1. The CAMELIA trial CAMbodian Early vs. Late Introduction of Antiretrovirals ANRS 1295/12160 - CIPRA KH001/10425 trial 22 nd July 2010 Late Breaker Session B-1, XVIII IAS Conference, Vienna, Austria F.X. Blanc, T. Sok, D. Laureillard, L. Borand, C. Rekacewicz, E. Nerrienet, Y.Madec, O. Marcy, S. Chan, N. Prak, C. Kim, K.K. Lak, C. Hak, B. Dim, C.I. Sin, S. Sun, B. Guillard, B. Sar, S. Vong, M. Fernandez, L. Fox, J.F. Delfraissy, A.E. Goldfeld.

2. START TB TREATMENT AND HAART SIMULTANEOUSLY START TB TREATMENT FIRST AND DELAY HAART PROS Lower risk of HIV disease progression or death in advanced patients (CD4 < 50 cells/mm 3 ) Avoid overlapping side effects Avoid PK interactions Lower pill burden Lower risk of IRIS CONS Overlapping side effects PK interactions Higher pill burden Risk of immune reconstitution disease Higher risk of HIV disease progression or death in advanced patients (CD4 < 50 cells/mm 3 ) Adapted from J Acquir Immune Defic Syndr 2007; 46: S9-S18. HAART in TB-HIV: Early or late?

3. 2003: CD4 < 200/mm 3 : - Start TB treatment. - Start ART as soon as TB treatment is tolerated (between 2 weeks and 2 months) - Efavirenz-containing regimens 2010:- Start ART in all HIV-infected individuals with active TB, irrespective of the CD4 cell count. Strong recommendation, low quality of evidence. - Start TB treatment first, followed by ART as soon as possible afterwards (and within the first eight weeks). Strong recommendation, moderate quality of evidence. - Use efavirenz as the preferred NNRTI in patients starting ART while on TB treatment. Strong recommendation, high quality of evidence. WHO recommendations

4. CAMELIA study design (2003-2004) - Prospective, randomized, open-label, two-armed trial with no placebo - Designed as a superiority trial to answer the question of the best timing for the introduction of HAART in severely immunosuppressed (CD4 ≤ 200/mm 3 ) HIV-infected adult patients with newly diagnosed TB in Cambodia - 2 arms: late introduction of ART (reference arm: 8 weeks) vs. early (2 weeks) introduction of the same HAART - Primary endpoint: survival at the end of the trial (intent-to- treat analysis) ANRS 1295/12160 - CIPRA KH001/10425 study

5. CAMELIA strategy Switch D4T to AZT ANRS 1295/12160 - CIPRA KH001/10425 study

6. - 2 sponsors: French ANRS and U.S. NIH/DAIDS (CIPRA) - Partnership with Cambodian Health Committee - 5 study sites (rural and urban) in Cambodia - 661 patients were AFB+ at inclusion (pulmonary or extra- pulmonary TB) with CD4 ≤ 200/mm 3 - 1 st patient enrolled on January 31 st 2006 - 6 DSMB meetings - Last patient enrolled on May 27 th 2009 - End of the study: May 2010 CAMELIA key points ANRS 1295/12160 - CIPRA KH001/10425 study

7. 661 patients randomized 282 culture + M.tb 38 culture - 332 randomized to the EARLY arm 329 randomized to the LATE arm 778 patients screened 12 NTM 294 culture + M.tb 31 culture - 4 NTM 117 patients not enrolled due to: - CD4>200 (n=78) - LFT impairment (n=24) - pregnancy (n=3) - TB treatment >1month (n=2) - CD4 >200 & LFT impairment (n=2) - death before randomization (n=2) - pregnancy & LFT impairment (n=1) - no CD4 at enrolment (n=1) - high bilirubine (n=1) - delay in blood sampling (n=1) - CD4>200 & pregnancy (n=1) - ART history & LFT impairment (n=1) CAMELIA recruitment M.Tb: Mycobacterium tuberculosis; NTM: nontuberculous mycobacteria ANRS 1295/12160 - CIPRA KH001/10425 study

8. Early arm (N=332)Late arm (N=329)p Gender Male Female 215 (64.8) 117 (35.2) 210 (63.8) 119 (36.2) 0.80 Age, years Median (IQR)35 (30 – 41)36 (30 – 42) 0.38 BMI, kg/m 2 Median (IQR)16.7 (15.3 – 18.3)16.8 (15.2 – 18.6) 0.90 Karnofsky score ≥80 50-70 ≤40 43 (13.0) 259 (78.0) 30 (9.0) 44 (13.4) 251 (76.3) 34 (10.3) 0.83 CD4, cells/mm 3 Median (IQR)25 (11 – 56)25 (10 – 55) 0.61 Viral load, log copies/mL Median (IQR)5.60 (5.20 – 6.02)5.66 (5.25 – 6.00) 0.25 Patient characteristics at enrollment ANRS 1295/12160 - CIPRA KH001/10425 study

9. Early arm (N=320) Late arm (N=325) p Location of TB Pulmonary Pulmonary & extra-pulmonary Extra-pulmonary 221 (69.1) 71 (22.2) 28 (8.7) 222 (68.3) 73 (22.5) 30 (9.2) 0.97 Drug resistance None Isoniazid (INH) monoresistance Streptomycin monoresistance Rifampin monoresistance INH polydrug resistance Multidrug resistant (MDR) No DST Missing 217 (67.8) 23 (7.2) 17 (5.3) 3 (0.9) 16 (5.0) 6 (1.9) 37 (11.6) 1 (0.3) 240 (73.8) 10 (3.1) 4 (1.2) 24 (7.4) 7 (2.2) 30 (9.2) - 0.10 Characteristics of tuberculosis ANRS 1295/12160 - CIPRA KH001/10425 study

