Surveillance of transmitted HIV drug resistance among women attending antenatal clinics in Dar es Salaam, Tanzania

Antiviral Therapy 13 Suppl 2:77 82 Surveillance of transmitted HIV drug resistance among women attending antenatal clinics in Dar es Salaam, Tanzania Geofrey R Somi 1, Tabitha Kibuka 2, Karidja Diallo 3, Tulli Tuhuma 1, Diane E Bennett 4, Chunfu Yang 3, Charles Kagoma 5, Eligius F Lyamuya 6, Roland O Swai 1 and Sidibe Kassim 3 * 1 National AIDS Control Program, Dar es Salaam, Tanzania 2 Centers for Disease Control and Prevention/Tanzania, Dar es Salaam, Tanzania 3 Centers for Disease Control and Prevention Atlanta, GA, USA 4 World Health Organization, Geneva, Switzerland 5 World Health Organization, Dar es Salaam, Tanzania 6 Muhimbili University College of Health Sciences, Dar es Salaam, Tanzania *Corresponding author: E-mail: skassim@cdc.gov Background: In resource-limited settings where antiretroviral treatment (ART) access is being scaled-up, the World Health Organization (WHO) recommends surveillance of transmitted HIV drug resistance (HIVDR). We used the WHO HIVDR threshold survey method to assess transmitted HIVDR in Dar es Salaam where ART was introduced in 1995 and where ~11,000 people are currently on ART. Methods: From November 2005 to February 2006, dried blood spot (DBS) specimens were made from remnant specimens collected during the national HIV serosurvey from 60 primagravidas <25 years old attending six antenatal clinics for routine syphilis testing. Genotyping was performed at the Centers for Disease Control and Prevention, Atlanta, Georgia, USA. Protease and reverse transcriptase drug resistance mutations were identified using the Stanford University HIV drug resistance database. We used the National Institutes of Health genotyping tool for HIV-1 subtyping. HIVDR prevalence categorization was based on the WHO threshold survey binomial sequential sampling method. Results: Among the 60 eligible specimens collected, 50 DBS were successfully amplified using RT-PCR. Sequencing was performed on the first 39 specimens: 13 (33.3%) were subtype A1, 13 (33.3%) subtype C, and 4 (10.3%) subtype D, the remainder differed in the closest subtype based on protease versus reverse transcriptase. No resistance mutations were seen; HIVDR to all drug classes was categorized as <5%. Conclusions: Our survey indicates that prevalence of transmitted HIVDR among recently infected pregnant women in Dar es Salaam is low (<5%). The survey should be repeated during the next HIV sentinel survey in Dar es Salaam and extended to other regions where ART is being scaled up. Introduction Since the beginning of the 21st century, access to antiretroviral treatment (ART) has dramatically increased in resource-limited settings where the majority of HIV-infected individuals who need this life-saving treatment live [1]. The increase in the number of individuals receiving ART is the result of several initiatives including the United States President s Emergency Plan for AIDS Relief (PEPFAR) [2] and the Global Fund to Fight AIDS, Tuberculosis and Malaria [3]. To facilitate the implementation of ART programmes in resource-limited settings, public health guidelines have been developed to ensure favourable treatment outcomes and to minimize the emergence of HIV drug resistance (HIVDR) [4]. Key components of these guidelines include the standardization and simplification of antiretroviral (ARV) regimens and the avoidance of substandard treatment. The adoption of these guidelines in scaling-up ART programmes in countries in Africa and other parts of the world has led to treatment outcomes comparable to those attained in North America and Western Europe [5]. Despite these achievements, the widespread use of ARV drugs will lead to the transmission of HIV variants resistant to ARV drugs, as has been reported in the United States and Europe [6 8]. To inform ART programmes in resource-limited settings where screening for HIVDR is not routinely available, the 2008 International Medical Press 1359-6535 77

