Linking HIV and Antiretroviral Drug Resistance Surveillance in Peru: A Model for a Third-Generation HIV Sentinel Surveillance : JAIDS Journal of Acquired Immune Deficiency Syndromes

Secondary Logo

Journal Logo

Epidemiology and Social Science

Linking HIV and Antiretroviral Drug Resistance Surveillance in Peru

A Model for a Third-Generation HIV Sentinel Surveillance

Lama, Javier R. MD, MPH*; Sanchez, Jorge MD, MPH*; Suarez, Luis MD; Caballero, Patricia MD; Laguna, Alberto MD, MTMH, PhD§; Sanchez, Jose L. MD, MPH§∥; Whittington, William L.H. AB; Celum, Connie MD, MPH; Grant, Robert M. MD, MPH# Peruvian HIV Sentinel Surveillance Working Group

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes 42(4):p 501-505, August 1, 2006. | DOI: 10.1097/01.qai.0000221677.29693.dd
  • Free


Antiretroviral (ARV) drug resistance surveillance is limited by small and unrepresentative samples of recently HIV-1-infected persons, biased recruitment of persons with symptomatic primary HIV infection, and lack of comparable information regarding HIV incidence rates, secondary resistance, and treatment prevalence.1 Given the expansion of ARV programs worldwide, it is expected that resistance to HIV will continue to emerge and spread and that appropriate surveillance programs are needed to be implemented, both in countries where highly active antiretroviral therapy (HAART) is widely accessible and in areas where it is being introduced.2 In places where HAART programs are rolling out, linking HIV drug resistance surveillance to HIV infection surveillance is desirable to allow important and interacting parameters of the epidemic to be measured in the same sampling frame, including infection prevalence, infection incidence, treatment prevalence, and drug resistance.3 This approach also avoids the cost of setting up new surveillance systems, which is particularly important for resource-poor settings.

Until the Peruvian Ministry of Health ARV program expanded public HAART access in 2004, a minority of HIV-infected persons had access to treatment. Partially effective and nonsuppressive [eg, mono or dual nucleoside reverse transcriptase inhibitors (NRTIs)] therapies were commonly used, influenced by costs, accessibility, and short-term clinical outcomes.4 Use of less effective ARV regimens raises concerns about selection and transmission of HIV-1-resistant strains, which could impair virologic responses to HAART regimens that are now increasingly used in these settings. Alternatively, minimal ARV exposure before the advent of potent combinations in Peru could help prevent the extensively mutated multidrug-resistant HIV variants that have been reported in the United States and Europe.5-7

Peru, where men who have sex with men (MSM) account for the largest proportion of HIV-infected cases,8,9 provides a unique opportunity for research into the epidemiology of drug-resistant HIV in the context that HAART is recently becoming more widely available through the Global Fund to Fight AIDS, Tuberculosis, and Malaria and the Peruvian Ministry of Health co-sponsored program. Baseline assessments will characterize a population that is relatively treatment-naive at the time of scaling up ARV access. To gain further insight into the prevalence and patterns of ARV drug resistance in Peru in 2002, we linked ARV drug resistance surveillance to the regular second-generation HIV sentinel surveillance, which had been biannually conducted among MSM in Peru since 1996.


Men who were at least 18 years of age and who had had sexual intercourse with one or more men during the previous year were eligible to participate, regardless of history of HIV screening, serostatus, or treatment. There was no effort to oversample HIV-infected persons.

During a 3-month period, recruiters and peer educators visited previously mapped venues in which MSM socialized to contact potential participants and to refer them to sentinel clinic sites. Surveillance activities were conducted in 6 different cities: Lima, Sullana, Piura, Arequipa, Iquitos, and Pucallpa. Specimens were obtained between October 2002 and March 2003. Cities in prior surveillance activities were chosen to include the most highly populated provinces in the country. Cities in the 2002 surveillance were selected based on their having the highest HIV prevalences and incidences of HIV infection in the 2000 HIV sentinel surveillance. Sample sizes were selected to allow prevalence in each city to be estimated with a defined 95% confidence interval (CI).

Comparable sampling strategies were used in all cities. Recruiters and peer educators representing the diverse MSM subcultures mapped venues where MSM socialize, including saunas, movie theaters, video arcades, places of commercial sex work, discotheques, bars, beauty parlors, houses for transvestites, and athletic courts. Once mapped, all of these centers were visited using standard procedures to contact potential participants and to refer them to the study clinics. Recruiters used diverse methods to refer participants, including posters, flyers, and informational meetings.

