Program Supplement [International Conference on AIDS (14th: 2002: Barcelona, Spain)]

load(VL) and country. io endpoint was time to 2nd 3-drug
regimen failure(RF). RF included virologic failure(VF), toxicity,
or premature discontinuation. 20 endpoints: time to ist RF or
1st VF. Hazard ratios(HR, 95% CI) from Cox regression with
Bonferroni correction.
Results: Pts were followed a median of 28 months. Median
baseline CD4 278, VL 4.9 log, age 36; 72% men, 53% nonwhite. Effect of starting with ZDV+3TC vs ddl+d4T on 10
endpoint depended on whether subjects started with EFV or
NFV (interaction test, p=0o.o2). Interaction was stronger for two
20 endpoints: time to 1st RF (p=o.oo2), time to 1st VF
(p=o.ooo6), and resulted from benefits of ZDV+3TC+EFV
compared to other 3-drug regimens. Analyses of 10 endpoint
adjusted for two comparisons showed trend favoring starting
with (a)ZDV+3TC vs (b)ddl+d4T in combination with (c)EFV
(HR=o.68; 0.46-1.o01) but not (d)NFV and of starting with (c) vs
(d) in combination with (a) (HR=o.71; o.48-1.o6) but not (b). 20
endpoint comparisons significantly favored (a) & (c) for time to
1st RF (HR=o.35; 0.22-0.58)(HR=o.40; 0.25-0.66) and ist VF
(HR=o.39; 0.24-o.66)(HR=o.34; 0.21-0.56), respectively.
Median time to severe or dose-modifying toxicity was 104 wks
on ZDV+3TC arms and 52 wks on ddl+d4T arms.
Conclusion: Starting with ZDV+3TC vs ddl+d4T showed a
strong trend to delay sequential RF and significantly delayed
1st RF and ist VF when combined with EFV but not NFV.
Starting with EFV vs NFV showed a strong trend to delay
sequential RF and significantly delayed 1st RF and ist VF when
combined with ZDV+3TC but not d4T+ddl.
Corresponding author: Robbins, Gregory, MGH ID-Unit, 55 Fruit
Street, GRJ-5o4, Boston, MA 02114, USA, United States,
Tel: +i 617 724-0076, Fax: [phone],
Email: [email protected]
[ LBOR2OB1
Antiretroviral strategies in naive HIV+ subjects: comparison
of 4-drug versus sequential 3-drug regimens (ACTG 384)
Sha er Robert1, Robbins, Gregory', Smeaton, Laura', De
Gruttola, Victor", Pettinelli, Carla1, Snyder, Sally, D'Aquila,
Richard1, Vella, Stefano2, Hirsch, Martin', Merigan,
Thomas', the ACTG Team, for ('United States; 2Italy)
Intro: 6-arm study of 980 nalVe Pts. 2X3 factorial design:
A. ZDV+3TC vs d4T+ddl, B. EFV vs NFV, C. 3 vs 4 drugs. This
abstract compares sequential 3-drug regimens vs single 4-drug
regimens (objective C). Design: 4-drug arms: ddl+d4T+NFV+EFV
(n=178); ZDV+3TC+NFV+EFV (n=182). 3-drug arms:
ddl+d4T+EFVZDV+3TC+NFV; ddl+d4T+NFVZDV+3TC+EFV;
ZDV+3TC+EFV dd+d4T+NFV; ZDV+3TC+NFVddI+d4T+EFV
(n=155/arm). NFV & EFV were blinded. Randomization and
analyses were stratified by viral load (VL) & country. 10
endpoint: time to failing a single 4-drug regimen or two
sequential 3-drug regimens. Regimen failure (RF) included
virologic failure (VF), toxicity, or premature discontinuation. 20
endpoints: time to ist RF or 1st VF. Hazard ratios (HR, 95% CI)
from Cox regression with Bonferroni correction.
Results: Patients were followed a median of 28 months.
Median baseline CD4 278, VL 4.9 log, age 36; 72% men, 53%
non-white. There was no evidence of difference in time to 10
endpoint between single 4-drug regimens compared to two
sequential 3-drug regimens. 20 endpoint analysis showed that
4-drug regimens delayed time to 1st RF (HR=o.53; o.39-o.71)
and 1st VF (HR=o.45; o.33-o.61) compared with NFV+2 NRTIs
but not EFV+2NRTIs. 4-drug regimens delayed time to 1st RF
(HR=o.63; 0.40-0.99) and ist VF (HR=o.57; 0.35-0.93)