GR Somi et al. WHO recommends surveillance of transmitted drugresistant HIV in recently infected persons using a minimum-resource strategy to signal when the level of HIVDR requires action. The recommended method (described in detail in this supplement) is the HIVDR threshold survey [9], which classifies HIVDR prevalence as low (<5%), medium (5 15%) or high (>15%). The maximum sample size is 47 consecutive specimens, which is sufficient to classify HIVDR prevalence in one of the three categories. However, sampling can be stopped before reaching the maximum sample size if HIVDR prevalence can be classified as low or high with the data available using the HIVDR threshold survey sampling and classification table [9]. The United Republic of Tanzania, composed of mainland Tanzania and the islands of Zanzibar, has an estimated total population of 33,584,607 inhabitants. In 2005, HIV prevalence in mainland Tanzania was estimated at 6.5% (5.8 7.2%) in the adult population (15 49 years old), and the number of adults and children living with HIV/AIDS in the country was estimated at 1,400,000 [10]. Although not documented, ART started in Tanzania with the introduction of zidovudine monotherapy in 1995. This was shortly followed by combination therapies including zidovudine and didanosine, indinavir and didanosine, and stavudine and indinavir. Until the current government programme started in 2003, access to ART was very limited. The government programme is based on the WHO guidelines for use of ART in resource-limited settings [4]. The first-line regimen of choice in adult treatment is stavudine, lamuvidine and nevirapine in fixed-dose combination. Three alternative first-line regimens have also been used: zidovudine, lamuvidine and nevirapine; stavudine, lamuvidine and efavirenz; or zidovudine, lamuvidine and efavirenz. The second-line regimens include abacavir, lopinavir/ritonavir and didanosine or abacavir, saquinavir/ritonavir and didanosine. In 2005, >29,000 persons were receiving ART in the country, with ~11,000 in Dar es Salaam alone. All patients in the national ART programme are managed following WHO guidelines for monitoring persons on combination ART, which is limited to clinical monitoring and CD4 + T-cell count testing. Viral load monitoring and HIVDR testing are not routinely available. HIV sentinel serosurveys among pregnant women [11] have been an important source of data on the epidemiology of HIV in the country since the beginning of the epidemic. The HIV sentinel surveillance among pregnant women in Tanzania is anonymous and unlinked. The survey is conducted in selected antenatal care (ANC) clinics in the different regions of the country. All pregnant women aged 15 49 years who present at the ANC for their first antenatal care visit are offered a blood draw for syphilis test; remnant serum is made anonymous and used for the HIV serosurvey. Methods of the HIV serosurvey have been described previously [12]. HIV prevalence was ~11% among pregnant women attending ANCs in Dar es Salam; among women aged 15 to 24 years, HIV prevalence was 7.6% (Tanzanian National AIDS Control Program, 2006 report). Because no HIV case surveillance is conducted in Tanzania, and because the majority of HIV cases diagnosed in voluntary counselling and testing centres are advanced cases, pregnant women attending ANC who are newly tested for HIV represent the best population to identify individuals relatively recently infected by HIV. We conducted HIVDR surveillance among pregnant women in ANCs in Dar es Salaam during the 2005/2006 HIV and syphilis infection sentinel serosurveillance. Methods Following the HIVDR threshold survey recommendations [9], eligible women for this survey were HIV-positive primigravida women <25 years old. From November 2005 to February 2006, during the HIV sentinel surveillance among pregnant women attending ANC, two anonymous unlinked dried blood spot (DBS) specimens were prepared (one each for HIV sentinel surveillance and transmitted HIVDR surveillance) using remnant specimens collected for routine syphilis testing in six ANCs in Dar es Salaam. DBS specimens for HIVDR surveillance and HIV sentinel surveillance were prepared at the same time. Following standard operating procedures, DBS specimens were dried overnight at each survey site and shipped to a central laboratory the day following specimen collection for storage. Overall, 3,563 DBS specimens were collected during the HIV sentinel survey. At the central laboratory, based on the pregnancy and age criteria for HIVDR surveillance, 1,172 (32.9%) specimens were selected and sent by courier to the Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA, for genotyping. After completion of the HIV test in the central laboratory in Dar es Salaam, 60 of the 1,172 specimens sent to Atlanta were found to be positive and thus eligible for genotyping. For HIVDR testing, total nucleic acids were extracted from DBS using a modified NucliSens silicabased extraction method [13]. A segment of 1,023 bp containing the protease and reverse transcriptase (RT) regions of the HIV-1 genome was amplified with an inhouse RT-PCR method and sequence determination was performed with an ABI 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Drug resistance mutations were identified using Stanford 78 2008 International Medical Press

