1. Virological Correlates Associated with Treatment Failure at Week 48 in the Phase 3 Study of Maraviroc in Treatment-Naive Patients Jayvant Heera 1, Mike Saag 2, Prudence Ive 3, Jeannette Whitcomb 4, Marilyn Lewis 5, Lynn McFadyen 5, James Goodrich 1, Howard Mayer 1, Elna van der Ryst 5, and Mike Westby 5 1 Pfizer Global Research and Development, New London, CT, USA 2 University of Alabama at Birmingham, Birmingham, AL, USA 3 University of the Witwatersrand, Johannesburg, South Africa 4 Monogram Biosciences Inc., South San Francisco, CA, USA 5 Pfizer Global Research and Development, Sandwich, UK Presentation Number 40LB 15th CROI Boston, USA, February 3–6, 2008

2. Heera J, et al. 15th CROI 2008; Presentation 40LB 2 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Randomization 1:1 MERIT Study: Phase 3 Trial Design Maraviroc (MVC 300 mg BID) + Combivir (ZDV+3TC)* Efavirenz (EFV 600 mg QD) + Combivir (ZDV+3TC)* Primary analysis 0 Week 48 Week 96 Screening (6 weeks) Patients stratified by HIV-1 RNA < and ≥100,000 copies/mL at screening Geographic location: Northern Hemisphere and Southern Hemisphere Patient eligibility criteria: ≥16 years of age Treatment naive CCR5-tropic (R5) HIV-1 infection HIV-1 RNA ≥2,000 copies/mL No evidence of resistance to EFV, zidovudine (ZDV), or lamivudine (3TC) *Patients experiencing toxicity to ZDV or 3TC were permitted to substitute an alternative nucleoside reverse transcriptase inhibitor MVC QD arm discontinued at end of Phase 2b (Week 16) for failure to meet protocol-defined criteria to continue (205 patients completed 16 weeks) Saag M, et al. 4th IAS 2007. Abstract WESS104 MERIT Study

3. Heera J, et al. 15th CROI 2008; Presentation 40LB 3 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Week 48 Efficacy Results Saag M, et al. 4th IAS 2007. Abstract WESS104 MERIT Study – 48 weeks 0 10 20 30 40 50 60 70 80 90 100 361360 69.3 65.3 Patients (%) N= MVC + CBVEFV + CBV Includes all patients who received at least one dose of study medication HIV-1 RNA <50 copies/mL (ITT) –4.2* (–10.9 † ) *Difference (adjusted for randomization strata) † Lower bound of 1-sided 97.5% confidence interval; non-inferiority margin = –10% CBV = Combivir Time (weeks) EFV + CBV MVC + CBV Mean change in from baseline in CD4+ count (cells/mm 3 ) 142 cells/mm 3 169 cells/mm 3 Change in CD4+ cell count (LOCF)

4. Heera J, et al. 15th CROI 2008; Presentation 40LB 4 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Summary of Discontinuations Through 48 Weeks Saag M, et al. 4th IAS 2007. Abstract WESS104 MERIT Study – 48 weeks Reason for discontinuation EFV + CBV N=361 MVC + CBV N=360 All, n (%) 91 (25.2)97 (26.9) Adverse event, n (%) 49 (13.6)15 (4.2) Lack of efficacy, n (%) 15 (4.2)43 (11.9) Other reason, n (%) 9 (2.5)14 (3.9) Withdrew consent or lost to follow-up, n (%)18 (5.0)25 (6.9) Includes all patients who received at least one dose of study medication

5. Heera J, et al. 15th CROI 2008; Presentation 40LB 5 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Objectives ● To understand the virological correlates associated with treatment failure at Week 48 Methods ● Tropism was measured throughout the study using the Trofile™ assay ● Resistance to NRTIs and EFV was evaluated by the PhenoSense GT™ assay ● Resistance to MVC was evaluated using the PhenoSense Entry™ assay, with plateaus in maximum percent inhibition (MPI) <95% as a marker of resistance ● MVC plasma concentrations were determined using sparse PK sampling and combined with adherence data

6. Heera J, et al. 15th CROI 2008; Presentation 40LB 6 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Percentage of Patients with HIV-1 RNA <50 Copies/mL at Week 48 by Tropism Result at Baseline 10 20 30 40 50 60 70 80 90 100 339 33111 14 69.3 68.0 54.6 7.1 Patients (%) N= R5D/M 0 Includes all patients who received at least one dose of study medication MVC + CBV EFV + CBV MERIT Study – 48 weeks ●25 (3.5%) patients had D/M virus at baseline ●In both treatment groups the proportion of patients with D/M virus at baseline who achieved undetectable HIV-1 RNA was reduced relative to patients with R5 virus

7. Heera J, et al. 15th CROI 2008; Presentation 40LB 7 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Possible Correlates of Tropism Changes From Screening to Baseline: Combined Treatment Groups Tropism change from screening to baseline, N=24 R5 at screening and baseline, N=697 Mean screening HIV-1 RNA, log 10 copies/mL 4.794.84 Mean screening CD4+ count, cells/mm 3 (min, max) 201 (23, 431) 271 (3, 1,528) Viral subtype B C Other/undetermined 17 (4.2%) 4 (1.9%) 3 (4.2%) 390 (95.8%) 212 (98.1%) 68 (95.8%) MERIT Study – 48 weeks

8. Heera J, et al. 15th CROI 2008; Presentation 40LB 8 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Analysis of Virological Correlates Associated with Lack of Efficacy ●Virological analysis of patients who discontinued due to lack of efficacy –43 patients on MVC –15 patients on EFV

