Highlights of the

XIV International AIDS ConferenceJuly 7-12, 2002; Barcelona, Spain

Selected and summarized by

Joseph J. Eron, Jr, MDAssociate Professor of Medicine

University of North Carolina at Chapel Hill

Supported by an unrestricted educational grant from

T-20: Phase 3 Studies in Highly Experienced Patients

Study Outline

• Two studies: TORO 1 (Americas) and TORO 2 (Europe and Asia)

• Highly treatment-experienced subjects:– Median 7.4 yrs prior therapy

– 75% prior AIDS-defining illness

– Median plasma HIV-1 RNA >100,000 copies/mL

– Median CD4+ cell count ~ 100 cells/mm3

• Over 600 subjects, randomized to change therapy to either:– New regimen optimized by genotype or phenotype, or

– Optimized regimen plus T-20

• Primary end point: change in plasma HIV-1 RNA

• 24 weeks follow-up

Abstracts: LbOr19A/B

T-20: Phase 3 Studies in Highly Experienced Patients (2)

Results at Week 24

• Significantly greater CD4+ cell count increases in the T-20 arms

• Injection-site reactions were the most frequent AE on T-20, although overall discontinuation rates between treatment arms were similar

Study Viral Load Reduction (log10 copies/mL)

T-20 + Optimized Regimen Optimized Regimen P value

TORO 1 -1.70 -0.76 < .001

TORO 2 -1.43 -0.65 < .001

Tenofovir DF in Treatment-Naive Subjects

Study Outline

• Gilead 903 study: randomized, double-blind, placebo-controlled

• Treatment-naive subjects (N = 600)– plasma HIV-1 RNA level > 5000 copies/mL (median = 81,300)

– any CD4+ cell count (median = 279)

• Randomized to initiate therapy with either:– tenofovir DF (QD) and stavudine placebo (BID), or

– stavudine (BID) plus tenofovir DF (QD)

– each combined with open-label efavirenz (QD) + lamivudine (BID)

Abstract: LbOr17

Tenofovir DF in Treatment-Naive Subjects

Results at Week 48

• Equivalent virologic and immunologic outcomes:– HIV-1 RNA < 50 copies/mL in 81% to 82% of both arms (ITT, M=F)

– CD4+ cell count increases of 167-169 cells/mm3 in both arms

• Similar rates of adverse effects:– More peripheral neuropathy on stavudine

– Lactic acidosis: 3 on stavudine, none on tenofovir DF

– Triglycerides: +74 mg/dL on stavudine; no change on tenofovir DF

– Cholesterol: +53 mg/dL on stavudine; +25 mg/dL on tenofovir DF

• Question: Resistance pattern after TDF/3TC/EFV failure?– Will K65R mutation appear more commonly in previously naive patients?

• Question: Long-term effects of regimens on metabolic parameters, eg, fat redistribution, bone density?

ACTG 384: Head-to-Head Comparison of 6 Initial Regimens

Study Outline

• Factorial design:– Zidovudine/lamivudine (ZDV/3TC) vs stavudine/didanosine (d4T/ddI) as NRTI

backbones

– Efavirenz (EFV) vs nelfinavir (NFV) vs both as the additional agent(s)

– Comparison of sequential 3-drug vs single 4-drug therapy

First-line Regimen Second-line Regimen

Stavudine + didanosine + efavirenz Zidovudine + lamivudine + nelfinavir

Stavudine + didanosine + nelfinavir Zidovudine + lamivudine + efavirenz

Zidovudine + lamivudine + efavirenz Stavudine + didanosine + nelfinavir

Zidovudine + lamivudine + nelfinavir Stavudine + didanosine + efavirenz

Stavudine + didanosine + efavirenz + nelfinavir Not applicable

Zidovudine + lamivudine + efavirenz + nelfinavir Not applicable

Abstracts: LbOr20A/B

ACTG 384: Head-to-Head Comparison of 6 Initial Regimens (2)

Primary End point

• Time to failure after exposure to all 3 classes, ie:– Time to failure of the 4-drug regimen

– Time to failure of the second 3-drug regimen

Baseline Characteristics

• N = 980

• 72% male, 47% white

• Median baseline CD4+ cell count = 278 cells/mm3

• Median baseline plasma HIV-1 RNA level = 4.9 log10 copies/mL

• Median follow-up = 28 months

ACTG 384: Head-to-Head Comparison of 6 Initial Regimens (3)

Results

• Factorial design thwarted by interactions between the regimen components:– Activity of efavirenz differed with ZDV/3TC vs d4T/ddI

