Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Tuesday, July 9, 1996
possible that viral strain may have some role in the observed difi i iprogres
sion since the rapid progressors in the Glenochil cohort have a Or, r ) decline
than their matched controls.
I. BMJ 1995; 310: 289 292.
McMenamin J.The Scottish Centre for Infection & Environmental ei i i, It!),
Ruchill Hospital, Glasgow G20 9NB.Tel: 014 I -946-7 I 20. Fax:0 I41 I
email:[email protected]
Tu.C.432
THE PROGRESSION OF HIV INFECTION IN A COHORT OF I i\EI'iOPHIILIC MEN
FOLLOWED FOR UP TO 16 YEARS AFTER SEROCONVERSION
Sabin Caroline A*, Phillips AN*, Bhagani 5, Pasi KJ, Elford J*, Lee CA/. II!.,ii,i t 'iiCentre
and Departments of Haernostasis and *Primary Care and Populat r -, 1 oyal Free
Hospital, London, UK.
Objective:To describe the progression of HIV disease in a cohort cl iiric'nolhir. men
infected with HIV for up to 16 years, and to assess the impact of co if ictio i wit hepatitis
C virus (HCV) on morbidity and mortality.
Methods: IIll haemophilic men became infected with HIV between I9/9 and 11985 after
being treated with infected blood products.The men have now beren followed f spectively
for up to 16 years after HIV seroconversion. All men were infected wvith I K IV either before
or at the same time as infection with HIV. Cumulative progression rat, io ailSn, non AIDS
clinical events and death are estimated using Kaplan-Meier rnethoci-.
Results: By the Ist January 1996,53 of the men had developed AIDS,, cuuv:itve progression rate of 62% by 16 years after HIV seroconversion (median tin to the development
of AIDS, 14.2 years). Before developing AIDS, many of the patients!,we developed clinical
conditions related to HIV infection; mild bacterial infections (43%), slir poblenis (41 1%), oral
candida (22%), thrombocytopenia (14%) or herpes zoster (12%). t1.'f t-e ris ihave now
died, a cumulative death rate of 54% (median survival time, 14.3 yeirs). Of itti "rnen,
nine patients did not have an AIDS diagnosis at the time of death, fur o f v,.iho o died of
liver failure. Of the 45 patients with an AIDS diagnosis who have died, I! Cther 6 patients
died in liver failure.Thus, liver failure due to infection with I-HCV hi,e c.r, ibrute, 1to the
deaths of 10 men in this cohort. Of the 57 men currently alive, 8 have..I)S nrid 49 remain
AIDS-free, including 6 patients who have remained AIDS-fiee for o. ie th i irs. 36/19
of the rnen currently AIDS fiee and alive have developed some indi, ri 5 l their HIV
infection and the CD4 count has fallen below 50 cells/mrn3 in 7 (I1%.;), is be'en 200 and
50 cells/rnm3 in 8 (I 6%), between 500 and 200 cells/mmr3 in 26 (5>.,),ind rem ains above
500 cells/mrn3 in 8 (16%).
Conclusions: The rmedian time to development of AIDS in haernmophilir, r i r b ts t least 14
years. Co-infection with -ICV has lead to increasing morbidity arn d i-.ti h f)ir hver disease.
C.A.Sabin, Department of Primary Care and Population Sciences Ri HSMiI. cR wl ind Hill
Street, London NW3 2PF, England.lel. 0171 794 0500 ext. 4752. F:'. ) 17 i 79l 1224.
Email: [email protected]
Tu.C.433
INCREASED INCIDENCE OF CANCER AMONG HOMOSEXUAL MEN, NEW YORK
CITY AND SAN FRANCISCO, 1978-1990
H e olNancyA1, Koblin BA2, Zauber AG3,Taylor PE2, Buchbinldeir SP, K rz MH4, Stevens
CE2. UCSF, SF, CA,; 2NY Blood Center, NYC; 3Memorial Sloan I.tta inrr,, NYC; 4D'tept of
Public Health, SF, CA
Objective:To determine whether there is an increase in cancer irnac< rir rc as among
homosexual men.
Methods: The incidence of non-Kaposi's Sarcomna cancer ws exarn i i cid Ir or I9 i8 to 1 990
among 15,656 homosexual men who participated in studies of hep, titi hB virus infection in
the late 1970's in NewYork City and San Francisco. Cancers in the two 'iohorIts were identified through searches of the New York State and California State C ric'r Reg tr ies,
respectively, for 1978 to 1990.To assess the magnitude of the exc's ci decit i c ancer ir
the combined cohorts, the obser ved number of cancers was comp.i I:d io x1 r reected
number derived from annual US cancer incidence rates among adr!t rnal's fi the years
1984-88 from the Surveillance, Epidemiology and End Results (SEFR) P'r og., n.