9. Heera J, et al. 15th CROI 2008; Presentation 40LB 9 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo R5 (32) Missing (1) NR/NP (2) DM/X4 (8) Tropism Baseline Tropism Failure R5 (13) (7) BLQ (9) (2) NR/NP (5) (2) D/M or X4 (16) NRTI res Tropism Shift and NRTI Resistance at Failure for Patients on MVC (n=43) 27* NR/NP = no result/non-phenotypable BLQ = HIV-1 RNA <500 copies/mL tropism tests were cancelled or censored for these samples *Excludes two patients with no resistance data at the last timepoint on treatment, but with resistance at earlier visits

10. Heera J, et al. 15th CROI 2008; Presentation 40LB 10 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo RTI Resistance Selected in Virus from Patients with Treatment Failure Tropism at failure* MVC (300 mg BID) NnM184V2NRTIres R5 D/M or X4 Other 22 19 2 22 19 2 10 (45%) 19 (100%) 0 2 (9%) 5 (26%) 0 Total43 29 (67%)7 (16%) Tropism at failure* EFV (600 mg QD) Nn EFV res +M184V +2NRTI res R5 Other 14 1 13 1 8181 3131 0101 Total15149 (64%)4 (29%)1 (7%) * Last valid tropism result while on treatment N = total patients in group; n = total patients with a valid resistance test; 2NRTI res = genotypic resistance to 3TC and ZDV (or substituted NRTI); EFV res = genotypic resistance to EFV; Other = missing baseline tropism result (n=1) or no valid tropism data during failure (n=2)

11. Heera J, et al. 15th CROI 2008; Presentation 40LB 11 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo R5 (32) Tropism Baseline Tropism Failure R5 (11) (4*) BLQ (6) (5) NR/NP (5) (2) D/M or X4 (10) (1) PK BLQ Incomplete Adherence is a Significant Contributor to Treatment Failure in Patients with R5 Virus at Baseline 50% NR/NP = no result/non-phenotypable BLQ = HIV-1 RNA <500 copies/mL tropism tests were cancelled or censored for these samples PK BLQ = MVC plasma concentrations below the limit of quantification corresponding to a rebound in viral load * Includes one patient with no BLQ PK but with documented interruption between visits corresponding to rebound in VL.

12. Heera J, et al. 15th CROI 2008; Presentation 40LB 12 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Characteristics of Patients who Failed 300 mg BID MVC with R5 Virus Country RaceSexClade MVC resistance* NRTI resistance** PK BLQ Belgium BlackMAGYesM184V Argentina WhiteMBFYesM184V, M41M/L Poland WhiteMBM184V USA WhiteMBM184V South Africa BlackMCM184V Australia WhiteMBYes Puerto Rico WhiteMB(Note 1) South Africa BlackMCYes South Africa*** BlackMCYes South Africa BlackFCYes UK WhiteMB South Africa OtherMC *By phenotype;**By genotype; *** Tropism was recorded as “BLQ” at last on-treatment timepoint, but ‘R5’ at all other timepoints PK BLQ = serum levels of MVC during periodic sampling were below limit of quantification Note 1: documented non-adherence between visits coinciding with viral load rebound

13. Heera J, et al. 15th CROI 2008; Presentation 40LB 13 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Example Patient Profile 1 Treatment start Treatment end MPI 99 Time since first administration (day) CD4+ count (cells/mm 3 ) R5 MPI 92 1 2 3 4 5 6 0100300200 0 400 600 800 NR M184V FPV/r + TDF + 3TC MVC + CBV R5 HIV-1 RNA (log 10 copies/mL) FPV = fosamprenavir; TDF = tenofovir

14. Heera J, et al. 15th CROI 2008; Presentation 40LB 14 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Mean Change from Baseline in CD4+ Count (cells/mm 3 ) by Tropism Result at Baseline and Time of Failure MVC + CBV EFV + CBV NR = not-reported, non-phenotypable, below limit of quantification, or missing N=0 N=15 R5  R5R5  D/M or X4 R5  NR non-R5  AnyTotal N=6 N=9N=10 N=42N=11 Mean change from baseline in CD4+ count (cells/mm 3 )

15. Heera J, et al. 15th CROI 2008; Presentation 40LB 15 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Tropism Summary ● 13 patients (3.8%) receiving MVC had a change in tropism result from R5 to D/M between screening and baseline – The response to MVC was significantly reduced in this subgroup – Tropism changes were 50% less frequent in patients with clade C HIV-1 than in patients with clade B or other clades ● For subjects with R5 virus at baseline, no appreciable difference in treatment response was seen between the MVC and EFV treatment groups ● A retrospective analysis will be performed to investigate the impact of an enhanced tropism assay on MERIT outcomes ● CXCR4-using virus was detected at failure in 10/32 (31.3%) MVC-treated (300 mg BID) subjects with R5 virus at baseline ● Patients failing MVC had higher CD4+ cell counts at failure than EFV, irrespective of tropism result at failure

16. Heera J, et al. 15th CROI 2008; Presentation 40LB 16 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Resistance Summary ● Lamivudine resistance mutation (M184V/I) was most common in patients failing in the MVC arm – 19/19 patients failing with D/M or X4 virus – 10/22 patients failing with R5 virus ● EFV-related resistance mutations were most common in patients failing in the EFV arm – 9/14 patients with resistance test data ● Resistance to MVC in patients failing with R5 virus was uncommon – 2/12 patients studied ● Viral load rebound in patients failing the MVC arm was more commonly associated with BLQ plasma levels of MVC at the time of failure

17. Heera J, et al. 15th CROI 2008; Presentation 40LB 17 1/9/2007 - 730pmeSlide - P3591 - Template - Blue - No Logo Acknowledgements ● The patients participating in the MERIT study ● The investigators and study site staff ● The Pfizer MERIT study team