compared with EFV+d4T+ddl but not EFV+ZDV+3TC. Severe or
dose-modifying toxicity was lower among regimens containing
ZDV+3TC compared with ddl+d4T.
Conclusions: Overall, there were no significant differences in 10
endpoint between single 4-drug and two sequential 3-drug
regimens. 4-drug regimens significantly delayed 1st RF and 1st
VF compared to NFV-containing 3-drug regimens regardless of
NRTIs used. 4-drug regimens significantly delayed 1st RF and
1st VF compared to EFV-containing 3-drug regimens only in the
ddl+d4T containing arms.
Corresponding author: Robbins, Gregory, MGH ID-Unit, 55 Fruit
Street, GRJ-5o4, Boston, MA 02114, USA, United States,
Tel: [phone], Fax: [phone],
Email: [email protected]
[ LBOR211
Effect of HAART on incidence of anal intraepithelial neoplasia
grade 3 among HIV-positive men who have sex with men
Palesk aelaHolly, Elizabeth, Ralston, Mary, Jay, Naomi, Berry,
Michael, Darragh, Teresa (United States)
Background: HAART has led to declines in the incidence of
Kaposi's sarcoma and non-Hodgkins lymphoma but its effect on
HPV-associated anogenital cancers is unclear. The incidence of
anal cancer is 37-fold higher among HIV+ MSM compared with the
general population of men. Anal cancer most likely is preceded by
anal intraepithelial neoplasia (AIN) 3. The purpose of this study
was to determine the effect of HAART on the prevalence and
incidence of AIN 3 among HIV+ men who have sex with men
(MSM).
Methods: Prospective cohort study of AIN among HIV+ MSM at
the University of California, San Francisco. At baseline and
every 6 months, a questionnaire was administered to obtain
data on behavioral and medical risk factors for AIN including
use of HAART. At each visit anal cytology and high-resolution
anoscopy with biopsy of visible lesions were obtained. The
level of AIN was classified using the more severe result of
cytology and histology.
Results: Of 433 HIV+ men enrolled at baseline, 15% were not on
any HAART regimen. 12% were diagnosed with AIN 3 at baseline.
The risk of prevalent AIN 3 was 2.6 (95% Cl 1.3-5.3) for men on
HAART compared with those not on HAART after adjustment for
CD4+ level. Over a 24-month follow-up period 51% (95% Cl 43 -59) of HIV+ men on HAART developed AIN 3 compared with 37%
(95% Cl 25-50) of men not on HAART (p=.11)
Conclusion: The prevalence and incidence of AIN 3 were higher
among HIV+ men on HAART compared with HIV+ men not on
HAART even after adjustment for CD4+ level. HAART does not
reduce the incidence of AIN 3 among HIV+ MSM. The proportion
of men with AIN 3 who will progress to cancer if left untreated is
unknown. AIN 3 may take several years to progress. Our data
suggest that the incidence of anal cancer will continue to rise in
the HAART era as men with AIN 3 live longer. Screening
programs to identify and treat AIN 3 before progression to
cancer should be considered for HIV+ MSM even if they are on
HAART.
Corresponding author: Palefsky, Joel, 505 Parnassus Ave,
Room M12o3 Box 0126, UCSF, San Francisco, CA 94143,
United States, Tel: [phone], Fax: [phone],
Email: [email protected]

XIV International AIDS Conference BARCELONA - JULY 7-12

27