10. NDeathsFollow-up time*Mortality rate** (95% CI)p Early arm33259712.48.28 (6.42 – 10.69) 0.002 Late arm32990653.713.77 (11.20 – 16.93) * expressed in person-years ** per 100 person-years 12 patients (1.8%) lost to follow-up. 8,955 protocol visits, <2% missed visits. SIGNIFICANT REDUCTION OF MORTALITY IN THE EARLY ARM ANRS 1295/12160 - CIPRA KH001/10425 study

11. Log-rank p-value: p=0.0042 Survival probability (95% CI) Early armLate armLog-rank p-value Week 5086.1 (81.8 – 89.4)80.7 (76.0 – 84.6)0.07 Week 10082.6 (78.0 – 86.4)73.0 (67.7 – 77.6)0.006 Week15082.0 (77.2 – 85.9)70.2 (64.5 – 75.2)0.002 Kaplan-Meier survival curves ANRS 1295/12160 - CIPRA KH001/10425 study

12. Multivariate analysis Adjusted HR (95% CI)* p ArmEarly Late 1 1.52 (1.12 – 2.05) 0.007 BMI≤16 16-17 17-18.5 >18.5 1.68 (1.07 – 2.63) 0.93 (0.53 – 1.60) 1 1.11 (0.66 – 1.87) 0.01 Karnofsky score≥80 50-70 ≤40 1 1.78 (0.97 – 3.26) 4.96 (2.42 – 10.16) <0.001 TB identification and location* Pulmonary Extra-pulmonary Pulm. and extra-pulm. NTM 1 1.19 (0.68 – 2.07) 2.26 (1.62 – 3.16) 2.84 (1.13 – 7.13) <0.001 Drug resistanceNo** Yes Yes, MDR 1 0.98 (0.63 – 1.51) 8.02 (4.00 – 16.07) <0.001 * Also adjusted for site and CD4 level at baseline (stratification factors) Factors independently associated with mortality Cox proportional hazard model ANRS 1295/12160 - CIPRA KH001/10425 study

13. NPR/IRISFollow-up time*Incidence** (95% CI)p Early arm3321102 728.54.03 (3.34 – 4.86) <0.0001 Late arm329483 333.51.44 (1.09 – 1.91) IRIS significantly more frequent in the early arm ANRS 1295/12160 - CIPRA KH001/10425 study * expressed in person-months ** per 100 person-months Time after TB treatment initiation (weeks)

14. W26W50W78W102W126W150 Early arm Undetectable VL VL > 250 copies/mL Not available 259 (90.2) 25 (8.7) 3 (1.1) 264 (95.6) 9 (3.3) 3 (1.1) 184 (93.0) 6 (3.0) 8 (4.0) 173 (93.5) 7 (3.8) 5 (2.7) 142 (93.4) 6 (4.0) 4 (2.6) 111 (95.7) 3 (2.6) 2 (1.7) Late arm Undetectable VL VL > 250 copies/mL Not available 241 (88.3) 25 (9.2) 7 (2.5) 237 (95.6) 9 (3.6) 2 (0.8) 145 (91.8) 8 (5.1) 5 (3.1) 143 (92.9) 5 (3.2) 9 (3.9) 126 (96.2) 3 (2.3) 2 (1.5) 96 (96.0) 1 (1.0) 3 (3.0) p0.800.820.340.810.640.63 Plasma viral load (VL) measured by real time PCR for HIV-1 RNA plasmatic quantification (ANRS kit). >95% undetectable viral load at week 50 ANRS 1295/12160 - CIPRA KH001/10425 study

15. W26W50W78W102W126W150 Early N Median (IQR) 283 65 (26 – 125) 273 118 (67 – 191) 189 169 (101 – 270) 180 194 (134 – 299) 148 210 (128 – 324) 115 230 (152 – 321) Late N Median (IQR) 265 59 (14 – 111) 247 112 (53 – 175) 153 165 (88 – 243) 148 177 (106 – 285) 129 187 (119 – 288) 97 201 (127 – 322) p0.110.220.810.190.570.51 Median CD4 increase at week 50: 114/mm 3 Week 0: median CD4+ cell count was 25/mm 3 ANRS 1295/12160 - CIPRA KH001/10425 study CD4 increase from baseline

16. 1. Mortality was reduced by 34% when HAART was initiated 2 weeks vs. 8 weeks after onset of TB treatment. 2. Irrespective of study arm, HAART has been extremely successful, as evidenced by >95% of patients with undetectable viral load. 3. Despite extremely low CD4+ cell count at inclusion, patients enrolled in this pivotal strategic trial have been extremely adherent. 4. HAART initiation 2 weeks after onset of TB treatment could potentially save 150,000 of the 450,000 annual HIV- TB deaths. CONCLUSIONS ANRS 1295/12160 - CIPRA KH001/10425 study

17. Sponsors: ANRS and NIH/DAIDS Cambodian Health Committee Institut Pasteur du Cambodge Médecins Sans Frontières – Belgium Cambodian Ministry of Health Cambodian National TB Program (CENAT) Cambodian National AIDS Program (NCHADS) Study sites: Khmer-Soviet Friendship Hospital (Phnom Penh), Donkeo Provincial Hospital (Takeo), Calmette Hospital (Phnom Penh), Svay Rieng Provincial Hospital and Siem Reap Referral Hospital Investigators, nurses, technicians, monitors, social workers… Members of the DSMB and the Scientific Advisory Board And especially all the patients and PLWHA representatives who joined us in this challenge. ACKNOWLEDGEMENTS ANRS 1295/12160 - CIPRA KH001/10425 study