– Activity of ZDV/3TC differed with nelfinavir vs efavirenz

• In the arms receiving sequential 3-drug regimens:– Time to first failure was substantially longer with ZDV/3TC/EFV

– Time to second failure appeared substantially longer if either the first or second 3-drug regimen was ZDV/3TC/EFV

• Comparing sequential 3-drug vs single 4-drug regimens:– No additional benefit from receiving nelfinavir with ZDV/3TC/EFV

• Significantly more toxicity from d4T/ddI vs ZDV/3TC backbone

• No significant differences in CD4+ cell count responses

• Results strongly support use of ZDV/3TC/EFV as initial therapy and suggest that the ddI/d4T backbone may be suboptimal

PIs Associated With More Cardiac Events Than NNRTIs

Study Outline

• Multicenter, randomized trial in Italy:– 776 patients treated with a PI-containing regimen

– 775 patients treated with a non-PI regimen

• End points: new myocardial infarction or new-onset angina

Baseline Characteristics

• Average age = 36 years

• Median CD4+ cell count at baseline = 325 cells/mm3 (range, 170-850)

• 87% smokers

Follow-up

• After 3 years, 587 PI and 621 non-PI patients remained on original regimen

Abstract: WeOrB1307

PIs Associated With More Cardiac Events Than NNRTIs (2)

Results

• 23 episodes of new heart disease in PI group– 6 transmural MI, 6 non-Q-wave MI, 11 new-onset or unstable angina

• 2 episodes of new heart disease in non-PI group – 1 MI, 1 angina

• Highly statistically significant difference in heart disease-free survival between arms– Incidence of heart disease in PI group > 10-fold that of the non-PI group, and

> 50-fold that of the general population

• Heart disease associated with lipodystrophy signs (OR = 26.9), elevated lipids (OR = 14.2), smoking (OR = 9.7), male sex, and HIV treatment.

• Nonblinded study, so the risk of an ascertainment bias (ie, cardiac events were sought more carefully in PI-treated patients) should be kept in mind

FRAM Study: Defining Lipodystrophy

Study Outline

• Aim: Compare randomly selected HIV-infected subjects and healthy controls and identify statistically significant differences and any linkages between components of lipodystrophy

• Three types of evaluation:– Self-report re: body habitus changes

– Clinical evaluation of presence/degree of visible lipoatrophy

– Body composition measures including whole-body MRI and DEXA scanning

• N = about 1200 HIV-infected subjects from 16 US clinics and 300 controls

• Subjects and controls were well matched for baseline characteristics, but subjects were lighter than controls

• This report focused on only a subset of enrolled men, not all subjects

Abstract: TuOrB1140

FRAM Study: Defining Lipodystrophy (2)

Initial Results

• Subjects were more likely to have lipoatrophy when compared with controls, by all 3 measures

• Nonuniform pattern of subcutaneous lipoatrophy:– Loss from legs > arms > lower trunk > upper trunk

• Visceral adipose tissue nonsignificantly lower in subjects vs controls

• No linkage between changes in amounts of subcutaneous vs visceral fat

• No significant difference in prevalence of “buffalo hump” in subjects vs controls

• Rates of accumulation of visceral fat or “buffalo hump” fat could not be compared in this cross-sectional study

Same-Clade Superinfection: Research and Health Implications

Case Report

• Anecdotal case described by Bruce Walker (Mass Gen)

• Subject enrolled in primary infection study:– Early initiation of HAART followed by sequential STIs

– Detailed longitudinal virologic and immunologic evaluations

• Evidence of superinfection– Subject initially controlled replication off-therapy for several months after final STI

> Subject had HIV-1-specific cytotoxic T lymphocytes directed at multiple epitopes

– Viral load increased after 1 month

– Genotyping indicated acquisition of a clade B virus that appeared genetically distinct from original infecting virus

– Patient admitted episode of unprotected sex

– No immunologic control of new virus; patient declined clinically despite some shared CTL epitopes between the initial and probable superinfecting HIV-1 variants

Abstract: WeOrA197

Same-Clade Superinfection: Research and Health Implications (2)

Implications

• Immune responses that were able to control initial strain could not prevent acquisition or control replication of a closely related second strain

• Alarming implications for vaccine research, currently focused on generating HIV-specific immune responses

• Superinfection with different HIV strains appears possible– Superinfected strain may be more pathogenic/drug-resistant

– Reinforces importance of prevention counselling of infected patients

Risk Factors for Progression in Naive Patients Starting HAART

Study Outline

• Pooled analysis of data from 12,574 patients in 13 cohort studies

• Intention-to-treat analysis of subjects starting 3-drug therapy, 80% with 2 NRTIs plus 1 PI

• Baseline characteristics:– Median age = 38 years

– 21% CDC stage 3 disease

– Median CD4+ cell count = 250 cells/mm3

– Mean plasma HIV-1 RNA = 4.9 log10 copies/mL

• Follow-up = 24,310 person-years:– 870 patients experienced at least 1 AIDS event

– 344 died

– 1094 either developed AIDS or died

Abstract: TuOrB1140; Lancet 2002;360:119-129.

