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

776 Abstracts 41169-41173

12th World AIDS Conference

41169 Comparison of mycobacteria growth indicator tube and
Lowenstein-Jensen medium for the recovery of
Mycobacterium tuberculosis from clinical specimens in
Senegal
Edgard Adam Macondo1, Cheikh Saad-Bouh Boye1, Fatou Ba2,
Gaye-Diallo Aissatou1, Ndeye Coumba Toure-Kane1, Ndiaye Aston Gueye1,
Souleymane Mboup1. 'PO. Box 7325 Dakar, Laboratoire De Bacteriologie
Virologie, Dakar; 2Programme National de Lutte Contre La Tuberculose Dakar,
Senegal
Objectives: To compare the rate and mean time for the recovery of M. tuberculosis in an African laboratory where Lowenstein-Jensen (LJ) is the commonly
used medium for the isolation of M. tuberculosis with a simple and nonradioactive
liquid medium MGIT.
Method: 531 specimens included 495 sputa from patients suspected to have
TB were collected from peripheral laboratories in Dakar and Thies. Specimens
were processed, inoculated and read on both LJ and MGIT as recommended by
the manufacturers.
Results: Among 531 specimens, 121 were microscopy positive (M+) and the
rest were microscopy negative (M-). A total of 173 (32.5%) isolates were obtained
including 103 M+ and 70 M-. LJ recovered 95 (54.9%) whereas MGIT recovered
159 (91.9%) (p < 0.01). LJ failed to recover 78 (45.0%) isolates and MGIT failed
to recover 14 (18%). Among 103 isolates M+, LJ recovered 66 (64.0%) and MGIT
recovered 98 (95.1%). Among 70 isolates M-, LJ recovered 29 (41.1%) and MGIT
61 (87.1%). The mean times for the recovery of isolates were 20.1 (8-45) days in
LJ and 10.5 (2-49) days in MGIT. For M+ isolates, 17.3 days in LJ and 7.8 days
in MGIT. For M-isolates, 21.4 days in LJ and 14.9 days in MGIT.
Conclusion: MGIT has shown a better rate and shorter mean time for the
recovery of M. tuberculosis in both M+ and M- specimens than those of LJ. Its
short mean time and high sensitivity allow the earlier diagnosis of TB wich is
important particulary in developing countries for carrying susceptibility tests.
41170 Community-acquired bacterial and fungal infections in
HIV-positive patients hospitalized in AIDS Diagnosis and
Therapy Center in Warsaw, Poland
Janusz Starczak', E. Czekalska-Lachowicz2, G. Skowrohska2,
U. Komorowska2, E. Tobolewska2, A. Lipniacki3, A. Horban3. 1 Wolska St. 37,
Warszawa; 2Hospital for Infectious Diseases, 3AIDS Diagnosis & Theraphy
Center, Warszawa, Poland
Objective: the aim of presented study was to analyze community-1 acquired
infections (recognized in first 48 hours of hospitalization) of HIV-positive patients.
Material and Methods: 3146 specimens obtained from 734 HIV-positive patients, hospitalized in AIDS Diagnosis and Therapy Center in Warsaw between
January 1996 and December 1997, were tested. Cerebrospinal fluid, blood,
urine, sputum, bronchoalveolar lavages and swab samples from rectum, ulcers,
abscesses and wounds were analyzed. Microorganisms were identified with
culture and biochemical tests, in some cases followed by serological analysis
(latex or slide agglutination). Mycobacterium sp. and CI. difficile detection was
supported and verified with Polymerase Chain Reaction technique. In this case
PCR amplified fragments of DNAs were identified with electrophoretic analysis or
hybridization with probes, detected in ELISA test.
Results: In 5 cases out of 12 positive cerebrospinal fluid Cryptococcus neoformans was recognized, next was mycobacterial infection - 4 cases. Out of 38
positive blood samples 9 were infected with S. aureus. Leading agent of 181
urinary tract infections was E. coli - 37 cases and Enterococcus sp. - 32 cases.
Fungal infections were the most common in bronchopneumonia - in 162 out of
457 cases C. albicans or other Candida were recognized, second agent was S.
aureus (49). In 63 cases of ulcer, abscesses or infected wounds dominated S.
Aureus - 25 isolates.
Conclusions: (1) Infections of respiratory system and urinary tract were the
most common in analyzed group of HIV-infected patients.
(2) In our study Candida albicans, S. aureus and E. coli were recognized
as leading agents of community-acquired infections of HIV-positive hospitalized
patients.
41171 Feasibility of HIV rapid test in Mozambique: Technical
aspect
Jose-Fernando Langal, Guillaure Guiraud2, S. Tembe3, E.V. Chicovele3.
'Banco de Sangue, Hospital, Central Maputo, Av. Agostino Neto, Maputo;
2French Cooperative-Technical Assistant, Maputo, MZ; 3Technician, Maputo,
Mozambique
Introduction: The routine HIV antibody screening test used in Mozambique
are, one ELISA test (Genelavia Mixt, SANOFI) and one rapid test (HIV Spot,
Genelabs). As we showed in a precedent study, HIV Spot find about 2% of false
negative results. In many districts of Mozambique only HIV Spot is available to
screen for HIV infection.