HIV drug resistance surveillance in Tanzania University s HIV drug resistance database [14]. Relevant major mutations were determined based on the WHO surveillance list for transmitted drug resistance [15], and prevalence of HIVDR was categorized using the binomial sequential sampling and classification plan described in detail in this supplement [16]. HIV-1 subtyping was determined based on the protease and RT regions of the HIV-1 pol gene, using the genotyping tool provided by the National Institutes of Health [17]. When the tool reported that the two regions most closely resembled different subtypes, we assigned subtype based on the protease region first, followed by the subtype based on the RT region. The protocols for HIV sentinel surveillance and the surveillance of transmitted HIVDR were approved by appropriate Institutional Review Boards in Tanzania and at CDC. Results Of the 60 DBS specimens eligible for genotyping, amplification of 50 (83.3%) samples proved successful using RT-PCR and sequence analysis was obtained for the first 39 specimens. No more specimens were sequenced because HIVDR prevalence classification could be made based on this number of sequences. According to the NIH genotyping tool, the HIV-1 subtypes of the 39 specimens sequenced were determined: 13 (33.3%) subtype A1, 13 (33.3%) subtype C, 4 (10.3%) subtype D, 2 (5.1%) subtype C/A1, 2 (5.1%) subtype CRF08_BC/C, 2 (5.1%) subtype D/CRF10_CD, 1 (2.6%) subtype A1/D, 1 (2.6%) subtype CRF15_01B/A1 and 1 (2.6%) subtype C/D. No mutations included in the WHO list for surveillance of transmitted HIVDR [15] were seen in the first 39 specimens genotyped. Using the binomial sequential sampling and classification table (Table 1) [16], the results from the first 34 specimens were sufficient to classify the prevalence of HIVDR for all relevant drugs and drug classes as <5% in this population. On the basis of the Stanford University HIV drug resistance database [14], some polymorphic mutations of interest were identified. Most were protease polymorphisms, while very few presented RT (non-nucleoside reverse transcriptase and nucleoside reverse transcriptase) mutations (Table 2). Discussion In this survey, we did not identify any mutations associated with transmitted resistance according to the WHO list for surveillance of transmitted HIVDR. Among the polymorphic mutations identified, only M46L is associated with low-level resistance to some protease inhibitors [18]. These findings suggest that prevalence of transmitted HIVDR in the city of Dar es Salaam is still low, and current ARV drugs used in first- and second-line regimens should continue to be effective if good adherence is achieved. Our findings are supported by an analysis by Vardavas et al. [19], which found that in countries like Tanzania, where ART is being expanded, prevalence of transmitted HIVDR will be low for the next few years because of the low level of ART coverage. In Tanzania, ART coverage was estimated at ~7% in 2005 [10]; at that level, the number of new infections attributable to persons exposed to ART is almost insignificant. The national ART programme has an ambitious target to reach about 400,000 people in more than 450 treatment sites by 2008. When these numbers are achieved, the risk of transmitted HIVDR is likely to grow substantially. Combining HIVDR surveillance with the HIV sentinel surveillance significantly reduced the cost of data collection for HIVDR surveillance. The use of the same type of specimen for both surveys was an additional advantage, particularly DBS specimens that require minimum laboratory infrastructure for specimen processing and transport. Use of DBS specimens for genotyping can still be challenging because of the difficulties of ensuring that adequate genome material is preserved in the sample. In this survey only 83.3% (50/60) of eligible specimens were amplifiable using PCR, despite the initial training provided to healthcare workers responsible for DBS specimen collection and the ongoing supervision throughout the survey. This might limit the ability to generalize the results, although bias is unlikely to be related to any factors associated with non-optimal specimen handling. Optimum conditions