Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

324 Abstracts 22306-22309

12th World AIDS Conference

women >40 years old, 0/21 with cervical cancer and 6/85 (7%) controls were
HIV-1-seropositive (p = 0.5); however, among women aged <40 years, 4/8 (50%)
with cervical cancer and 4/31 (13%) controls were HIV-1 seropositive (OR = 6.8
95% Cl = 0.9-56.6; p = 0.04). HIV-1-seropositive women with cervical cancer
had a mean CD4 count of 400 cells/ttl, and they were significantly younger than
HIV-negative women with cervical cancer (mean age = 32 vs. 47 years, p =.02).
Despite their younger age, (80%) of HIV-1-positive women had clinical stage III
or IV cervical cancer compared with 76% of HIV-negative women.
Conclusions: These preliminary results suggest that in women aged less than
40 years cervical cancer appears to be associated with HIV-1 infection. Given
the advanced stage of disease for most women with cervical cancer, and the
HIV prevalence rate of 14% in sexually active women in Abidjan, routine cervical
cancer screening should be considered as a public health priority for all women
in Abidjan.
623* / 22306 Incidence of squamous intraepithelial lesions (SIL)
in HIV-infected women
Marie-Christine Delmas1, A. Agarossi2, C. Bergeron3, F. Meier4, J. Paavonen6,
J.D. Poveda5, J.A.R. Van Den Hoek7. 1European Centre on AIDS, 14 Rue val
D'osne 94415 St. Maurice Cedex; 3Cerba-lpeca, Pontoise; 4Louis Mourier
Hospital, Colombes; 5CBMS Institut Pasteur, France; 2L. Sacco Hospital, Milan,
Italy; 6 University of Helsinki, Helsinki, Finland; 7Municipal Health Service,
Amsterdam, Holland
Objectives: To study risk factors associated with occurrence and evolution of SIL
in HIV(+) women.
Methods: The analysis was based on data collected in a multicentre European
cohort of HIV(+) women with a known date of infection. Women were followed
every 6 months with gynaecological examination and Pap smear test, and HPV
detection by Southern blot (SB) and PCR was done at inclusion.
Results: By January 1998, 512 women were included in the cohort and interpretable Pap smear results were available for 485 women. At inclusion, the
prevalence of SIL was 15.9% (low grade: 12.0%, high grade: 3.9%) and atypical
squamous cells of undetermined significance (ASCUS) were evidenced in 10.3%
of the women. HPV was detected in 50.7% of the 404 women with available
HPV detection results. In the 234 women with initial normal Pap smear and
followed-up, the cumulative probability of SIL at 18 months after inclusion was
15.1% and was correlated with the immunosuppression level (from 11.1% if >500
CD4// L to 32.8% if < 200). The cumulative probability of SIL at 18 months was
higher in HPV(+) (up to 54.4% for HPV types 16, 18 or 33 detected by SB) than
in HPV(-) women (9.6%). After adjustment for HPV infection and age at inclusion,
the relative risk of developing a SIL, as compared with women with >500 CD4//IL,
was 1.5 [95% CI: 0.8-3.1] in women with 200-500 CD4, 2.3 [0.9-5.6] in those
with < 200 CD4 and antiretroviral therapy, and 5.4 [1.7-17.6] in those with <
200 CD4 and no antiretroviral therapy. Adjusted for these factors, no difference
in the occurrence of SIL was seen according to the time since HIV infection.
The cumulative probability of regression from a low grade SIL was 34.0% at 1
year: 41.2% if >500 CD4/pIL, 37.3% if 200-500, and only 21.6% if < 200. The
cumulative probability of progression from a low to a high grade SIL was 11.4%
at 1 year.
Conclusion: A high incidence and a low regression rate of SIL were observed
in HIV-infected women with severe immunosuppression. Our results suggest a
preventing effect of antiretroviral therapy in the occurrence of SIL in severely immunosuppressed women. The potential effect of HIV viral load in the occurrence
of SIL and responses to treatment for cervical intraepithelial neoplasia remain to
be assessed in these women.
S22307 Cost-effectiveness of screening for anal squamous
intraepithelial lesions in HIV-infected men
Sue J. Goldie1, K.M. Kuntz1, K.A. Freedberg2, M. Welton3, M.C. Weinstein1,
J.M. Palefsky3. 1718 Huntington Avenue, Harvard School of Public Health 2nd
fl., Boston, MA; 2Boston University School of Medicine, Boston, MA; 3University
of California, San Francisco, San Francisco, CA, USA
Objectives: To evaluate the benefits (life expectancy), costs (dollars), and costeffectiveness (dollars per year of life saved) of screening for anal squamous
intraepithelial neoplasia (ASIL) in homosexual or bisexual HIV (+) men.
Methods: We developed a state transition Markov model which incorporates
incidence, likelihood of progression and regression of ASIL, operating characteristics of screening tests, natural history of HIV, competing mortality risks, and
costs of screening, treatment, and follow-up for ASIL and HIV. Data were obtained
from literature, national databases, and a large prospective study of the natural
history of ASIL in homosexual and bisexual HIV (+) men. Strategies included:
(1) No screening; (2) Biennial anal Papanicolaou smears (Pap); (3) Biennial anal
anoscopy; (4) Annual Pap; (5) Annual anoscopy; (6) Semiannual Pap; (7) Semiannual anoscopy. Selected input data for men with CD4 200-500 cells/pl included
monthly incidence of ASIL (3%), sensitivity and specificity of anal Pap smear (81%
and 63%), relative risk of progression (2.5 times that of HIV- men), cost of anal

