dc.description.abstract | Correct diagnosis of HIV depends on HIV variants that are in circulation
in a particular community as well as the type of HIV test kit that is applied. HIV
serotypes have been found to vary from one geographical area to another.
Commercial HIV test kits used in Kenya are made up of products of HIV variants
prevalent in Europe and The USA
In an effort to determine sensitivity of the common test used in Kenya an
evaluation of six such kits together with an analysis of serotypes of HIV prevalent
in Nairobi was undertaken. The kits were manufactured by Organon Teknika
(Vironistika Uniform Il), Wellcorne (Murex), Sanofi Pasteur (Genelavia), PBS
Orgenics (Irnrnunocornb), Genelabs (HIVspot) and Cambridge Biotech (Capillus).
Evaluation was done ust,pg -a panel of 72 serum samples of local origin.
Kits based on antigens derive? frorrf the core and envelope region and those made
up of whole viral lysate showed higher sensitivities (92% and 90.3%) than kits
made up of antigens from the envelope region only. Seroconversion samples
were not detected by all the test kits. Haemolysis of the blood samples decreased
sensitivity of Genelavia (80%) and Capillus (86%) while freeze-thawed samples
decreased sensitivity of all the kits except Murex. It is inferred that kits based on
antigens representing the HIV envelope region only were less sensitive than kits
based on antigens from more than one antigenic region of the virus.
On using one of the more sensitive kits (Vironistika Uniform II) for HIV testing,
out of 3339 patients screened, 588 were confirmed as HIV positive. However
five samples derived from spouses of confirmed Hl V positive patients were
ELISA negative. On Western blotting they had antibodies against only the core
proteins of mv and could not therefore be assumed to be Hl V positive. Further
analysis using PCR denoted they were HIV positive. The existing criterion for
positive Western blot interpretation was limited to include these samples as HIV
positive. It is apparent that Vironistika Uniform II did not detect HfV infection in
spouses because they may have been in a seroconversion stage.
A panel of peptides representing the V3 loop of envelope gp120 derived
from seven isolates (MN, HXB2, CDC4, SC, Z2, Z6 and ELI) were used for
serotypic characterisation through a self-made ELISA. Among the Hl V positive
samples the MN serotype was the most prevalent (21.5%) followed by Z2
(19.1%) and HXB2 (17.2%). SC, CDC4, Z6 and ELI had prevalences of 11.5%,
6.2, 5.3% and 3.8% respectively. Some serum samples (14.8%) did not show
reactivity to any of the peptides meaning that they may belong to other serotypes.
To determine the immune status of HIV positive patients, CD4+ cell count
was carried out by Flow Cytometry (Facscan, Beckton Dikinson) and the absolute
count correlated with reactivity to the panel of peptides. A majority of patients
(77%) with a CD4+ cell count of 200 x 106 cells per m!. and above had sera
reactive to three or more peptides. In contrast, 79% of the sera that did not react
with any peptide was derived from patients with a CD4+ cell count of 200 x 106
and below. It was suggested that serotypes with amino acid sequense glycineproline-
glycine-arginine (GPGR) at the V3 loop of gp 120 are more prevalent in
Nairobi than serotypes with amino aside sequence glycine-proline-glycineglutamine
(GPGQ). This is in contrast with results obtained through genotyping
studies. Asymptomatic HIV positive patients had a higher degree of serotype
cross-reactivity while the symptomatic patients had a lower rate of serotype crossreactivity.
It is recommended that polymerase chain reaction be used for clinical
decision making as a confirmatory procedure in cases where serological outcome
is negative but where HIV infection is strongly suspected. It is also recommended
that further research be carried out to supplement serotyping in order to
confirm the env and Kag sequences of HIV isolates in Kenya that may be
incorporated into HIV testing kits i12lri'crease their sensitivity. | en |