Those men who agreed and provided written informed consent underwent a structured interview for sexual risk behavior, and physicians obtained a medical history and conducted a targeted physical examination. This study protocol was approved by the Asociacion Civil Impacta Salud y Educacion, University of Washington, and Naval Medical Research Center Institutional Review Boards. HIV-infected participants and those with a diagnosis of a sexually transmitted infection received counseling, education, and treatment. HIV-infected participants were referred for standard care, which did not include publicly funded ARV therapy in 2002.

A blood sample was obtained for initial HIV-1 antibody screening by enzyme immunoassay (Vironostika, Organon Teknika, Durham, NC), with confirmation by Western blot (Biorad Laboratories, Hercules, CA). Recent HIV-1 infection was detected among those HIV positive who had an optical density signal-to-cutoff ratio less than 0.75 in a sensitive/less sensitive enzyme immunoassay testing (Vironostika, Organon Teknika).10 Laboratory testing was blinded to prior HIV infection status.

Plasma was separated from whole blood samples in all participants diagnosed with HIV infection and then aliquoted and stored at −84°C. Further testing for mutations on the HIV-1 reverse transcriptase and protease gene sequences conferring drug resistance was conducted using ViroSeq HIV-1 Genotyping System (manufactured by Celera Diagnostics and distributed by Abbott Molecular Inc, Chicago, IL). Consensus sequences from all genotyped individuals were aligned and manually edited, and neighbor-joining phylogenetic trees were used to seek evidence of laboratory contamination. Genotypic resistance was defined as the presence of mutations associated with impaired drug susceptibility or virologic response as specified by the International AIDS Society-USA mutations panel.11 No further attempts were made to reamplify unsuccessfully genotyped sequences.

Observed point prevalences were computed and HIV incidences with 95% CIs were estimated.10


A total of 3280 MSM met entry criteria in the 6 Peruvian cities and consented to participate. Overall, HIV prevalence was 13.9% and incidence was estimated to be 5.1 per 100 person-years (95% CI: 3.1-8.3). Iquitos and Lima had the highest HIV prevalence and incidence rates (Table 1). Of the 456 HIV-infected participants, 143 (31.4%) reported that they were already aware of their HIV infection before testing. Only 20 (14.0%) of the 143 men who had prior knowledge of their HIV infection reported ever being on ARVs, of whom 12 reported ARV treatment during the preceding 6 months. No data on specific ARV regimens were obtained.

Observed HIV Prevalences and Estimated HIV Incidences Among Men Who Have Sex With Men in Peru, 2002

Sequencing was completed in plasma samples from 375 (82.2%) of the 456 HIV-infected participants. Antiretroviral drug resistance was detected in 12 (3.3%) of 359 treatment-naive and 5 (31.3%) of 16 treatment-experienced participants with successful genotyping (Table 2). All sequences derived were subtype B. There were no differences between genotyping successes and failures with respect to demographics, risk behavior, and proportion of treatment naive.

Prevalence of Antiretroviral Resistance Among Men Who Have Sex With Men in Peru, 2002

Among all treatment-naive HIV-infected men from Lima, the overall prevalence of resistance to any drug class was 3.9%. Resistance to one or more drugs within each drug class was 3.0% for NRTIs and 2.5% for protease inhibitors (PIs), with no resistance for nonnucleoside reverse transcriptase inhibitors (NNRTIs) detected among chronically infected participants. In cities other than Lima, resistance was only observed in 2 treatment-naive chronic HIV-1-infected participants from Sullana, both with K103N mutation present, which confers resistance to NNRTIs.

There was one treatment-naive man, also recently infected, from Lima with 3-class multidrug-resistant infection. The drug resistance genotype revealed evidence of the following mutations: M184V, K103N, and L90M, expected to confer resistance to lamivudine, all NNRTIs, and PIs, respectively. Although this multidrug resistance was only observed in a single participant, he was a sex worker who reported a large number of sexual partners (100) in the prior 3 months.

Among treatment-experienced MSM, the overall prevalence of resistance to NRTIs was 31.3%; to NNRTIs, 6.3%; and to PIs, 18.8%. All men with ARV resistance were from Lima (Table 2).