Aims: to evaluate the feasibility, technical aspect and cost effectiveness of 8
HIV rapid test as an alternative of HIV Spot.
Method: saliva and sera samples from 18 known AIDS patients were tested
(with their consent). Saliva samples were collected by Orascreen (Beacon) and
sera were processed by 7 different rapid tests: HIV Spot, Capillus (Camb. Biotech),
Multispot (SANOFI Pasteur), Genie II HIV 1/2 (SANOFI Pasteur), Immunocomb
II Biospot (PBS Org.), Recombigen (cambridge Biotech) and Test Pack (Abbott).

For each assay, was collected complete information like contain, presentation,
principle, storage, material required, reader variability, time to perform, ease to
perform, commercial costs per test, desavantage to perform). Dilutions for 2
specimens were processed like the others Each test was interpreted by two
independent persons from the laboratory.
Results: All sera were detected reactive in Western Blot against HIV1 antigen
and all rapid test were positive for all samples and dilution. Only 13 showed a
reactivity with saliva test; the main problem can be the variability of the quantity
of saliva collected with the device. About reader variability, Capillus and Multispot
showed the best results with 0%. Multispot and Genie II need extra-material. The
less easy to perform are Genie II and Bispot. In term of time to perform, the worst
is Recombigen (40 minutes). HIV Spot can be stored at ambiente temperature.
Discussion: HIV Spot is the first test in term of feasibility essentialy because
of its cost, time to perform and storage (at 250C). Because of some lakes of HIV
Spot with HIV2 or HIV1/0 antibodies, it will be interested to estimate sensibility
and specificity of all rapid test in district conditions, including all storage and staff
problem. The best way for the future is probably to change the kit but in the same
time, the strategy of blood transfusion.
141172 Prospective evaluation of patients with AIDS and
abdominal surgery
Eduardo Ferat', G. Guzman-Valdivia2, L. Rosales2, G. Luna2, S. Treviho2,
L. Nieto2, M. Santoscoy3. 'Merida 108, 06700 Distrito Federal Mexico City;
2Hospital General Gabriel Mancera, Ditrito Federal; 3Hospital General Gabriel
Mancera, Mexico City, Mexico
Objectives: To asses the utility of surgery in pts. with AIDS and abdominal
disease, and to know the diseases of the abdomen that lead to surgery of patients
thus having a clinical outcome.
Design: Prospective, longitudinal, descriptive and observational study.
Methods: Adults with AIDS having the following symptoms were included:
Abdominal pain, life threating gastrointestinal bleeding (GIB), or abdominal tumor. All patients were evaluated by the same surgeon, who decided to performe
surgery based on the following symptoms: Peritoneal irritation, palpable mass or
specific organ disease demonstrated by imaging studies. In those cases in which
there was not any specific finding, liver, ileum, and lymph node biopses were
performed. Presumptive preoperative and definitive postoperative diagnosis, as
well as the clinical outcome was recorded. Patients were followed for 30 days
after the surgery.
