1. Analysis of Zidovudine pharmacokinetics to determine whether there is a genetic component to the variability and to determine the bioequivalence of seven formulations 25.05.12 1 Angelina Fahmay

2. AZT Case study 1) Involving 100mg AZT :  24 healthy subjects (12 Female, 12 Male)  Weights of healthy subjects (Range: 50-77kg) (Average: 62kg)  Given FOUR different oral formulations of 100mg of AZT at SEVEN day intervals  ONE reference and THREE generics (Oral)  Blood samples collected after: 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 2.5, 3, 4, 5, 6, 8h 2) Involving a combination of 300mg AZT + Lamivudine:  Another/different set of 24 healthy subjects (12 Female, 12 Male)  Weights of healthy subjects (Range: 51.5 – 82.8 kg) (Average: 64kg)  Given THREE different oral formulations of the combination at SEVEN day intervals  ONE referenc e and TWO generics (Oral)  Blood samples collected after: 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 4, 6, 9, 12h 25.05.12 2

3. Research questions  Is the pharmacokinetic variability seen in the AZT data likely to be due to a genetic component?  Model the data using NONMEM VII (Non Linear Mixed Effects Modeling)  Compute the genetic component coefficient (R GC ) Between Occasion Variability (BOV) and Between Subject Variability (BSV) for clearance to look for possible genetic association  Does compartmental modeling and non-compartmental analysis produce similar results in a bioequivalence study of AZT data?  Perform a non-compartmental analysis (R software)  Use a compartmental modeling approach to estimate bioavailability 25.05.12 3

4. Zidovudine  Zidovudine also known as azidothymidine (AZT)  Synthesized in 1964 by Jerome P.Horwitz as a potential anti- cancer drug  First Anti-HIV drug  Nucleoside reverse transcriptase inhibitor (NRTI)  Chemically known as: 3’-azido-3’-deoxythymidine  An analogue to the nucleoside deoxythymidine triphosphate (dTTP)  The difference in structure that causes chain termination is that AZT lacks a hydroxyl group on the 3’ of the deoxyribose sugar moiety and instead contains an azido group 25.05.12 4

5. Mode of action of AZT  Zidovudine activated by cellular kinases  Zidovudine triphosphate  Zidovudine triphosphate acts as a competitive inhibitor for reverse transcriptase in the place of deoxythymidine triphosphate (dTTP) i.e. nucleoside analogue  Zidovudine triphosphate is then incorporated into the growing viral DNA chain  chain termination occurs 25.05.12 5

6. Plots of Raw data using R created  Individual plots using ID column (an individual plot for each of the 48 subjects)  Individual plots using PID column (an individual plot for each formulation for each patient)  A plot of all the data (all 48 healthy subjects)  A plot for each formulation (i.e. 1-7) 25.05.12 6

7. Plots of the Raw data using R Concentration vs. time plots for formulation 1- 4 for 100mg AZT data Concentration vs. time plots for formulation 1- 3 for 300mg AZT + 5TC data 25.05.12 7

8. NONMEM models Omega fixed Omega on Cl Omega on Cl, V Omega on Cl, V, Ka Omega on LAG LAG theta 0.2 LAG theta 0.5 BLQ values included Omega on Cl, V, Ka, LAG 2 Comp Omega on Cl, V, LAG PID into ID 25.05.12 8

9. Final model : 8h RUNMIN OBJClearance Volume of distribution Absorption rate constant LAG time Proportional error Additive error Form 2Form 3Form 4Form 2Form 3 100mg 300mg 8h ✓ 16922.35415017311300.2490.3986.6-0.02310.176-0.597-0.02110.159 ID column used (ID ranged from1-48) 1 compartment Omega on Ka fixed Dataset: Includes both 100mg + 300mg data M3 method used for BLQ LOQ = 3 Weight added as a covariate for clearance and volume of distribution Bioequivalence BOV on clearance was added 25.05.12 9

10. BLQ  11.11% of the 1944 observations had a concentration of AZT in plasma below the LOQ 25.05.12 10

11. Weight as a covariate on Clearance and Volume of Distribution A plot to show the relationship between weight and clearance (eta 1) / volume of distribution (eta 2), respectively, in run8g.mod i.e. before weight was added as a covariate A plot to show the relationship between weight and clearance (eta 1) / volume of distribution (eta 2), respectively, in run8h.mod i.e. after weight was added as a covariate 25.05.12 11

12. Goodness of Fit Plots 25.05.12 12

13. Visual Predictive Check 25.05.12 13

14. BIOEQUIVALENCE STUDY RunForm 4/1 (100mg)Dataset 8d 0.00341 100mg dataset used onlyBioequivalence 8e - 300mg dataset ued onlyBioequivalence 8f -0.511 100mg + 300mg Bioequivalence 8g -0.518 100mg + 300mg BOV and BSV 8h -0.597 100mg + 300mg Weight added to Cl + V 40.30 % 100.341% 25.05.1214 Bioequivalence for each of the 7 formulations using Non-compartmental and compartmental analysis Comparison of bioequivalence for formulation 4 throughout final set of model runs

15. Genetic component on clearance R GC CL= 1 – (0.00867/0.047) = 0.8155 (to 4 sig fig) BSV GREATER than BOV σ 2 = Between occasion variance (BOV) ω 2 = Between subject variance (BSV) 25.05.12 15

16. Metabolism of Zidovudine ZIDOVUDINE AMT GAMT GAZT UGT 3 potential N- glycosylation sites in UGT2B7 (asparagine at positions 67, 68 and 315 Genetic polymorphism of the UGT2B7 enzyme is proven to NOT be significant Decrease the affinity of AZT for glucurinodation. 25.05.12 16