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

134 Abstracts 13240-13244

12th World AIDS Conference

39% in adm/pt, and 33% in hospital charges/pt. Decreases for OD, MORT, and
HCU were even greater between 1996 and 1997 (data not, shown). Although
decreases in LOS and hospital charges occurred as a result of improved standards of inpatient and outpatient care before HAART, changes in OD, MORT, and
hospital adm occurred after and reflect the impact of implementation of HAART.
13240| Population-based surveillance for cryptococcosis in the
US: 1993-94 vs. 1996-97
Rana Hajjeh1, D. Rimland2, M. Sattah2, S. Nash1, E. Graviss3, R. Hamill3.
'CDC; DBMD; MS C-09, 1600 Clifton Rd., N.E. Atlanta, GA 30333; 2Emory
University and VAMC, Atlanta, GA; 3Baylor College of Medicine and VAMC,
Houston, TX, USA
Background: Cryptococcosis (CC) is the most common life-threatening fungal
opportunistic infection in persons with AIDS. Azole drugs, known to be protective
against CC, have been increasingly used recently. In addition, since licensure
in 1995, protease inhibitors have significantly improved management of HIV
infection, and led to a decrease in incidence of opportunistic infections.
Objectives: To examine trends in occurrence of CC following changes in
management of HIV infection.
Methods: We reinitiated population-based active surveillance for CC in two
areas of the US: metropolitan Atlanta and Houston (population 6.3 million), where
we had also conducted surveillance during 1993-94. An incident case was defined by the first isolation of Cryptococcus neoformans from any clinical site or
positive latex agglutination test on serum or CSF.
Results: During the recent surveillance period (7/1/96 to 6/30/97), we detected
191 CC cases corresponding to an annual incidence (per 1,000 persons living
with AIDS) of 18 in Atlanta compared to 31 in 1993-94 (p-value < 0.001), and
10 in Houston compared to 27 in 1993-94 (p-value < 0.0001). Of CC cases
identified in 1996-97, 86% (n = 168) occurred among persons with AIDS (mean
CD4 count = 39), compared to 91% in 1993-94 (p-value > 0.05). The proportion
of persons who were male (88% vs. 92%) in 1996-97 and 1993-94, respectively,
and the case-fatality rates (10% vs. 9%) were not significantly different (p-value
> 0.05). However, 65% (110) were black, compared to 50% (164) in 1993-94
(p-value = 0.001).
Conclusions: There has been a substantial decrease in the incidence of CC
since 1993-94, likely due to the increase in use of azoles and protease inhibitors.
These data also suggest that black persons with AIDS in these areas had less
access to routine medical care and chemoprophylaxis. Continued surveillance
for CC is important to monitor trends in disease occurrence and the impact of
therapeutic changes in the management of HIV infection.
S13241 Effect of antiretroviral and other antiviral therapies on the
incidence of Kaposi's sarcoma and trends in Kaposi's
sarcoma
Jeffrey Jones, D.L. Hanson, M.S. Dworkin, J.W. Ward, H.W. Jaffe. t Centers of
Disease Control; Mailstop E-47 CDC 1600 Clifton Rd, Atlanta, Georgia 30333,
USA
Background: Antiretorviral therapy may decrease the incidence of Kaposi's sarcoma (KS) by slowing the progression of HIV disease and other antiviral therapy
has been associated in some studies with a lower incidence of KS. We evaluated
the effect of antiretroviral and other antiviral therapies on the incidence of KS
in a large multicenter HIV/AIDS surveillance system and examined trends in KS
incidence.
Methods: We used data abstracted from medical records in over 90 hospitals and clinics 1/90-7/97 in 9 US cities that participate in the Adult/Adolescent
Spectrum of HIV Disease (ASD) project. Incidence was calculated per 100 person-years (py) and the effect of therapies (time dependent [td]) on risk for KS
was calculated by using multivariate Poisson regression which controlled for sex,
race, age (td), HIV exposure mode, CD4+ cell count (td), and calendar year (td).
Antiretroviral therapy was defined as either mono- or dual (mono/dual) therapy
or as triple therapy. Acyclovir, ganciclovir, and foscarnet were the other antiviral
therapies evaluated.
Results: Of 30,293 HIV-infected persons in the study representing 54,573
person years, 82% were men, 56% were nonwhite, and by HIV-exposure mode,
63% male-male sex (including those who injected drugs), 15% injecting drug use,
8% heterosexual contact, and 14% other exposure mode or no mode recorded.
Overall, KS incidence declined an estimated 10% per year (observed incidence
4.8/100 py in 1990 to 1.5/100 py in 1997). Mono/dual therapy (n = 21,080; relative
risk [RR] =.87; 95% confidence limit [CI],.78-.97), triple therapy (n = 802; RR =.41; 95% CI,.20-.83), and foscarnet (n = 550; RR =.59; 95% CI,.351-.997) but
not acyclovir or ganciclovir, were independently associated with a decreased risk
for KS.
Conclusion: In our study population, antiretroviral therapy and foscarnet, which
has in vitro activity against HHV-8 (the putative etiologic agent of KS), were associated with a lower risk for KS. The incidence of KS is declining in this large US
population and should continue to decline as new antiviral agents are developed

and used widely.