Results: From Jan 97 to Jan 98, 25 pts were evaluated and eigthteen fullfilled
the inclusion criteria. 16 were males and 2 females with and average age of 34.13
~ 5.7. Sixteen pts were in C3 stage and two in B3. Preoperative diagnosis was:
Acute abdominal pain in 9 pts; Lithiasic Cholecystitis in 2; Acalculous Cholecystitis
in 2 and Tumor, Cholangitis, Appendicitis, Lower GIB and Abscess (one case
each). An incidental duodenal perforation was the only transoperative complication. Four patients had to taken to surgery again because of early complications.
The definitive diagnosis was: Intestinal Lymphoma in 4 pts., Mycobacterial infection in 4 (3 intestinal, 1 hepatic), Intestinal CMV in 3, and Splenic abscess,
acute appendicitis, acute cholecystitis, inespecific ileitis, mesenteric adenitis and
xantogranulomatous cholecystitis (one case each). There were two deaths, the
rest of the patients had a satisfactory outcome.
Conclusions: The most common causes of abdominal pathology in our patients, were Lymphoma, Tuberculosis and CMV. Specific diagnostic tests should
be performed in order to exclude these pathologies.
411731 Impact of human immunodeficiency virus type 1 (HIV-1)
subtype on HIV antibody detection in Burkina Faso
Philippe Van De Perre', Lucile Gautier-Charpentier1, O.R. Ouedraogo2,
S.A. Simonon', M.N. Meda1, D.H. Dahourou3, S.R. Soudre4, B.F. Barin5.
1 Centre Muraz BP 153, Bobo-Dioulasso; 2National HIV Reference
Laboratory, Ouagadougou; 3Blood Center Transfusion, Bobo-Dioulasso;
4National Comity for AIDS Control, Ouagadougou, Burkina Faso; 5URHC
Bretonneau Virology Laboratory, Tours, France
Objective: To identify a potential impact of HIV-1 subtype variability on five Enzyme Linked Immuno-Sorbent Assays (ELISA) screening tests results, in Burkina
Faso.
Methods: In the context of a study on HIV testing algorithms, two hundred
and sixty sera collected among blood donors, hospitalized patients and pregnant
women in Bobo-Dioulasso and Ouagadougou (Burkina Faso) were identified
as HIV antibody positive by five commercial ELISA (ICE HIV-1.0.2 and Murex
HIV-1/HIV-2 from Murex, Genelavia Mixt from Sanofi Diagnostics Pasteur, Enzygnost Anti HIV-1/2 Plus from Behring and Vironostika Uni-Form II from Organon
Teknika) and were confirmed by Western Blot (WB). These sera were tested with
a home made Peptide Enzyme Immunoassay (PEIA) for subtyping. This test uses
specific HIV-1 V3 loop peptides to discriminate subtype A, B, C, D and E. All the
results were analysed for each subtype.
Results: Among the 260 positive sera, 64 (24.5%) were serotype A, 120

(46.1%) serotype C, and 2, 3 and 1 (0.8, 1.2 and 0.6%) were serotype B, D
and E respectively. Fifty-three (20.3%) presented cross reactions mainly between
subtypes A and C and 17 (6.5%) were non typable. Optical densities (OD) were
overflow for all tests except for Vironostika Uni-Form II where OD mean was
2.414. Among non typable sera, four samples had a nearly complete WB profile
and an ICE HIV-1.0.2 OD which was not overflow and ranged from 0.341 to 1.636.
These samples will be submitted to genotyping (HMA and sequencing).