Pretreatment HIV-1 Resistance in Argentina: Results from the Second Surveillance Study Following World Health Organization Guidelines (2019)

Introduction

Over the last decades, and especially in the last one, there has been an extraordinary rise in the use of antiretroviral therapy (ART) worldwide. By the end of 2021, 28.7 million people were on ART globally, out of an estimated 38.4 million people living with HIV-1. ¹

The access to ART has saved the lives of millions of people with HIV/AIDS; nevertheless, its use is associated with the possibility of developing drug resistance. The emergence and transmission of HIV drug resistance in the context of universal access to antiretroviral treatment constitute a serious problem, especially in regions where ART regimens include drugs with a lower genetic barrier for the development of resistance.

The emergence of resistance not only represents a problem at individual levels but also at population level, since it could increase the risk of HIV transmission due to the lack of viral suppression, resulting in increased numbers of HIV infections and HIV-associated morbidity and mortality. Since 2004, the World Health Organization (WHO) and partners in the Global HIV Drug Resistance Network (HIVResNet) have been developing a public health strategy to prevent and assess HIV drug resistance in resource-limited countries in the context of scaling up antiretroviral treatment. ²

The implementation of studies aimed at the surveillance of HIV-1 resistance allows to generate strategic information that can be used to take actions to control resistance. These actions include the recommendation of first- and second-line schemes that present improved long-term effectiveness, the intensification of viral load monitoring, the definition of the strategic use of HIV-1 genotyping, and the use of pre- and post-exposure prophylaxis, among others. ³ In 2017, the WHO launched the Global action plan on HIV drug resistance 2017–2021, describing a package of interventions and resources to guide the response to HIV drug resistance (HIVDR), including the report on HIV-1 drug resistance, as well as HIVDR guidelines and implementation tools, focused on dolutegravir-based antiretroviral regimens, monitoring HIV care service delivery, and strategies to ensure uninterrupted drug supplies. ⁴

To date, levels of HIVDR in countries scaling up ART remain manageable; however, they are slowly increasing. For example, in 21 of 30 surveys reported to WHO, pretreatment HIV-1 drug resistance to nevirapine (NVP) or efavirenz (EFV) in populations initiating first-line ART reached levels above 10%. ⁵ In Argentina, more than 65,000 individuals are currently on ART in the public health system. ⁶

The first national survey of HIV-1 resistance in Argentina was done in 2014, ⁷ which included 330 individuals showed overall data of primary resistance to non-nucleoside reverse transcriptase inhibitors (NNRTIs) of 11% (17% in people previously exposed to ARVs), to nucleoside/nucleotide reverse transcriptase inhibitors (NRTI) of 3%, and 2% to protease inhibitors (PI). In 2019, with more than 50% of people on ART treated with NNRTIs, the ART guidelines had switched from NNRTI as a first-line regimen to dolutegravir-based regimens in 2019. ⁸ It was necessary to update the data regarding this very important factor for HIV-1 management.

Materials and Methods

This study was carried out by the Ministry of Health of the Argentine Republic with the support of PAHO and the Argentine Group for HIV Drug Resistance Surveillance 2019, in accordance with the WHO guidelines for HIVDR surveillance (Appendix A1). A random selection was made among all the centers where ART was initiated through the public health system of the country (115 centers), represented by 19 centers selected among the main treatment centers in the country (10% of the centers with the lowest volume of individuals were discarded, guaranteeing a sampling frame that represents 90% of the new ART prescriptions).

The participating centers of the study were selected by sampling type “Probability Proportional to Proxy Size (PPPS).” The sample size was calculated, aimed at obtaining an HIVDR point prevalence rate with a confidence interval (CI) of maximum +5% for two populations: All individuals starting ART and all individuals starting ART without previous ART exposure.

The following assumptions were made for the sample size calculation: Clinics to be sampled using PPPS method-estimated prevalence of pretreatment HIVDR, 10%; proportion of genotyping failure, 20%; proportion of ART initiators with (or unknown) prior exposure to ART, 25%; and proportion of individuals starting NNRTI-based regimens in Argentina, 70%. Participants were classified as having pretreatment drug resistance if they harbored HIV with drug resistance mutations that conferred resistance to at least one of the antiretroviral drugs according to the Stanford algorithm. Fisher's exact or Chi-squared test was used to compare proportions, and Mann–Whitney U to compare continuous variables between ART-exposed and nonexposed groups.

