9 / 0.25 / TT / 36 / 23 / 73 / (4)
10 / 0.25 / TC / 16 / 23 / 73 / (4)
11 / 0.25 / CC / 4 / 23 / 73 / (4)
12 / 0.25 / TT / 12 / 24 / 71 / (5)
13 / 0.25 / CC / 8 / 24 / 70 / (5)
14 / 0.5 / TT / 12 / 24 / 71 / (5)
15 / 0.5 / CC / 8 / 24 / 70 / (5)
16 / 1 / TT / 12 / 24 / 71 / (5)
17 / 1 / CC / 8 / 24 / 70 / (5)
18 / 2 / TT / 12 / 24 / 71 / (5)
19 / 2 / CC / 8 / 24 / 70 / (5)
20 / 0.25 / MIX / 12 / 20-24 / 46-84 / (6)
21 / 0.25 / MIX / 12 / 24 / 71 / (7)
22 / 0.25 / MIX / 12 / 23.8 / 76.4 / (8)
23 / 0.25 / MIX / 9 / 19-23 / 62-97 / (9)
24 / 0.25 / MIX / 10 / 22.1 / 78 / (10)
25 / 0.25 / MIX / 10 / 23 / 73 / (11)
26 / 0.25 / MIX / 10 / n/a / 79 / (12)
27 / 0.5 / MIX / 9 / 22.8 / 66.3 / (13)
28 / 0.25 / MIX / 12 / 23 / 75 / (14)
29 / 0.25 / MIX / 12 / 18-24 / 52-85 / (15)
Visualization of the dose normalised plasma concentration time profiles (normalization to a 0.25 mg oral dose) stratified for OATP1B1 genotype (Figure 1).
Figure 1: Oral repaglinide plasma concentration-time profiles after single dose studies (all normalised to a dose of 0.25mg) in populations with different OATP1B1 genotypes (TT, TC, CC and MIX) – sources of data listed in Table 1
Table II: Individual AUC and Cmax values of repaglinide (normalised to 0.25mg)
AUC (ng.h/mL) / ReferenceTT / TC / CC
14.51
14.11
12.81
11.19
10.97
10.89
9.88
9.80
9.10
8.88
8.36
7.57
7.31
6.74
6.44
6.26
/ 14.68
13.46
13.42
13.24
13.24
10.10
9.36
9.06
8.84
7.44
7.05
6.57
/ 20.00
16.51
16.03
14.16
/ (1)
AUC (ng.h/mL)
TT / TC / CC
14.48
14.12
12.79
11.17
10.95
10.88
9.91
9.78
9.29
9.10
9.03
8.87
8.32
7.99
7.70
7.51
7.28
6.73
6.54
6.41
6.24
4.95
4.39
3.75
/ n/a / n/a / (3)
Cmax (ng/mL)
TT / TC / CC
4.03
2.67
2.79
4.15
3.97
3.70
4.55
2.49
5.63
2.17
4.25
4.27
2.43
2.67
4.37
4.15
3.38
3.64
2.22
5.49
1.36
4.47
6.25
4.24
5.97
3.06
6.34
3.40
2.64
4.34
2.57
4.11
4.02
3.75
6.24
2.96
5.08
4.14
4.64
2.47
3.84
3.42
3.15
/ n/a / 4.25
7.34
4.09
5.50
4.55
8.23
8.68
6.96
7.68
6.40
4.58
6.59
4.86
2.69
8.94
5.01
3.01
3.61
5.62
7.22
6.11
6.54
7.60
6.53
8.92
7.27
5.92
6.24
3.86
/ (5)
AUC (ng.h/mL)
TT / TC / CC
6.56
6.78
3.58
4.42
3.79
3.39
4.31
4.87
4.21
5.58
4.11
5.07
6.01
4.99
6.06
4.11
5.28
4.32
3.85
5.07
3.44
4.91
4.79
8.16
4.87
4.46
5.32
6.78
4.68
6.06
3.70
4.57
3.18
4.39
4.28
6.01
4.84
4.57
4.91
3.86
4.72
6.16
3.45
4.68
3.28
4.53
4.50
5.59
/ n/a / 7.00
10.06
5.77
8.03
7.65
10.47
10.73
9.51
5.82
9.21
4.72
7.53
6.83
14.82
18.88
8.16
4.61
10.15
10.57
7.41
6.84
6.21
7.84
9.22
9.83
7.24
10.49
4.91
16.22
13.15
/ (5)
AUC (ng.h/mL)
MIX
3.75
2.50
6.34
4.26
5.17
2.45
3.16
2.61
8.48
/ (10)
MIX