Risk Factors for Progression in Naive Patients Starting HAART (2)

Results

• Risk of progression/death at 1, 2, or 3 years estimated based on baseline CD4+ cell count, log HIV-1 RNA, age, transmission group, and CDC stage

• Baseline CD4+ cell count < 200 cells/mm3 associated with highest risk of progression

• Baseline viral load was significant only if > 5.0 log10 copies/mL

• Other risk factors: age > 50 years; injection-drug use; CDC stage 3 disease

• Viral load at month 6 of therapy was a significant factor at all levels.

Comments

• Optimal CD4+ cell count (> 200) for therapy initiation remains unknown– Factors such as age, IDU, and high viral load may weigh toward early therapy

• Interactive risk calculator available online:– http://www.art-cohort-collaboration.org

Discontinuing HAART in Prematurely Treated Patients

Study Overview

• Views on initiating HAART have recently become more conservative: many patients now on treatment would not have met current criteria for starting

• Should such patients discontinue HAART?

• Johns Hopkins cohort of patients who interrupted HAART– N = 101

– 44% “prematurely treated”; 30% toxicity/nonadherence; 8% virologic failure

– HAART resumed if CD4+ cell count < 200 cells/mm3 or by physician/patient decision

Abstract: ThOrB1439

Discontinuing HAART in Prematurely Treated Patients (2)

Predictors of Successful Interruption

• 33/101 patients resumed HAART:– 25% had rising HIV-1 RNA; 24% had falling CD4+ cell count; 24% had both

– Resumers had lower mean CD4+ cell count and less viral suppression at time of interruption

• Pre-HAART CD4+ cell count (but not viral load) was strongest predictor of duration of interruption:– 7-fold greater likelihood of resumption if pre-HAART CD4+ cell count < 200 cells

mm3 (P = .001), compared with pre-HAART baseline CD4+ cell count > 500 cells/mm3

– 4-fold greater likelihood if pre-HAART CD4+ cell count = 200-350 cells/mm3 (P = .015)

• Patients who met current guidelines for therapy at time of interruption were 3-fold more likely to resume therapy than those who did not (P = .004)

Update: HIV Eradication by HAART Is Impossible

• Updated data from Siliciano (Johns Hopkins)

• One important reservoir = HIV-infected resting T cells – Transcriptionally silent, so invisible to immune surveillance

– Unaffected by current drugs

– Replenished by cell division, not from viral replication

– Half-life is at least 6 months, so decay rate is measured in decades

– No decrease in decay rate in optimally treated, chronically infected patients with no “blips” of viremia

– Even if this reservoir could be flushed, others exist

• However, long-term suppression appears possible– In patients with long-term suppression, virus released into circulation is wild-type

– No resistance even to drugs with low genetic barrier in many patients with long-term suppression

Abstract: MoOr103

CLASS Study of First-Line Regimens

Study Outline

• Study of antiretroviral sequencing; initial ITT analysis of 3 first-line regimens:– Abacavir/lamivudine/efavirenz (ABC/3TC/EFV)

– Abacavir/lamivudine/amprenavir/ritonavir (ABC/3TC/APV/RTV)

– Abacavir/lamivudine/stavudine (ABC/3TC/d4T)

• Planned interim analysis at 48 weeks

Baseline Characteristics

• N = 291– 70% were black or Hispanic

– Mean baseline viral load = 4.19 log10 copies/mL; 42% > 100,000 copies/mL

– 35% had CD4+ cell counts < 200 cells/mm3

– 28% had CDC stage B or C disease

Abstract: TuOrB1189

CLASS Study of First-Line Regimens (2)

Results

• Suppression to < 50 copies/mL highest in the efavirenz arm:– 76% vs 52% (APV/RTV) and 62% (d4T); P = .047

• Efavirenz also superior if baseline viral load > 100,000 copies/mL– 77% < 50 copies/mL vs 53% and 55%

• Similar increases in CD4+ cell count

• 26 (9%) premature discontinuations; only 5 (2%) due to AEs

• 6% abacavir hypersensitivity

Comments

• RT mutation pattern in failures with ABC/3TC backbone will be instructive

• Importance of regimen convenience cannot be underestimated

• Similar comparisons should be undertaken with more compact PIs, eg, fosamprenavir or atazanavir