ERDI1

1
Exposure-Response Relationships and Drug
Interactions of Sirolimus

• James Zimmerman, PhD
• Clinical Research and DevelopmentWyeth Research

TM
2
Main Topics in This Presentation

• Background Information
• Exposure-Response Relationships
• Sirolimus whole blood exposure
• Logistic regression analysis of acute rejection
• Drug Interactions
• Conclusions

3
Background Information
TM
4
SIROLIMUS

• Nomenclature
• Sirolimus, rapamycin, RAPAMUNE?, RAPA
• Immunosuppressive agent
• Prophylaxis of renal allograft rejection
• Structurally similar to tacrolimus

MW 914.2 g/mol
5
Mechanism of the Immunosuppressive Effect

• Not a calcineurin inhibitor or antimetabolite
• Uniquely binds to mTOR (mammalian target of
  rapamycin)
• Blocks cytokine-mediated proliferative responses
• T cells
• B cells
• Mesenchymal cells

6
mTOR Is a Critical Kinase in Cell Cycle
Progression
7
Approved Regimens for Sirolimus in U.S.

• Sirolimus Fixed Oral Dose in Combination with
  Cyclosporine (initial regimen)
• Recommended as soon as possible after
  transplantation.
• Loading dose 6 mg (day 1)
• Maintenance dose 2 mg/day (thereafter)

Sirolimus 4 hours after cyclosporine capsules or
oral solution, Modified
8
Approved Regimens for Sirolimus in U.S.

• Sirolimus Concentration-Control
• During sirolimus and cyclosporine combination
  therapy
• In pediatric and hepatic-impaired patients
• After co-administration of strong
  inhibitors/inducers of CYP3A and P-gp
• After marked changes in cyclosporine doses

Sirolimus 4 hours after cyclosporine capsules or
oral solution, Modified
9
Approved Regimens for Sirolimus in U.S.

• Sirolimus Concentration-Control
• During and after cyclosporine withdrawal (2
  to 4 months after transplantation)
• Recommended for use in patients with low to
  moderate immunological risk
• Not recommended in patients with Banff grade III
  acute rejection, vascular rejection,
  dialysis-dependency, serum creatinine gt 4.5
  mg/dL, re-transplants, multi-organ transplants,
  and high PRA, nor in black patients.

10
Sirolimus Whole Blood Exposure
TM
11
Factors Affecting Observed Sirolimus Whole Blood
Exposure

• First-Pass Extraction
• Dosage Form (solution, tablet)
• Demographics (ethnicity, sex, age)
• Disease (hepatic impairment)
• Food (high-fat meal)
• Assay Method
• Drug Interactions

12
First-Pass Extraction and Dosage Form Affect
Sirolimus Oral Availability
NONMEM 2-stage population PK (Zheng et al. Clin
Pharmacol Ther 1996.) Bioavailability comparison
(Wyeth)
13
Demographics Affect SirolimusOral-Dose Clearance
(CL/F)
NONMEM Population PK (GloboMax) Two-Stage
Population PK (Wyeth) Interstudy comparison
(Wyeth)
14
Hepatic Impairment and a High-Fat Meal Affect
Sirolimus PK
HI Hepatic Impairment, S Solution, T Tablet
15
The Assay Method Used Affects Observed Sirolimus
Concentrations
Due to metabolite cross-reactivity of the
immunoassay Immunoassay values were used in the
logistic regression analyses.
16
Pharmacokinetic Measures of Sirolimus Exposure in
Phase 2 and 3 Clinical Trials

• Trough Concentrations
• Measured in all patients
• AUC(0-24h)
• Measured in selected patients
• Permitted determination of the Cmin(24h) vs
  AUC(0-24h) relationship
• R2 0.96 (1-30 ng/mL) in 42 renal transplant
  recipients from a phase 3 pivotal trial
• Troughs are a useful surrogate for AUC

17
Sirolimus Cmin vs AUC Relationship(Phase 3
trial RAPA CsA Corticosteroids)
18
Logistic Regression Analysis of Acute Rejection
TM
19
Logistic Regression Analysis Topics

• Objective
• Pharmacokinetic and Statistical Methods
• Database Characterization
• Exposure-Response (PK/PD)
• Based on raw data
• Based on logistic regression analysis
• Predicted Probability Analysis
• Concentration and Dose Predictions

20
Objective of the Logistic Regression Analysis

• To evaluate the optimal dose of sirolimus in
  renal transplant patients who are at high risk
  and lower risk for acute rejection
• High-risk patients were defined as
• Black patients
• Patients with PRA ? 50
• Patients with HLA mismatches ? 4
• Patients with retransplants
• Patients with multiorgan transplants