2.26
5.54
2.57
2.97
2.91
17.72
3.35
4.26
4.56
/ (12)1
MIX
3.27
2.82
2.82
4.47
6.02
4.36
2.67
8.53
8.08
3.23
2.89
2.89
6.92
3.72
3.46
4.14
4.06
3.76
3.16
3.01
6.17
4.74
3.57
3.53
4.10
3.91
5.83
2.71
4.21
3.23
3.16
10.98
4.25
4.92
10.53
9.59
/ (16)
MIX
8.49
5.29
3.40
2.97
9.88
8.08
4.27
4.22
4.19
/ (13)
Cmax (ng/mL)
MIX
5.07
4.38
8.93
6.14
5.97
3.94
3.77
2.55
2.46
MIX
11.69
9.30
7.99
6.46
5.67
5.67
5.38
4.65
4.43
2.76
2.25
/ (6)
1 reported in Backman et al. (2009)
Visualization of data presented in Table 2 as box plots (Figure 2).
Figure 2: Individual AUC and Cmax values reported in the literature for TT (1, n = 87, 43 AUC and Cmax, resp.), TC (2, n= 12, 0), CC (3, n= 34, 29) and MIX (4, n= 63, 20) normalised to a repaglinide dose of 0.25mg
2. Results – empirical model
Table III: Results of the optimisation performed in NONMEM using an empirical model
Fitting (FOCE INTER) / Bootstrap1Parameters / Final model / SE / Mean / SE
OFV / -1138.583
CL/F (L/h) / 53.4 / 1.85 / 53.4 / 1.88
ka (h-1) / 1.90 / 0.127 / 1.92 / 0.140
Vc/F(L) / 11.8 / 1.11 / 11.9 / 1.21
Q/F (L/h) / 27.3 / 2.08 / 27.3 / 2.18
Vp/F (L) / 35.9 / 1.88 / 35.8 / 1.91
tlag (h) / 0.218 / 0.00240 / 0.218 / 0.00265
Coefficient(F)2 / 0.308 / 0.00709 / 0.310 / 0.0779
Coefficient(CL)2 / 0.171 / 0.00469 / 0.169 / 0.0518
σadd / 0.0691 / 0.00649 / 0.0674 / 0.00670
ηCL/F (%) / 16.0 / 29.6 / 15.7 / 29.6
ηka (%) / 19.5 / 135 / 10.3 / 151
ηVc/F (%) / 24.2 / 47.0 / 19.7 / 50.8
ηQ (%) / 24.8 / 31.7 / 23.8 / 35.9
ηVp/F (%) / 7.62 / 26.0 / 24.1 / 27.3
ηtlag (%) / 5.14 / 48.1 / 5.03 / 45.8
1 bootstrap analysis performed n = 1000; 2 coefficient on F and CL for SLCO1B1 521CC carriers relative to the typical value of the reference OATP1B1 genotype (TT)
Figure 3: Conditionally weighted residuals (CWRES) vs. time and individual predictions (plasma) and absolute IWRES vs. individual predictions (plasma)
Figure 4: Observed plasma concentrations vs. individual (left) and population (right) predictions
Figure 5: Individual observations and predictions separated by genotype of oral repaglinide plasma concentration-time profiles using an empirical 2-compartmental model
3. NONMEM implementation of the semi-mechanistic model
Code starts
$PROBLEM Hybrid PK model of repaglinide which includes a mechanistic permeability rate limited liver model
$DATA Repaglinide_dataLN_final.csv IGNORE=@
; All data represent mean C-t profiles
; 29 individual profiles
; Applying log-transformation to the data
$INPUT ID TIME DV AMT RATE CMT MDV EVID GENOTYPE
$SUBS ADVAN13 TRANS1 TOL=6