We have successfully incorporated HIV resistance surveillance into HIV sentinel surveillance in Peru and found low prevalences of HIV-1 drug resistance, likely reflecting the low treatment rates before 2004, and a high frequency of wild-type drug failure among treated persons. The high incidence of HIV-1 infection that is evident in Peru is also expected to favor the spread of more robust drug-susceptible viruses.12 Conversely, ARV resistance was prevalent among the small number of ARV-treated persons, most frequently affecting NRTIs that were commonly used in either mono or dual drug therapy before the 2002 sentinel surveillance. The low baseline prevalence of ARV resistance before public ARV treatment "rollout" in Peru contrasts with the history of ARV treatment in developed countries, where high levels of NRTI resistance occurred before introduction of HAART.5 Nonetheless, one case of 3-class primary drug resistance was observed, which may reflect migration from more heavily treated populations. Drug resistance was also observed in 2 participants from Sullana, a northern border city with Ecuador, suggesting that surveillance should be expanded to a regional level. Notably high HIV prevalence and incidence were observed in selected cities, confirming that MSM represent an important core group for surveillance and intervention in Peru.13 The high proportion of infected persons who were previously unaware of their infection status supports wider promotion of voluntary testing and counseling services. In very high incidence populations such as in Peru, testing using more sensitive methods that detect viral antigens or nucleic acids should be considered so that very recent infections may be detected.

Information regarding prevalence and trends of primary drug resistance in developing countries is limited. Even from locations having extensive access to testing technology, reports of primary drug resistance are difficult to interpret; because sample sizes are small, the prevalence and incidence of HIV infection and treatment are not determined such that the prevalence of ARV resistance among persons at risk for transmitting HIV is not known from the same sampling frame. This large population-based surveillance program in Peru offers a unique opportunity to address these important gaps in the knowledge regarding how rapid scale-up of combination treatment will impact the regional epidemiology of ARV-resistant HIV. The surveillance data reported here support the current HIV treatment policies in Peru, which recommend initiation of standard treatment regimens without drug resistance testing for individual patients. Use of standard regimens with minimal laboratory requirements allows for more persons to be treated. Drug resistance and treatment surveillance data also guide policy regarding adherence strategies (directly observed therapy vs. self-administered therapy and various hybrids), second-line regimen selection, and acquisition, and help set regional priorities for treatment rollout, laboratory development, and counseling.

This sample of MSM represents a convenience-based sample of the population of MSM in these 6 Peruvian cities, who were recruited by experienced outreach workers. This surveillance activity followed standard guidelines14,15 and recruited a large sample of geographically diverse participants from 6 different cities representing the diverse MSM subcultures in Peru. Nonetheless, the sample may be biased depending on the recruiters' knowledge of and access to different subgroups of MSM in Peru. We were not able to assess the association between high-risk sexual behavior and presence of ARV resistance because of the low frequency of such mutations. Assessment of genotypic drug resistance using ViroSeq has been reliable for analysis of HIV-1 subtype B and non-B subtype viruses.16-18 Analysis of a subset of the specimens from genotyping assay failures indicated low or undetectable plasma RNA levels, suggesting that viral RNA degradation in surveillance specimens rather than the presence of highly divergent viral variants accounted for most of the assay failures.

Our results indicate that implementation of programs for ARV resistance surveillance in developing settings is feasible. Our methodology has several advantages over previous studies on the epidemiology of drug resistance, including (1) the proposed research is in a well-defined area; (2) the samples are population-based, allowing estimates of population prevalence of drug resistance in treated and untreated persons; and (3) persons at risk for transmitting HIV are included in the well-characterized sample, allowing estimates to be weighted by individual sexual behaviors, and to extrapolate to potential transmission within sexual networks and to model at a population level. Furthermore, using specimens gathered for other surveillance purposes minimizes costs.

Second-generation HIV sentinel surveillance becomes a less useful approach for monitoring changes in the HIV epidemic after HAART programs are scaled up because therapy-induced increases in survival will increase HIV prevalence regardless of changes in the incidence of HIV infection. Thus, an innovative approach is needed to capture the dynamic nature of the epidemic in the HAART era. We propose that an integrated approach of community-based surveillance of sexual behavior in high-risk populations should be linked to monitoring HIV incidence, treatment rates, primary and secondary ARV resistance, and phylogenetic diversity in third-generation HIV sentinel surveillance model programs. This information is needed to guide both prevention and treatment programs.