S13242 1 Risk of presentation with cryptosporidiosis as an
AIDS-defining condition in California
Asheena Khalakdinal, F. Tabnak2, T. Haight3, R. Sun2, J. Colford3. 1U.C.
Berkeley School of Public Health and CA. Dept. of Health Services, Berkeley,
CA; 2Calif Dept. of Health Services, Sacramento, CA; 3U.C. Berkeley School of
Public Health, Berkeley, CA, USA
Objectives: To identify risk factors associated with the presentation of cryptosporidiosis as an AIDS-defining condition.
Design: Case-control study.
Methods: Surveillance data from California's HIV/AIDS Reporting System from
1982-1997 were analyzed. The cases were 1,300 patients with cryptosporidiosis
as an AIDS-defining diagnosis and the controls were 89,814 patients with other
AIDS-defining conditions. Gender, race/ethnicity, sexual orientation, injection drug
use (IDU), and CD4 count at diagnosis were among the variables analyzed by
stratified analysis and multivariate logistic regression.
Results: This study suggests a significantly lower risk among African-Americans compared with other ethnicities for presentation Additionally, there is evidence that homosexual/bisexual males are more likely to present with cryptosporidiosis than heterosexual males. Finally, our analyses show that females
are at a lower risk of presenting with cryptosporidiosis than males. These effects
were not altered by adjustment for sexual orientation, gender, and injection drug
use, and CD4 count at AIDS diagnosis.
White (59,533) OR (95% Cl) Black (13,787) OR (95% Cl)
Male Gay Non-IDU 1.00 (referent) 0.60 (0.47, 0.75)
IDU 1.01 (0.81, 1.26) 0.67 (0.42, 1.08)
Non-Gay Non-IDU 0.46 (0.28, 0.76) 0.29 (0.16, 0.54)
IDU 0.74 (0.42, 1.31) 0.17 (0.00, 0.99)
Female Non-Gay Non-IDU 0.33 (0.13, 0.85) 0.14 (0.00, 0.50)
IDU 0.22 (0.09, 0.58) 0.18 (0.00, 0.64)
NOTE: Male, Gay, Non-IDU is the reference group for all other groups
Conclusions: These results confirm reports from earlier, smaller studies. Immunologic differences or differential exposure to Cryptosporidium among the
subgroups may account for the differences observed. These striking differences
must be borne in mind when designing trials of candidate therapeutic agents for
cryptosporidiosis. One limitation of this study may be the use of retrospective
surveillance data with potentially incomplete reporting.
13243 Characteristics of HIV-associated histoplasmosis in
Argentina
Jorge Benetucci, Messina Garcia, O.M. Bases, S.M. Oliva, A. Maranzawa,
L. Redini, L. Puga. Hospital MuhFiz, Ward 17/FUNDAI Foundation, Buenos Aires,
Argentina
Objective: To analise the epidemiologycal and clinical characteristics of histoplasmosis (H), the value of diagnosis methods, the relationship of CD4 cells count
and delay (D) in the diagnosis with global mortality and the possibility of surviving
this clinical event.
Method: Seventy patients who had H were enrolled and studied thoroughly
between 1983 and 1997. 93% were male and the average age was 32.46% were
IVDU, 29% homo/bisexual and 12% heterosexual.
Result: Three clinical forms were observed: cutaneous (CH), disseminated
(DH) and mucosal (MH). In the first case, diagnosis was done by scraping (71%)
or biopsy (19%). DH were diagnosed by blood culture (52%), biopsies of different
tissues (30%), bone marrow culture (3%) or two associated methods (25%).
Mucosal forms were diagnosed by scraping of mucosal lesions in 80% of the
cases.

CI. forms
Cutaneous
Dissem.
Mucosal

~

n
32 (45%)
33 (47%)
5 (7%)

~

X CD4 X Day D
106 61
67 49
78 49

~

X Survival
10 mo.
5 mo.
10 mo.

~

Dead
16 (50%)
24 (73%)
1 (20%)

There were not significant differences in mortality of CH and DH as regards CD4
cells count or D on diagnosis. The global mortality was not significantly different
between those clinical forms (p = 0.06). 4/16 patients with CH and 9/33 with DH
died of H (p = 0.0015) while only 1/5 with MH died because of the mycosis.
Conclusion: Patyients with CH had greater CD4 cells count and D than patients with DH or MH. The best method of diagnosis was scraping for CH and
MH and the blood culture for DH. It was more likely for patients with DM to die
because of the mycosis than for patients with CM.
13244 Opportunistic infections in seropositive patients in
private hospitals in Nairobi, Kenya
Surendra Patel, D.M. Owilli. PO Box 39779, Nairobi, Kenya
Objective: To determine the nature of opportunistic infection present in HIV
seropositive patients.
Study Design: Five years retrospective controlled study between January 1993
to December 1997.
Methods: 1. The Medical records of five hundred adult patients with confirmed
HIV infection were retrieved and scrutinized. 2. The following data was extracted:
(a)-Age, sex, race, occupation and marital status (b)-Clinical manifestations with