and procedures, such as the minimum quantity of blood needed to prepare DBS specimens and storage and transport requirements should be determined to ensure a higher quality of specimens for HIVDR testing. It is also important to evaluate other specimen collection, storage and transport possibilities other than frozen plasma specimens, which are impractical in many resource-limited settings, to improve the yield for HIVDR testing. The extension of the survey to other regions of the country using specimens collected during HIV sentinel surveillance will be difficult because of the minimum sample size (34 specimens) required to classify the HIVDR prevalence as low (which is expected in most areas). During this survey, only ~1.7% of the 3,563 specimens collected in the six survey sites were eligible for genotyping. Assuming that HIV prevalence and the proportion of primigravida women <25 years old in survey sites in Dar es Salaam are equivalent in other cities, one would need to collect at least Antiviral Therapy 13 Suppl 2 79

GR Somi et al. Table 1. Classification of transmitted HIVDR prevalence among women attending ANCs in Dar es Salaam, Tanzania 2005 Specimens Lower Running total of specimens with HIVDR mutations Upper genotyped, n limit (LL) (ResTot) A/B/C/D limit (UL) Dar1 ND 0/0/0/0 ND Dar2 ND 0/0/0/0 ND Dar3 ND 0/0/0/0 ND Dar4 ND 0/0/0/0 ND Dar5 ND 0/0/0/0 ND Dar6 ND 0/0/0/0 ND Dar7 ND 0/0/0/0 ND Dar8 ND 0/0/0/0 ND Dar9 ND 0/0/0/0 ND Dar10 ND 0/0/0/0 ND Dar11 ND 0/0/0/0 ND Dar12 ND 0/0/0/0 ND Dar13 ND 0/0/0/0 ND Dar14 ND 0/0/0/0 5 Dar15 ND 0/0/0/0 5 Dar16 ND 0/0/0/0 5 Dar17 ND 0/0/0/0 5 Dar18 ND 0/0/0/0 5 Dar19 ND 0/0/0/0 5 Dar20 ND 0/0/0/0 5 Dar21 ND 0/0/0/0 5 Dar22 ND 0/0/0/0 5 Dar23 ND 0/0/0/0 5 Dar24 ND 0/0/0/0 5 Dar25 ND 0/0/0/0 6 Dar26 ND 0/0/0/0 6 Dar27 ND 0/0/0/0 6 Dar28 ND 0/0/0/0 6 Dar29 ND 0/0/0/0 6 Dar30 ND 0/0/0/0 6 Dar31 ND 0/0/0/0 6 Dar32 ND 0/0/0/0 6 Dar33 ND 0/0/0/0 6 Dar34 1 0/0/0/0 6 2 7 2 7 2 8 2 8 Stop Stop Stop Stop Genotyping of specimens stops when a classification of HIV drug resistance (HIVDR) prevalence has been made on the basis of this chart. A classification of HIVDR prevalence is made when, among the specimen number genotyped listed in the first column, the running total of specimens found with transmitted HIVDR (ResTot in the third column) is less than the lower limit (LL) specified in the second column or greater than the upper limit (UL) specified in the fourth column. When one of these conditions occurs, HIVDR prevalence can be classified either as <5% (if ResTot is less than the LL on the same horizontal line) or >15% (if ResTot is greater than the UL). If neither of these conditions has occurred after the 47th specimen has been genotyped, prevalence is classified as 5% or 15%, respectively. A separate classification must be performed for each drug or drug class of interest. The table summarizes the separate analyses for resistance to at least one drug in any class (A), nucleoside reverse transcriptase inhibitor (NRTI) resistance (B), nucleoside reverse transcriptase inhibitor (NNRTI) resistance (C), and protease inhibitor (PI) resistance (D). Categorizations: resistance to at least one drug in any class <5%; resistance to at least one NRTI <5%; resistance to at least one NNRTI <5%; resistance to at least one PI <5%. ND, no decision. 80 2008 International Medical Press

HIV drug resistance surveillance in Tanzania 2,000 specimens in a given site in order to have 34 eligible for genotyping. None of the survey sites outside Dar es Salaam currently sees a sufficient number of eligible young primagravida pregnant women to collect this number of specimens. In these regions, it will be necessary to consider other strategies, such as the extension of the data collection period during HIV sentinel surveillance until the desired sample size is reached. With a low prevalence of transmitted HIVDR, it is recommended to conduct the survey periodically in the same geographical areas and initiate progressive extension to other regions of the country where the ART programme is being implemented. In this way hot spots can be identified and corrective measures taken early to prevent occurrence of HIVDR. It is also recommended to assess factors that might be associated with emergence of HIVDR in the ART programme, as well as assessing HIVDR emerging during ART [20]. Acknowledgements International Medical Press is acknowledged for their support with publishing this manuscript. The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of the CDC. Some