$MODEL NCOMP=5 COMP=(AbsSite) COMP=(LiverB) COMP=(LiverT) COMP=(Centr) COMP=(Peri)
$PK
; System parameters
QH = 92.7; typical hepatic blood flow (L/h)
VL = 1.6; L; typical liver volume (L)
VLT = VL*(1-0.115); typical liver volume excluding blood (L)
VLb = VL*0.115; liver blood volume (L)
NHEPS = VLT*1000*120; % number of hepatocytes/ g or mL of liver
UC = 1/1000000*60; % unit conversion µL/min to L/h
; Repaglinide (FIXED)
FG = 0.89; unitless
FUP = 0.026; unitless
BP = 0.62; unitless
FUB = FUP/BP; unitless
FUCELL = 0.072; unitless
CLINT = 128.1*40000*VLT
PDIF = 12.7; µL/min/Mcells (mean n = 3)
; Repaglinide (uncertain)
IF(GENOTYPE.EQ.3) THEN
TVCLUP=THETA(2)
ELSE
TVCLUP=THETA(1)
ENDIF
TVKA=THETA(3); per h
TVV1=THETA(4)
TVQ=THETA(5)
TVV2=THETA(6)
TVTLAG=THETA(7)
TVRUVADD = THETA(8)
; MODEL FOR RANDOM BETWEEN STUDY VARIABILITY
CLUP=TVCLUP*EXP(ETA(1))
V1=TVV1*EXP(ETA(2))
Q=TVQ*EXP(ETA(3))
V2=TVV2*EXP(ETA(4))
KA=TVKA*EXP(ETA(5))
ALAG1=TVTLAG*EXP(ETA(6))
RUVADD=TVRUVADD
; CONVERSION BETWEEN BLOOD AND PLASMA CONCENTRATIONS
S4 = BP
$DES
DADT(1)=-KA*A(1); % absorption site (amounts)
DADT(2)=(KA*FG*A(1)+QH*A(4)-QH*A(2)-PDIF*NHEPS*UC*FUB*A(2)+PDIF*NHEPS*UC*FUCELL*A(3)-CLUP*NHEPS*UC*FUB*A(2))/VLb; % hepatic outlet (concentrations)
DADT(3)=(PDIF*NHEPS*UC*FUB*A(2)+CLUP*NHEPS*UC*FUB*A(2)-PDIF*NHEPS*UC*FUCELL*A(3)-CLINT*UC*FUCELL*A(3))/VLT; % liver tissue (concentrations)
DADT(4)=(QH*A(2)-QH*A(4)-Q*A(4)+Q*A(5))/V1; % central compartment (concentrations)
DADT(5)=(Q*A(4)-Q*A(5))/V2; % peripheral compartment (concentrations)
$ERROR
FLAG=0
IF(AMT.NE.0) FLAG=1
IPRED = LOG(F+FLAG
IF(CMT.EQ.4) W = RUVADD;
IF(W.EQ.0) W=1;
IRES=DV-IPRED
IWRES = IRES/W
Y = IPRED + W*EPS(1);
$THETA
(0.001,510.0); THETA(1)
(0.001,260.0); THETA(2)
(0.001 4.0); THETA(3)
(0.001,10.0); THETA(4)
(0.001,25.0); THETA(5)
(0.001,20.0); THETA(6)
(0,0.200); THETA(7)
(0,0.05); THETA(8)
$OMEGA BLOCK(4)
0.04; PPVCLUP
0.01 0.04; PPVV1
0.01 0.01 0.04; PPVQ
0.01 0.01 0.01 0.04; PPVV2
$OMEGA BLOCK(1)
0.04; PPVKA
$OMEGA BLOCK(1)
0.04; PPVTLAG
$SIGMA 1 FIX
$ESTIMATION MAXEVALS=999999 SIGL=6 NSIG=2 CTYPE=4 PRINT=10 NOABORT METHOD=1 INTER;
$COVARIANCE Matrix=S
$TABLE ID TIME AMT CMT EVID GENOTYPE CLUP KA V1 Q V2 ETA1 ETA2 ETA3 ETA4 ETA5 ETA6
IPRED IRES IWRES CWRES NOPRINT ONEHEADER FILE=RPG_M4a.fit
Code ends
4. Results – Reduced PBPK model
Figure 6: Individual observed and predicted oral plasma concentration-time profiles of repaglinide; dashed lines represent simulations with individual and solid lines represent simulations with population parameters
A
B