Results: The standardized incidence ratio (SIR) for all cancers was.6 (9 co-ifidence
interval (C)= 1.4-1.8). Excesses were observed for Non-Hodgkin's I yll'ria ( LHL)
(SIR=-12.7; 95% CI= 11.0-14.6), Hodgkin's disease (HD) (SlR-2.5; 95% I I. )91 rid ansi
baI O -Id )an aa
cancer (SIR=24.2; 95%CI- 13.5-39.9). As seen with NHL, a cancer iii ir sVi ",,.,ociatecl
with HIV- I, HD incidence was significantly higher in more recent,5 cr r rin ed to earlier
years. No cases of HD were found among HIV-I antibody negative rier', ii illwas diagnosed near the time of initial AIDS diagnoses. Anal cancer incidenc e did N'I correlante with
HIV- I antibody status and did not tend to occur near the timoe oflSI t cir,irls.
Conclusions: This study confirms the association of NI-L with HIV-I i is, ii. Is ovides
additional evidence ol an association between I-D ansd HIV-I irifs.cti i r ir- i -,te irs anai
cancer was observed but did riot appear to be associated wiit f IfsI i 'r:, tI: ii hoc spsecific cancers, including liver ciricer, wse ri ot ini excess.
Nancy A. Hesnol, UCSF Box I1352, San Frasscisco, CA 941I43-1352,, r- '
Telephone: 41I5-502-6281I FAX: 41I5-4t76-8528 esrail: nanscyosunlsii.r:r,
Tu.C.434
THE PROBABILITY TO BE AIDS FREE BEYOND IQ0YEARS AFTER SF85 rCONVERSION.
Geskus, Ronald B., van Grienen, G.J.P,Albrechst-vasn Lent, N., Courririr. -:,'' ir i t,al
Health Service, department of Public Health and Ersirornmeni, Arrsr,L., -I i' iIsethser landcs.
Objective: Jo study the distribution of the timre fi-om seroconver a s 5' i n.:/1[; o'',ii long
time span in a group of ser ocoriverted hosrosexual men.
Methods: Eighty sine scroconverted hoisosexual sicn, or iginsting hi r:, - t i V w icrins'
trial conducted in Amsterdam between 1980 and 1983, were cnr il-cr irs sir is iif\IS
cohort study, and followed up to date.Thirty men hays lenigthss of 55;~ i: i:or s,,r:i,, lo h saelva
(time betwees last negative and] first positive test) of less shari one -e 5,i i i ire length
> 6 years.The msedians length is 2.3 years. We use two ssethods tlr 5 t a.. r-o,_,,rI i long

intervals, both based on the nonparametric maximum likelihood est,Iorru tinter val censored data. Method I imputes individual expected seroconversion dI, _, r,- r etirmite
of the distribution of seroconversion over time.Then a Kaplan-MeKirIi - rper
Tu.C.432 - Tu.C.440

formed. Cronfidence intervals are too small, since the uncertainty in the date of seroconversion is not taken into account. Method II uses the interval censored times from seroconversion to AIDS diretly Both methods were applied using both the 1987 and the 1993 CDC
/AIDS definition. For persons lacking CD4 counts the 1987 definition was used.
Results: In the rarge fIom 2.5 to I12.5 years, the estimated probability to be AIDS flee
according to mrethod I is almost linearly decreasing with time.The estimate at 12.5 years is
0.25 (95% Cl 0.15-0.41) for the 1987 and 0.18 (95% Cl 0.10-0.33) for the 1993 CDC definition. Method II gave probability estimates of 0.33 and 0.22 respectively One person was
still AIDS free after I5 years of follow up.
Conclusions: The estimates show a steady decrease in probability to be AIDS free over
time. Using the 1993 CDC definition supports the conjecture that the vast majority of peo
ple will develop AIDS within 20 years after seroconversion. Moreover, the estimator used
wit Srth'- 199) delrition is biased upwards, since only 29 persons had CD4 counts. Of
cour, e, [I is crnijecture is subject to the assumptior n thait te structure of the distribution
does rnot crha.ge beyond 12.5 years. Beyond 12.5 years, the estimated curve shows a tenderscy to flatten. IHowever, estimates get unreliable since little information is available.
G kus. Ronald B. Municipal Health Service, Nieuwe Achtergracht 100, 1018 WT
Ansterdan, the Netherlansds.lelephone: 31- 20-5555524 Fax: [phone]
Ernail: rgeskustcir uironet.nil
Tu.C.435
OBSERVED HIV- I DISEASE PROGRESSION TIMES IN GAY MEN WITH ACCESS
TO TREATMENTS
Hoover D.. Saiah A.J., Guccione M., Rosenberg PS., Chmiel J., Detels R. Kingsley L for the
Multicenter AIDS Cohort Study Baltimore, Chicago, Los Angeles, and Pittsburgh, the United
States of Arneric
Objective. stirate time to AIDS and death firom HIV- I seroconversion (HIV-SC) among
gay men with HIV SC dates starting three years before availability of AIDS treatments.
Methods: I,.rr hundred and eighty one (481) rmen were observed to seroconvert to HIV- I
fro i 198'i tc: 195 in the Multicenter AIDS Cohort Study (MACS).The exact date of HIV
SC was kncvrwn within six mronths for 408 or 85%. Cumulative AIDS and death incidence
were detnisrni-d with Kaplan -Meier methods. Relative risk (RR) regression evaluated effects
of demoiraphic rcvariates and other exposures on this incidence.The date of analysis was
Jantr iry I., 1995; excluding AIDS/death only 45 men (9.3%) were lost to follow-up more
tha one year before this date. Analysis of CD4 counts showed that this loss was unrelated
to stage of HIV I disease.