Pap ($26) cost of anoscopy ($140). Due to uncertainty in the estimates of high
grade ASIL (HSIL) progression to invasive cancer and the treatment efficacy for
HSIL, we made several assumptions to bias the analysis against screening.
Results: Screening men with CD4 cell counts between 200-500 cells/tl was
associated with discounted life expectancy gains between 1.9 and 2.6 months.
Annual Pap smears had a cost-effectiveness (CE) ratio of $12,600 per year of
life saved (YLS) compared to no screening. The additional benefit conferred by

increasing the screening frequency to twice a year with either Pap or anoscopy
was minimal ( 4 days). Results were most sensitive to the progression rate of
HSIL and efficacy of treatment. Anoscopy became more efficient than anal Pap
smears when the sensitivity of Pap was <60% or the specificity was <80%.
Conclusions: Screening HIV (+) men at high risk for ASIL with annual anal
Pap smears is likely to provide significant life expectancy gains. The C/E ratio of
annual Pap smears ($12,600/YLS) is comparable to other well accepted preventive medical interventions in HIV disease, such as prophylaxis with TMP/SMX for
Pneumocystis carinii pneumonia ($13,000/YLS), prophylaxis with azithromycin
for Mycobacterium avium complex ($31,000/YLS), and cervical cancer screening
in HIV-negative women ($180,000/YLS).
S22308 Screening HIV-infected women for cervical intraepithelial
neoplasia with serial PAP smears
Jonathan A. Cohn1, S. Gagnon2, M. Spence3, C. Brinson4, A. Stein5,
J. Hellinger6. 1 University of Maryland Medical Center, Baltimore, MD; 2University
of Kansas, Wilchita, KS; 3Hahnemann University, Philadelphia, PA; 4Central
Texas Medical Foundation, Austin; 5Community Research Initiative, Coral
Gables, FL; 6Community Research Initiative, Brooklyn, NY, AMFARS Cervical
Disease Study, New York, USA
Objectives: The USPHS/IDSA prophylaxis guidelines recommend initial screening for cervical intraepithelial neoplasia (CIN) with two Pap smears obtained over
6 months followed by annual smears. We compared semiannual cervical cytology
with annual colposcopic biopsies in order to examine this strategy.
Methods: In this prospective cohort study, HIV infected women with initial CD4
counts < = 500 cells/cu mm and neither prior cervical cancer nor treatment for
cervical dysplasia in the preceding two years had cervical smears taken at the
initial, 6 month and 12 month visits. Cervical biopsies were obtained at the initial
and 12 month visits. Women with CIN I at entry were followed with more frequent
cytologic and colposcopic exams. Those with high grade CIN were referred for
treatment.
Results: Of 109 women with initial colposcopic biopsies, 80 had no CIN, 17
had CIN I and 12 had CIN II/111. Repeat biopsies were obtained 6-12 months after
the initial visit on 51 of the 97 eligible women. Follow-up biopsies revealed CIN
in 2 of 8 women with CIN on initial biopsy (25%), in 3 of 16 with HPV/condyloma
on initial biopsy (19%), in 6 of 24 with inflammation or other benign findings on
initial biopsy (25%), and 0 of 3 women with normal initial biopsies. Among the
