Typing of HIV strains in Bangladesh

[Guest guest]

HIV strains in Bangladesh

Recent research at ICDDR,B (www.icddrb.org) in Bangladesh looked at viral

isolates from 272 HIV-1 positive patients, of which subtyping revealed 41% to be

the most common type – subtype C.

Analysis suggests that these subtype C strains were introduced to Bangladesh

from different parts of the world. Although most of the subtype C strains

obtained from injecting drug users were closely related, the strains obtain from

other patients were heterogeneous.

These data suggest that injecting drug users were not frequently transmitting

HIV to other at-risk populations. As Bangladesh is at the early phase of the HIV

epidemic, intervention programmes for those most-at-risk of infection should be

further strengthened to prevent a major epidemic.

National HIV surveillance in Bangladesh indicates that the HIV prevalence among

most-at-risk-population groups (sex workers, injecting drug users, males having

sex with males) is less than 1%.

However, the data regarding diversity of HIV strains and patterns of

transmission in Bangladesh are limited. Subtype data are important because the

major challenge in controlling AIDS lies in the diversity of HIV and its

enormous evolutionary potential. Gene sequence data can also be used to identify

the geographical origin of local HIV strains and their relationships with each

other and with global HIV strains.

ICDDR,B has been working to characterize Bangladeshi HIV strains by sequencing

their gag genes. During 1999–2005, blood samples positive for HIV were

collected through three different sources:

HIV surveillance which obtains samples from injecting drug users, female sex

workers, transgenders, men who have sex with men, heroin smokers, and patients

with sexually transmitted infections from different parts of the country.

Surveys which obtains samples from TB patients attending two TB clinics or

hospitalized in a TB hospital in Dhaka.

Samples collected from clients attending the voluntary counselling and testing

unit in ICDDR,B, Dhaka.

Genetic analysis revealed revealed a high similarity among 84% of the 31

Bangladeshi samples obtained from injecting drug users, which were similar to

strains described in the GenBank database from female sex workers in India.

The remaining 5 isolates obtained from injecting drug users were better matched

with Ethiopian and South African strains. The strains among patients seen in the

voluntary counselling and testing unit were diverse, and were similar to HIV

strains circulating in India, Zimbabwe and Ethiopia.

Five strains from female sex workers were closely related to each other with

more than 95% identity but not to the other 2 strains from female sex workers.

Among 7 strains isolated from female sex workers only 1 was similar to a strain

from an injecting drug user. Two of the heroin smokers had identical strains

(100%), which best matched strains from injecting drug users (99%). The strain

from 1 patient with a sexually transmitted infection was 100% identical to a

strain from female sex worker.

A phylogenetic tree was constructed which included the Bangladeshi strains with

reference HIV strains from different geographical regions available in the

GenBank database. Most of the isolates from injecting drug users (n=26)

clustered very closely together with Indian strains with the exception of 4

strains clustering with Ethiopian and South African strains.

Isolates from other most-at-risk populations did not show the same degree of

homogeneity. For example, the isolates from sex workers were mapped to different

branches with Chinese, Zimbabwean and South African isolates.

Transgender isolates clustered with either Chinese or Indian strains. The single

isolate from a Bangladeshi TB patient clustered most closely to an isolate from

the United States. Strains from the voluntary counselling and testing unit were

scattered throughout the phylogenetic tree and shared common origin with strains

from different countries like India, Myanmar, Zimbabwe and Ethiopia.

The clustering pattern of the HIV strains suggests that there was very little

overlap in the strains obtained from injecting drug users and those from other

Bangladeshi most-at-risk populations.

Phylogenetic analysis also indicated that some isolates from families clustered

suggesting that clients transferred the viruses to their wives and children.

Of the 70 clients from the voluntary counselling and testing unit, 77% reported

that they worked abroad. The other 16 were presumed infected from individuals

who had worked abroad (12 were wives and four were children of HIV-positive

migrants).

The majority of the migrants had travelled to Saudi Arabia and UAE while others

worked in India, Nepal, Malaysia and Singapore. All migrant workers reported

buying sex from women while abroad.

Genetic analysis confirms that the HIV-1 strains from some of these migrant

workers were most similar to HIV-1 strains from the countries where they worked.

These findings suggest that the HIV-1 infections in migrant workers were

acquired while working abroad.

For further information on this study, or on other HIV research at ICDDR,B,

contact

Dr Tasnim Azim,

Head, HIV/AIDS Programme and Virology Laboratory

ICDDR,B

tasnim@...

For further information on other public health research at ICDDR,B, contact

Jo Grzelinska

Communications

ICDDR,B

jo@...