Results: A total of 183 AIDS diagnoses and 146 deaths were observed. Median times to
AfIDS arid deralh from HIV SC with 95% confidence intervals were 8.3 years (7.9,9.0) and
9.6 years (9. 1,1 0. I) respectively By two, four, six, eight and ten years respectively after HIVSC., (2.7/% and 0.6%), (13.6% and 8.1%), (29.4% and 18.0%), (46.8% and 33.8%) and (65.2%
Arid 55.6%) Iad developed AIDS and died, respectively Older age was associated with
more rapid progression to AIDS and death (RR= 1.27 and 1.34 per decade respectively);
education, race, last seronegative CD4 count and exposure to hepatitis B were not. Among
seroconver ters who had reached an indicated need for treatments by a CD4 count
<200/;: 78%,,18%, 70% and 54% respectively had used AZT other antiretrovirals, PCP
prophylaxis and acyclovir, respectively
Conclusions: The mredian time to AIDS (8.3 years) and death (9.6 years) in this closely followed cohort are two years less than many previous estimates for homosexual men of similar
ages.We believe the true incubation period has lengthened.This "apparent" decrease probably
reflects inpr oved scertainment of AIDS and death in our data over time. Continued evaluation of these individuals and other cohorts is needed to confirm these findings.
Donaldr R. 1-hoover; 624 N.Broadwiy, Rm 784, Baltimore, MD, 21205, the United States
lelephone (410)6i4-0748 Fax:[phone]
Tu.C.440
DECLINING MOTHER-TO-CHILD HIV TRANSMISSION FOLLOWING PERINATAL
ZIDOVUDINE RECOMMENDATIONS, UNITED STATES
Sirnonds R, Nesheim S, Matheson P, Abrams E,Vink P Palumbo P Steketee R. for the
Perinatal AIDS CollaborativeTransmission Study (PACTS), CDC, Atlanta, GA, USA
Objective: I evaluate the risk and risk factors for mother-to-child HIV transnission since
recor nmer datsrons for using zidovudine (ZDV) to prevent motherto-chilrd transmission
were published in August 1994.
Methods: Since 1986, the CDC Perinatal AIDS Collaborative Transmission Study (PACTS)
h-s enrolled IV- infected pregRnant women and their children in a prospective study of
mother -to-chid HIV transmission.We arnalyzed data on prenatal and newborn ZDV use,
naternal CD4 lymphocyte count within 90 days of delivery, maternal clinical status, date
of deivery, duration of ruptured membranes, and child infection outcome. Children who
estrd p 0ositive twice by virologic tests (IIV co-culture, DNA-PCR, p24 antigen) or by HIV
antibody after 15 months of a ge or had AIDS were considered infected; children without
AIDS wo ststed negative twice by virologic tests (>I1 safte r 2 months of age) or by HIV
antirodty test wrusss' considered urnirfected.
Results: T 1 Irnother--child pairs with cd infections outcenes, 235 (20%) wene infected.
COf 101'9 rhidreir resi'n before 9/1/94, 217 (21I%) were infected, cosmpared with 18 (1196) oh
f6l boin f sa ts (R h.9, 95%C d o.2-3.0). Rink for transmission was lower with prenart alio iirda si ZDV use (RRbu.7, 95%C10.5-0.9), aind Migher with delivery following
isirpuro '-, nrshn -"is's (RfOM) for _>'I hsuis (RR=s 1.7, 95%,C1c I.3-2.2), ma ternal CD4+ hyms
sliacyro cuint <'r0O cells/pul (R2.0, 95%C1=' I.'1--2.8), or smaternal Class C HIV disease
(Phi.:: -u7,' 9rCI I I -2.7).The piesahence of prenastal iris/or neonatal ZDV use increased
hs-on I!'. raiioi1 wmeen who delivered before 9/1/94 to 80% amntg these delivering
aftes mrds (p-Ot0 I ).The proportion of deiveniss occiring 4 hours after ROM decreased
duR ism tihs period (55% vs 37%, p 0.001 ), whereas te pr-oportions of women wit CD4+
lyiip u,oyl counts <500 cells/n sind with Classs C disease did sot change significantly.
Conclusinss:-ils -e snssther-to--chrld trasmisssion risk [s declined asnong PACFS ernrees
Icl-'uic', l;ublhSc'rto f thre 71 tV ss'csrmendatnons. Loer tansnsission risk may be due
snot rsry tr us ss iasirg ZDV cise, list slso to denresasing pirsva leisre of other sk factors, such
a lo, c:~lan asf ROM.
R.J. Srir odr tif) i600 ClIisftontod, Mistop E-45, Atlantta GA 303 33, USA
-tel: 3!0-3 I S.C3:; Fix: 1-404-639-61 18; Enrai: rxs5Sacidh-ivl.ern.cdc.gov

'O
0
u
c
nO
0
(L')
u
C
C
0
U
C
0
c
nO
C
248