43 women with no CIN on initial biopsy, CIN was diagnosed on follow-up in 6
of 13 women (46%) with an epithelial cell abnormality on repeat cytology, and
3 of 30 (10%) women with no epithelial cell abnormality (p = 0.01). Sensitivity
of 1 or 2 follow-up Pap smears was 67% (95%CI 33, 91%) and specificity 79%
(95%CI 63, 91%). In this population, the incidence of CIN was 21% over one
year, and the negative predictive value of repeated Pap smears was 90% (95%CI
73, 98%).
Conclusions: CIN developed frequently among women who were without dysplasia by colposcopic biopsy at baseline. Given the high annual incidence of CIN
(21%) and modest negative predictive value of repeated Pap smears (90%) in
our cohort, we believe that continuous intensive screening for cervical disease is
prudent.
1 22309 Squamous intraepithelial lesion (SIL) in HIV-infected
women
Bercina Candoso1, J.M. Cabral2, R. Magarinho2, E.C. Fernandes2, A.A. Horta2,
O. Vasconcelos2, R. Sarmento E Castro2. 1Rua Espinho; 2Porto, Portugal
Objective: To evaluate frequency of cervical Squamous Intraephitelial Lesion
(SIL) in HIV-infected women; To evaluate immunosupression as cofactor, in the
development of SIL; To compare cervical smears to colposcopic evaluations, in
diagnosing percursors of cervical cancer; To evaluate the importance of HPV
typification.
Methods: We have observed 108 HIV-infected women, with an average age
of 28.5 years (range 19-50). In all, we performed cervical smears, colposcopic
evaluation and, when necessary, colposcopically directed cervical biopsies (n =
94). We classificate as Low Grade SIL (LoSIL) HPV infection and CIN1; as High
Grade SIL (HiSIL) CIN2 e CIN3. We have also investigated the presence of
factors that might be implicated in SIL ocurrence: age of first intercourse, number of sexual partners, partners risk smoking habits, previous sexual transmited
diseases (STD) and CD4 cell count. We have also performed HPV typification in
cytology and/or in biopsy, in every case presenting SIL.
Results: Cervical disease was present in 87% and SIL in 78.7% (LoSIL - 74%,
HiSIL - 26%) of this population. The average age found for LoSIL was 28.4 years
and for HiSIL 30.3 years - we have diagnosed HiSIL in a very precocious age
(19 years old). We found 50.5% of false-negative rate in cytologic screening for
SIL, from that, 9.4% were HiSIL. We verified also that 18.6% of LoSIL in cervical
smears were HiSIL, in histological diagnosis. We found that the frequency of SIL
increases with all the cofacors that we have above-mentioned and with severe
immunosupression (CD4 > 500 = 59% SIL; CD4 200-500 = 78% SIL; CD4 <

200 = 82.5% SIL). We found different HPV types in 70.5% cases (85% HPV of
high-grade of malignancy).
Conclusions: There is a higher frequency of SIL in this population, determinated by the degree of immunodeficiency, among other cofactors. The high
false-negative rate of cytologic screening in predicting SIL suggests that cytology
alone is not adequate in this group. The strategy is performing repeated cervical
smears and colposcopic evaluation should be considered in HIV positive women,