The minimum number of individuals to be included in the study was 360 individuals (since this study was associated with the analysis of early warning indicators, a higher number of individuals were finally enrolled). In the centers that failed to include at 6 months, the number assigned of individuals at the beginning of the study was compensated by another center in the same region with similar characteristics. Participant registration and sample collection began in April 2019 and ended in December 2019 (9 months).

Samples were analyzed by three national reference laboratories (Unidad de Virología, Hospital Muñiz, Laboratorio Central de Córdoba; Instituto de Investigaciones Biomédicas en Retrovirus y SIDA); and were sequenced by an in-house system using Big Dye Terminator v3.1 (Thermo Fisher) and Sanger sequencing. Data generated was in FASTA format created from chromatograms produced by the Sanger sequencing platforms. Raw chromatograms were analyzed using a web-based software RECall [Version v2.30 (last updated May 11, 2022)], as recommended by the WHO for laboratories in the HIVDR network. All samples with resistance mutations and random 10% of specimens with no resistance mutations were also analyzed in a WHO-accredited laboratory (CIENI, México).

The ethical principles for medical research were respected and approval of the study was requested by the different ethics committees of the participating centers. All participants accepted and signed the respective informed consent.

Results

Between April and December 2019, 447 individuals with HIV-1 from 19 centers in Argentina were enrolled in the study. The distribution of the centers and individuals enrolled are concordant with the characteristics of the HIV epidemic in Argentine, with more than 45% of cases concentrated in Buenos Aires city and surroundings.

Plasma samples were obtained from 425 individuals before initiation or restart of ART. The baseline characteristics of the population are described in Table 1.

The median viral load in naive individuals was 4.7 log10 copies/mL and 4.2 log10 copies/mL for exposed individuals. The median CD4 in naive individuals was 348.5 cells/mm³ (IQR 157–532) and 232 cells/mm³ for exposed individuals (that represents 21% of the study population) (IQR 96–342, p = .03). It is important to highlight that 30% of nonexposed individuals and 53.8% of history of previous ART treatment presented CD4+ cell count lower than 200 cells/mm³ at enrollment, while 53% and 76.3% presented values less than 350 cells/mm³, respectively. These data once again confirm the high prevalence of new late diagnoses and gaps in the cascade of HIV care in Argentina.

The diagnosis of HIV-1 was carried out in 85.2% of naive individuals in 2019, 2.2% in 2018, 1.6% between 2015 and 2016, and the remaining between 2003 and 2014. Of the individuals previously exposed to ARVs, the median time from diagnosis of HIV infection and study entry was 8 years (IQR 3–12 years). There is no information on the time with or without ART before the study in this population.

Among individuals previously exposed to ART, 93% had received it as antiretroviral therapy, followed by its prescription as prophylaxis of mother to child transmission (4%).

During the follow-up, centers reported data from 409 individuals, of whom 333 (81.5%) individuals had initiated ART, 84% of previously exposed individuals, and 72% of naive individuals. The median time between diagnosis and initiation of ART in naive subjects was 31.3 days (IQR 15–76 days). In individuals with prior exposure to ART, 25% restarted the last ART combination they had received. There was a significant difference with the ART reported to be prescribed in the 2016 report, according to the evolution of antiretroviral treatment recommendations. The proportion of treatment-naive individuals who started ART with recommended first-line regimens (based on dolutegravir or darunavir) was 74.3%.

Of the 425 samples obtained, 393 could be sequenced, representing 93% of the total. From the analysis of integrase and protease/retrotranscriptase regions, 50.4% of the samples were classified as recombinant B and F subtypes, 44.8% as B, 3.9% as C, and 0.84% as others (BC, AG).

The prevalence of mutations associated with resistance was detected in 27.7% (95% CI 25.6–34.9%) of the study population. NNRTIs were found in 19.6% of the samples (95% CI 16.3–24.5%), integrase strand transfer inhibitors (INSTI) in 6.1% (95% CI 6.1–11.9%), nucleoside/nucleotide reverse transcriptase inhibitors (NRTI) in 3% (95% CI 1.9–5.9%), and PI in 1.5% (95% CI 0.7–3.6%) (Table 2 and Fig. 1).

OPEN IN VIEWER

FIG. 1.

Distribution of resistance-associated mutations.

The mutation K103N/S/R was the most frequently found: 40.8% (95% CI 31.9–50.2%) of the samples (Table 3).

The high prevalence of INSTI resistance is due to the presence of some polymorphisms such as the G163RK naturally present in some non-B subtypes circulating in Argentina.

There was no difference in the rate of resistance-associated mutations (RAMs) according to gender (male 25% and females 24%), area of residency, or HIV subtype (except for G163RK).