Figure 7: Impact of uncertainty in the in vitro parameters CLmet, CLdiff, and consequently CLuptake on the prediction of DDI magnitude illustrated as the % change in repaglinide oral AUC. A, in the presence of the potent OATP1B1 inhibitor cyclosporine (300mg Neoral) and B, assuming irreversible metabolic inhibition and 90% contribution of the inhibited pathway to repaglinide metabolic intrinsic clearance
Figure 8:Predicted effect of CYP2C8*3 polymorphism (increase in CLmet) on repaglinide plasma exposure in individuals with reduced OATP1B1 activity (CC for the c.521T>C polymorphism). Each boxplot represents individual AUCs of 1000 simulated subjects for every scenario regarding the increase in repaglinide CLmet. The fractional change in mean AUC relative to baseline (x1 the original value of CLmet) is reported. Highlighted in grey is the range where the increase in metabolic clearance is most probable based on in vitro data associated with the CYP2C8*3 variant.
Figure 9:Predicted effect of either other CYP2C8 polymorphism or inhibition effect (decrease in CLmet) on repaglinide plasma exposure in individuals with reduiced OATP1B1 activity (CC for the c.521T>C polymorphism). Each boxplot represents individual AUCs of 1000 simulated subjects for every scenario regarding the decrease in repaglinide CLmet as defined in methods. The fractional change in mean AUC relative to baseline (x1 the original value of CLmet) is reported
References:
1.A. Kalliokoski, M. Neuvonen, P.J. Neuvonen and M. Niemi. Different effects of SLCO1B1 polymorphism on the pharmacokinetics and pharmacodynamics of repaglinide and nateglinide. J Clin Pharmacol. 48:311-321 (2008).
2.A. Kalliokoski, J.T. Backman, K.J. Kurkinen, P.J. Neuvonen and M. Niemi. Effects of gemfibrozil and atorvastatin on the pharmacokinetics of repaglinide in relation to SLCO1B1 polymorphism. Clin Pharmacol Ther. 84:488-496 (2008).
3.A. Kalliokoski, J.T. Backman, P.J. Neuvonen and M. Niemi. Effects of the SLCO1B1*1B haplotype on the pharmacokinetics and pharmacodynamics of repaglinide and nateglinide. Pharmacogenet Genomics. 18:937-942 (2008).
4.M. Niemi, J.T. Backman, L.I. Kajosaari, J.B. Leathart, M. Neuvonen, A.K. Daly, M. Eichelbaum, K.T. Kivisto and P.J. Neuvonen. Polymorphic organic anion transporting polypeptide 1B1 is a major determinant of repaglinide pharmacokinetics. Clin Pharmacol Ther. 77:468-478 (2005).
5.A. Kalliokoski, M. Neuvonen, P.J. Neuvonen and M. Niemi. The effect of SLCO1B1 polymorphism on repaglinide pharmacokinetics persists over a wide dose range. Br J Clin Pharmacol. 66:818-825 (2008).