of the authors are staff members of the WHO. The authors alone are responsible for the views expressed in this publication and they do not necessarily represent the decisions or stated policy of the WHO. Table 2. List of polymorphic mutations identified at resistance-associated pol gene positions among women attending ANCs in Dar es Salaam, Tanzania 2005 Mutation PI mutations M46L/V 2 F53Y 1 L33F 1 K20I/Q/R/T 6 M36I 35 L63A/I/K/P/Q 27 A71T 2 I93L 17 NRTI mutations V75L 1 Y115I 1 V118I 2 NNRTI mutations V108IL 1 Number of specimens NRTI, nucleoside reverse transcriptase inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor. Disclosure statement The authors declare that they have no competing interests. References 1. UNAIDS. AIDS epidemic update: December 2006. (Accessed 19 March 2007). Available at http://www.who.int/hiv/ mediacentre/2006_epiupdate_en.pdf 2. Office of the United States Global AIDS Coordinator. The President Emergency Plan for AIDS Relief (PEPFAR). US five-year global AIDS strategy; 2004. (Accessed 19 March 2007). Available at http://www.state.gov/ documents/ organization/29831.pdf 3. The Global Funds to Fight AIDS, Tuberculosis and Malaria. (Accessed 19 March 2007). Available at http://www.theglobalfund.org/en/ 4. WHO. Scaling up antiretroviral therapy in resourcelimited settings: treatment guidelines for a public health approach; 2003. (Accessed 19 March 2007). Available at http://www.who.int/3by5/publications/documents/ arv_guidelines/en/ 5. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1 infected patients in the first year of antiretroviral therapy: comparison between low-income and highincome countries. Lancet 2006; 367:817 824. 6. Clavel F, Hance AJ. HIV drug resistance. N Engl J Med 2004; 350:1023 1035. 7. Bangsberg DR, Charlebois ED, Grant RM, et al. High levels of adherence do not prevent accumulation of HIV drug resistance mutations. AIDS 2003; 17:1925 1932. 8. Little SJ, Holte S, Routy JP, et al. Antiretroviral drug resistance among patients recently infected with HIV. N Engl J Med 2002; 347:386 394. 9. Bennett DE, Myatt M, Bertagnolio S, Sutherland D, Gilks CF. Recommendations for surveillance of transmitted HIV drug resistance in countries scaling up antiretroviral treatment. Antivir Ther 2008; 13 Suppl 2:25 36. 10. UNAIDS. Report on the global AIDS epidemic; 2006. (Accessed 19 March 2007). Available at http://www.unaids.org/en/hiv_data/2006globalreport/ default.asp 11. UNAIDS.UNAIDS/WHO working group on global HIV/AIDS and STI surveillance. Guidelines for conducting HIV sentinel serosurveys among pregnant women and other groups; 2003. (Accessed 19 March 2007). Available at http://www.who.int/hiv/pub/ surveillance/anc_guidelines/ en/index.html 12. Roland SO, Somi GR, Matee MI, et al. Surveillance of HIV and syphilis infections among antenatal clinic attendees in Tanzania 2003/2004. (Accessed 19 March 2007). Available at http://www.biomedcentral.com/ 1471 2458/6/91\ 13. McNulty A, Jennings C, Bennett D, et al. Evaluation of dried blood spots for human immunodeficiency virus type 1 drug resistance testing. J Clin Microbiol 2007; 45:517 521. 14. Stanford University, HIV drug resistance database. (Accessed 19 March 2007). Available at http://hivdb.stanford.edu/pages/asi/ 15. Shafer RW, Rhee SY, Pillay D et al. HIV-1 protease and reverse transcriptase mutations for drug resistance surveillance. AIDS 2007; 21:215 223. 16. Myatt M, Bennett DE. A novel sequential sampling technique for the surveillance of transmitted HIV drug resistance by cross-sectional survey for use in low resource settings. Antivir Ther 2008; 13 Suppl 2:37 48. 17. Natrional Institutes of Health, HIV genotyping tool. (Accessed 18 July 2007). Available at http://www.ncbi.nlm.nih.gov/projects/genotyping/ formpage.cgi Antiviral Therapy 13 Suppl 2 81

GR Somi et al. 18. Baxter JD, Schapiro JM, Boucher CA, et al. Genotypic changes in human immunodeficiency virus type 1 protease associated with reduced susceptibility and virologic response to the protease inhibitor tipranavir. J Virol 2006; 80:10794 10801. 19. Vardavas R, Blower S. The emergence of HIV transmitted resistance in Botswana: When will the WHO detection threshold be exceeded? PLoS ONE 2007; 17:e152. 20. Jordan MR, Bennett DE, Bertagnolio S, Gilks CF, Sutherland D. World Health Organization surveys to monitor HIV drug resistance prevention and associated factors in sentinel antiretroviral treatment sites. Antivir Ther 2008; 13 Suppl 2:15 23. Accepted for publication 27 January 2008 82 2008 International Medical Press