According to previous exposure to ARVs at the population level, it was observed that in samples of naive individuals, 16.8% (95% CI 12.8–21.4) presented variants of resistance to NNRTIs. This represents a statistically significant increase (p = .03) with respect to the prevalence reported in 2014–2015 (10%, 95% CI 6–14). Moreover, it was observed that 5.7% (95% CI 2.9–15.9) had variants associated with resistance to INSTI in this population. It was not possible to compare the evolution of resistance to INSTI since it was not analyzed in the previous study. For experienced individuals, the prevalence of variants associated with resistance was 30.38% (95% CI 20.8–42.2) for NNRTIs and 7.7% (95% CI 2.9–15.9) for INSTI. It should be noted that 51% of the experienced individuals had previously received EFV or NVP.

Discussion

This study reflects not only the evolution of HIV-1-associated RAM in the Argentinean population but also some of the characteristics of individuals with recent HIV-1 diagnosis.

The median CD4 count in people starting ART was below 350 cells/mm³, which reflects a high rate of late diagnosis in this sample, consistent with national data. The degree of immunosuppression at the time of restarting ART in previously exposed people is more pronounced (232 cells/mm³), with a significant proportion of people in advanced stages of the disease.

The time frame observed in the study from the diagnosis of HIV-1 to the start of ART initiation was 31 days among people not previously exposed. This is in line with the strategy of early initiation of ART; nevertheless, it is still far from achieving the objective of the strategy of early ART initiation, the “test and treat” strategy. This highlights the need for strengthening the policies to achieve the goal of ART initiation within 7 days and ideally on the day of diagnosis.

The prevalence of HIV-1 resistance in people starting ART in Argentina has changed since the last surveillance study, with a significant increase in RAM associated with NNRTI resistance. Even though the country exhibits a moderate level of drug resistance, the current rate of resistance requires pre-treatment resistance studies to be carried out in both exposed and previously unexposed individuals if they received ART containing NNRTI. The use of EFV as a first-line treatment-based regimen for more than 2 decades, in the context of universal access to ART, is most likely the factor associated with the rise in the frequency of RAMs associated with resistance to this family of drugs. ⁹ The use of pre-treatment resistance studies when starting regimens containing INSTI or PI is not mandatory, but it is recommended.

As in the previous study, no association between ART resistance and gender, area of residence, or HIV-1 subtype was observed. The most frequent HIV-1 viral subtypes remain unchanged, with recombinant forms BF and B representing more than 90% of the samples sequenced.

The rate of samples that were not possible to amplify was lower than 10%, and there was no difference regarding the region of the country where the sample was taken or the laboratory responsible for the assay.

For the first time, the presence of pretreatment RAM at the integrase region was evaluated in Argentina. It is important to highlight that the percentage of RAM to INSTI was found to be higher than expected. This is directly related to a polymorphism (G163RK) associated with F present in Argentina as a recombinant form with B subtype, which confers a decrease in the response to some of the drugs included in this ARV class. ¹⁰ This polymorphism affects raltegravir and elvitegravir (reducing their potency), but not other drugs of the same class, like dolutegravir, behaving in these cases as an accessory mutation.

The frequency of this variant associated with resistance is higher than that found in other parts of the world, having reported frequencies of 0 to 1.2% in Ethiopia, ¹¹ Sub-Saharan countries, ¹² Korea, and the United States, ¹³ but higher levels for subtype F. ¹⁴ This polymorphism could be associated with the reduced efficacy of raltegravir-based regimens reported in Argentina. ¹⁵ If G163 and T97 were excluded from the analysis, the prevalence of INSTI RAM in naive individuals would be 2.85% (95% CI 1.31–5.35%).

Given this finding, it is relevant to carry out studies in the region to define the prevalence and impact of these polymorphisms on antirretrovirals, especially focusing on NSTI.

Conclusions

Moderate rates of resistance to NNRTIs both in exposed and unexposed people in Argentina are observed in this study. This finding could be related to the previous use of EFV as the first line of treatment.

On the other hand, the present work generates the framework of evidence that supports the use of schemes based on high genetic barrier integrase inhibitors such as dolutegravir as the first line of treatment, and the need for the use of resistance studies before prescribing treatments based on non-nucleoside inhibitors of the reverse transcriptase.

It is reported for the first time, the presence of mutation linked to first-line integrase inhibitors such as raltegravir in one of the most prevalent recombinant viral subtypes in Argentina. This finding highlights the need to conduct more studies focused on polymorphisms associated with subtypes and recombinant forms circulating in our region.

Finally, this study once again confirms that the surveillance recommended by the WHO is a feasible and very useful tool for HIV programs.