6.M. Niemi, J.T. Backman, M. Neuvonen and P.J. Neuvonen. Effects of gemfibrozil, itraconazole, and their combination on the pharmacokinetics and pharmacodynamics of repaglinide: potentially hazardous interaction between gemfibrozil and repaglinide. Diabetologia. 46:347-351 (2003).
7.L.I. Kajosaari, T. Jaakkola, P.J. Neuvonen and J.T. Backman. Pioglitazone, an in vitro inhibitor of CYP2C8 and CYP3A4, does not increase the plasma concentrations of the CYP2C8 and CYP3A4 substrate repaglinide. Eur J Clin Pharmacol. 62:217-223 (2006).
8.L.I. Kajosaari, J.T. Backman, M. Neuvonen, J. Laitila and P.J. Neuvonen. Lack of effect of bezafibrate and fenofibrate on the pharmacokinetics and pharmacodynamics of repaglinide. Br J Clin Pharmacol. 58:390-396 (2004).
9.M. Niemi, L.I. Kajosaari, M. Neuvonen, J.T. Backman and P.J. Neuvonen. The CYP2C8 inhibitor trimethoprim increases the plasma concentrations of repaglinide in healthy subjects. Br J Clin Pharmacol. 57:441-447 (2004).
10.J.T. Backman, J. Honkalammi, M. Neuvonen, K.J. Kurkinen, A. Tornio, M. Niemi and P.J. Neuvonen. CYP2C8 activity recovers within 96 hours after gemfibrozil dosing: estimation of CYP2C8 half-life using repaglinide as an in vivo probe. Drug Metab Dispos. 37:2359-2366 (2009).
11.J. Honkalammi, M. Niemi, P.J. Neuvonen and J.T. Backman. Dose-dependent interaction between gemfibrozil and repaglinide in humans: strong inhibition of CYP2C8 with subtherapeutic gemfibrozil doses. Drug Metab Dispos. 39:1977-1986 (2011).
12.A. Tornio, M. Niemi, M. Neuvonen, J. Laitila, A. Kalliokoski, P.J. Neuvonen and J.T. Backman. The effect of gemfibrozil on repaglinide pharmacokinetics persists for at least 12 h after the dose: evidence for mechanism-based inhibition of CYP2C8 in vivo. Clin Pharmacol Ther. 84:403-411 (2008).
13.M. Niemi, J.T. Backman, M. Neuvonen, P.J. Neuvonen and K.T. Kivisto. Rifampin decreases the plasma concentrations and effects of repaglinide. Clin Pharmacol Ther. 68:495-500 (2000).
14.L.I. Kajosaari, M. Niemi, M. Neuvonen, J. Laitila, P.J. Neuvonen and J.T. Backman. Cyclosporine markedly raises the plasma concentrations of repaglinide. Clin Pharmacol Ther. 78:388-399 (2005).
15.L.I. Kajosaari, M. Niemi, J.T. Backman and P.J. Neuvonen. Telithromycin, but not montelukast, increases the plasma concentrations and effects of the cytochrome P450 3A4 and 2C8 substrate repaglinide. Clin Pharmacol Ther. 79:231-242 (2006).
16.T.B. Bidstrup, P. Damkier, A.K. Olsen, M. Ekblom, A. Karlsson and K. Brosen. The impact of CYP2C8 polymorphism and grapefruit juice on the pharmacokinetics of repaglinide. Br J Clin Pharmacol. 61:49-57 (2006).
17.M. Gertz, C.M. Cartwright, M.J. Hobbs, K.E. Kenworthy, M. Rowland, J.B. Houston and A. Galetin. Application of PBPK modeling in the assessment of the Interaction Potential of Cyclosporine against Hepatic and Intestinal Uptake and Efflux Transporters and CYP3A4 Pharm Res 30:761-780. (2013).
1