The authors thank Mrs Maribel Acuña and Zoila Angeles for technical assistance; Abbott Molecular Inc for provision of kits; Drs Martin Casapia (Iquitos), Nora Ojeda (Sullana and Piura), Abner Ortiz (Pucallpa), Jorge Vergara (Lima), and Victoria Zamalloa (Arequipa) for local study leadership; Mrs Rosa Galvan for laboratory study coordination; all recruiters and peer educators; and mainly, all study participants.


1. Tang JW, Pillay D. Transmission of HIV-1 drug resistance. J Clin Virol. 2004;30:1-10.
2. Vella S, Palmisano L. The global status of resistance to antiretroviral drugs. Clin Infect Dis. 2005;41(suppl 4):S239-S246.
3. Blower SM, Gershengorn HB, Grant RM. A tale of two futures: HIV and antiretroviral therapy in San Francisco. Science. January 28, 2000;287(5453):650-654.
4. Verdonck K, Vidal M, Seas C, et al. Use of antiretroviral therapy among HIV-seropositive patients attended at a public reference hospital in Peru. Paper presented at: XV International AIDS Conference; 2004; Bangkok, Thailand.
5. Grant RM, Hecht FM, Warmerdam M, et al. Time trends in primary HIV-1 drug resistance among recently infected persons. JAMA. July 10, 2002;288(2):181-188.
6. Little SJ, Holte S, Routy JP, et al. Antiretroviral-drug resistance among patients recently infected with HIV. N Engl J Med. August 8, 2002;347(6):385-394.
7. Wensing AM, van de Vijver DA, Angarano G, et al. Prevalence of drug-resistant HIV-1 variants in untreated individuals in Europe: implications for clinical management. J Infect Dis. September 15, 2005;192(6):958-966.
8. UNAIDS. UNAIDS/Pan American Health Organization. AIDS surveillance in the Americas. Annual Report; 2004.
9. Bautista CT, Sanchez JL, Montano SM, et al. Seroprevalence of and risk factors for HIV-1 infection among South American men who have sex with men. Sex Transm Infect. 2004;80:498-504.
10. Janssen RS, Satten GA, Stramer SL, et al. New testing strategy to detect early HIV-1 infection for use in incidence estimates and for clinical and prevention purposes. JAMA. 1998;280:42-48.
11. Hirsch MS, Brun-Vezinet F, Clotet B, et al. Antiretroviral drug resistance testing in adults infected with human immunodeficiency virus type 1: 2003 recommendations of an International AIDS Society-USA Panel. Clin Infect Dis. July 1, 2003;37(1):113-128.
12. Sanchez MS, Grant RM, Porco TC, et al. A decrease in drug resistance levels of the HIV epidemic can be bad news. Bull Math Biol. July 2005;67(4):761-782.
13. Bautista CT, Sanchez JL, Montano SM, et al. Seroprevalence of and risk factors for HIV-1 infection among South American men who have sex with men. Sex Transm Infect. December 2004;80(6):498-504.
14. UNAIDS/WHO Working Group on Global HIV/AIDS/STI Surveillance. Guidelines for effective use of data from HIV surveillance systems. UNAIDS, Geneva; 2004:60.
15. Rehle T, Lazzari S, Dallabetta G, et al. Second-generation HIV surveillance: better data for decision-making. Bull World Health Organ. 2004;82:121-127.
16. Mracna M, Becker-Pergola G, Dileanis J, et al. Performance of Applied Biosystems' ViroSeq HIV-1 Genotyping System for sequence-based analysis of non-subtype B human immunodeficiency virus type 1 from Uganda. J Clin Microbiol. 2001;39:4323-4327.
17. Performance of the Celera Diagnostics ViroSeq HIV-1 Genotyping System for sequence-based analysis of diverse human immunodeficiency virus type 1 strains [computer program]. Version; 2004.
18. Beddows S, Galpin S, Kazmi SH, et al. Performance of two commercially available sequence-based HIV-1 genotyping systems for the detection of drug resistance against HIV type 1 group M subtypes. J Med Virol. 2003;70:337-342.

epidemiology; antiviral drug resistance; sentinel surveillance; HIV-1; developing countries; homosexual men

© 2006 Lippincott Williams & Wilkins, Inc.