21
Pharmacokinetic Methods(Dose Administration)

• Sirolimus solution or tablets (fixed doses)
• Maintenance Dose (MD) 2 mg/d or 5 mg/d
• Loading Dose 3 x MD
• Dose changes were allowed for toxicity at
  investigator discretion
• Cyclosporine (concentration control)

22
Pharmacokinetic Methods(Parameter Estimation)

• Pharmacokinetic parameters were based on
  concentration and dose data up to 75 days after
  transplantation
• Sirolimus and Cyclosporine Trough Parameters
• Cmin,TN AUC0-t/t
• Sirolimus Dose-Proportionality Parameters
• DoseTN AUD0-t/t
• Cmin,TN "CDoseTN

23
Statistical Method Used for Logistic Regression
Analysis

• Probability (pr) of acute rejection was
  determined from a general model for 1 of 2
  possible outcomes using the SAS Logistic
  Procedure

• Y 1 (rejection)
• Y 0 (no rejection)
• k number of independent variables

24
Independent Variables Tested by Logistic
Regression Analysis

• Cmin,TN (sirolimus and cyclosporine)
• Sex (female vs male)
• Race (black vs nonblack)
• Age (recipient and donor)
• Donor Type (cadaveric vs living)
• PRA (? 50 vs lt 50)
• HLA Mismatch (? 4 vs lt 4)
• Ischemia time

25
Statistical Model Development

• Univariate Logistic Regression
• Stepwise Multivariate Logistic Regression
• Criteria for entry/removal of variables p lt 0.15
• Scale identification (linear vs nonlinear) for
  significant continuous variables
• Interaction-term testing

26
Summary of Clinical Studies Included in the
Logistic Regression Analysis
All patients received RAPA, CsA, and
corticosteroids
27
Total Numbers of Patients Among the Combined
Clinical Studies by Risk Group

• The numbers of high-risk and lower-risk patients
  among the combined clinical studies were nearly
  identical.
• High-risk patients (n 914)
• Lower-risk patients (n 918)

28
Percentages of Rejections Among Individual
Clinical Studies by Risk Group
29
Numbers of Patients Among High-Risk Categories by
Rejection Status
30
Sirolimus Trough Concentrations Among Independent
Variables by Rejection Status
31
Concentration-Effect( Raw Data)
32
HLA Mismatch-Effect(Raw Data)
33
Univariate Logistic Regression Analysis
Race (black vs nonblack) p-value 0.18
34
Multivariate Logistic Regression
Analysis(Testing for Scale Linearity and
Interaction Terms)

• Sirolimus and cyclosporine Cmin,TN values were
  nonlinear by quartile breakdown and the
  Box-Tidwell transformation
• Sirolimus and cyclosporine Cmin,TN were therefore
  dichotomized in the final model
• Sirolimus ? 5 vs gt 5 ng/mL
• Cyclosporine ? 150 vs gt 150 ng/mL
• There were no significant interaction terms.

35
Final Multivariate Logistic Regression Model(All
Variables Dichotomized)
36
Method for Predicting Probabilities Among All
Combinations of Independent Variables

• Equation for Predicting the Probability (Pr) of
  Acute Rejection
• Pr 1/(1 e-?)
• ? ?0 ?1X1 ?2X2 ?3X3 ?4X4
• Subscript 1 sirolimus (continuous)
• Subscript 2 cyclosporine (x 1 or 0)
• Subscript 3 HLA mismatch (x 1 or 0)
• Subscript 4 sex (x 1 or 0)

1 significant effect , 0 no significant
effect
37
Results of Predicted Probability
Analysis(Sirolimus as a Continuous Variable)
38
Sirolimus Dose Adjustments in Patients With
Increased Probabilities of Acute Rejection

• 3-Step Approach
• Predict the sirolimus concentrations required to
  offset the increased probability of acute
  rejection
• Determine the dose proportionality of data used
  in the logistic regression analysis
• Estimate dose adjustments based on the previous
  steps

39
Method for Predicting RAPA Troughs to Offset the
Increased Probability of Rejection

• Equation for Predicting Sirolimus Trough
  Concentrations
• ? -Ln (1/Pr) - 1
• ? ?0 ?1X1 ?2X2 ?3X3 ?4X4
• Subscript 1 sirolimus (continuous)
• Subscript 2 cyclosporine (x 1 or 0)
• Subscript 3 HLA mismatch (x 1 or 0)
• Subscript 4 sex (x 1 or 0)

1 significant effect , 0 no significant
effect
40
Illustration of the Effect of HLA Mismatch on the
Probability of Acute Rejection
41
Predicted RAPA Troughs to Offset the Increased
Probability of Acute Rejection
42
Dose-Proportionality Analysis
43
Predicted Sirolimus Doses to Offset the Increased
Probability of Acute Rejection
44
Sirolimus Drug Interactions
TM
45
First-Pass Extraction is the Major Determinant
for Sirolimus Drug Interactions

• Substrate for intestinal and hepatic cytochrome
  P450 3A (CYP3A)
• 7 major metabolites
• Substrate for P-glycoprotein (P-gp)

Sattler et al. Drug Metab Dispos 1992.
Lampen et al. J Pharmacol Exp Ther 1998.
Lampen et al1998
46
Coordinate Effect of Intestinal P-gp and
CYP3A on Sirolimus Oral Availability
47
An Additional Pathway for Sirolimus Extraction

• In Caco-2 cell monolayers expressing CYP3A4
• Sirolimus was nonenzymatically degraded to the
  ring-opened product seco-rapamycin
• Seco-rapamycin was metabolized to M2 by an
  unidentified nonmicrosomal enzyme
• M2 was the major product at all sirolimus
  concentrations
• M2 and seco-rapamycin were actively secreted by
  P-gp

Paine et al. J Pharmacol Exp Ther 2002.
Lampen et al1998
48
Summary of Study Designs Used to Investigate
Sirolimus Drug Interactions
RTx renal transplant SD single dose, MD
multiple dose
49
Drugs That Significantly Affected Sirolimus
Exposure in Healthy Subjects
50
Known Effects of the Drugs That Significantly
Affected Sirolimus Exposure
S substrate I inhibitor (Zhang Y,
Benet LZ. Clin Pharmacokinet 2001.) xWatkins PB.
J Clin Invest 1989. Kim RB. Pharmaceut Res 1999.
51
The Effect of Relative Dose Time on the RAPA/CsA
Interaction in Healthy Subjects
52
The Effect of Dosage Form on theRAPA-CsA
Interaction in Healthy Subjects
An interstudy comparison of 10-mg oral RAPA
doses for solution (study 1) vs tablet (study 2)
and 300-mg oral CsA. Units ng?h/mL
53
Drugs That Did Not Significantly Affect
Sirolimus Exposure in Healthy Subjects
54
Known Effects of the Drugs That Did Not
Significantly Affect Sirolimus Exposure
S substrate I inhibitor (Zhang Y,
Benet LZ. Clin Pharmacokinet 2001.) Jacobsen W
et al. Drug Metab Dispos 2000. xKim RB. Drug
Metab Rev 2002. Brian WR et al. Biochem 1990.
55
Effect of Sirolimus on the Exposure (AUC) of
Other Drugs in Healthy Subjects

• Sirolimus did not affect the exposure of
• Diltiazem, cyclosporine, ketoconazole, rifampin,
  acyclovir, atorvastatin, digoxin, ethinyl
  estradiol, norgestrel, glyburide, nifedipine, and
  tacrolimus
• Sirolimus affected the exposure of
• Erythromycin AUC ? 69
• S-(-) Verapamil AUC ? 48

56
Conclusions
TM
57
Conclusions (Sirolimus Exposure-Response)

• The whole blood exposure of sirolimus was
  affected by first-pass extraction, dosage form,
  demographics, hepatic disease, high-fat meal,
  assay method, and other drugs.
• Based on a multivariate logistic regression
  analysis, the probability of acute rejection in
  1832 renal allograft patients was increased by
  sirolimus troughs ? 5 ng/mL, cyclosporine troughs
  ? 150 ng/mL, HLA mismatches ? 4, and female sex.

58
Conclusions (Sirolimus Exposure-Response)

• Individualization of sirolimus doses immediately
  after transplant, based on HLA mismatch and sex,
  would likely decrease the probability of acute
  rejection in renal allograft recipients receiving
  combined sirolimus, full-dose cyclosporine, and
  corticosteroid therapy.

59
Conclusions (Sirolimus Drug Interactions)

• Cyclosporine, diltiazem, verapamil, erythromycin,
  and ketoconazole significantly increased
  sirolimus whole blood exposure in healthy
  subjects.
• Rifampin significantly decreased sirolimus whole
  blood exposure in healthy subjects.
• Sirolimus significantly increased erythromycin
  and S-(-) verapamil plasma exposure in healthy
  subjects.

60
ACKNOWLEDGEMENTS

• Clinical investigators and bioanalytical
  laboratories for the Rapamune trials in
  Australia, Canada, Europe, and U.S.A
• Rapamune therapeutics group and clinical data
  management of Wyeth Research
• Kyle Matschke, MS, biostatistician of Wyeth
  Research