Transmitted Drug Resistance Among Recently Diagnosed Adults and Children in São Paulo,Brazil

Abstract

Transmitted drug resistance mutations (TDRM) have been a constant threat to treatment efficacy. We evaluated TDRM in plasma RNA of 217 antiretroviral therapy-naive patients from sites in the São Paulo metropolitan area, collected from 2012 to 2014. The partial HIV-1 polymerase region was sequenced using Big Dye terminators at an ABI 3130 Genetic Analyzer. TDRM was defined according to the Stanford database calibrated population resistance (CPR v.6.0), but other drug resistance mutations (DRM) considered at the IAS list (IAS, 2014) and at the Stanford HIV Database Genotyping Resistance Interpretation (GRI-HIVdb) were also described. Out of 78% (170/217) of patients with information on the time of diagnosis, most (83%, 141/170) had been recently diagnosed, with the first positive HIV serology at a median of 58 days (IQR 18–184). Subtype B predominated (70%), followed by subtype F (10%), BF (7.5%), C (7.5%), and BC (5%). TDRMs were observed in 9.2% (20/217, CI 95% 5.9% to 13.6%), mostly (5.2%) to nonnucleoside reverse transcriptase inhibitor (NNRTI) antiretroviral class. Among children and adolescents, only a single patient showed TDRMs. Additional non-CPR mutations were observed: 11.5% (25/217) according to IAS or 4.6% (10/217) according to GRI-HIVdb. Overall, 23.5% (51/217) of the cases had one or more DRM identified. TDRM prevalence differed significantly among some sites. These trends deserve continuous and systematic surveillance, especially with the new policies of treatment as prevention being implemented in the country.

Introduction

With enhanced access to antiretroviral therapy (ART), reductions in morbidity and mortality related to HIV infection have become even clearer, improving the quality of life of patients living with HIV/AIDS. However, the development of resistance to antiretroviral drugs and the potential for its transmission have been constant threats as they may reduce therapy efficacy.^1,2 The pretreatment genotypic test identifies resistance that might compromise initial therapy, and upon treatment failure allows a better virological response with a genotypic test-guided therapy.³ However, its cost and accessibility have limited its widespread use in resource-limited settings. Monitoring the prevalence trends of transmitted drug resistance mutations (TDRM) could help in the identification of more adequate antiretroviral combinations and generate prevalence estimates that, when reaching a given threshold, indicate the need for the introduction of the pretreatment genotypic test as part of routine care.^{4

–25}

In Brazil, most studies in TDRM prevalence documented low to intermediate rates,^{4,7
–9,10,15,19,25} but different potential limitations can compromise comparability of these studies such as differences in sample size and recruitment design. We reevaluate, using a similar methodology, previous study sites⁴ and new ones in the major metropolitan area of São Paulo, Brazil.

Materials and Methods

Blood samples from 226 HIV-1 antiretroviral therapy-naive patients were collected from May 2012 to July 2014 in three areas of the São Paulo metropolitan area: (1) sites in the city of São Paulo; (2) sites in neighboring cities southeast of São Paulo, ABCD (Santo Andre, São Bernardo do Campo, São Caetano do Sul, and Diadema); and (3) sites in Campinas, the second largest metropolitan area in the state, about 100 km from São Paulo. The study enrolled consecutive HIV-1 antiretroviral therapy-naive patients of all age groups recruited specifically for the study or from physicians' request as part of the clinical follow-up. Clinical and demographic data were obtained at baseline visit when informed consent was obtained from subjects or either parents or guardians. Patients younger than 12 years old were categorized as children, 12 years to 18 years as adolescents, and those over 18 years old as adults. Time from first HIV serological diagnosis allowed classifying patients as recent (less than 2 years from diagnosis) and long-term HIV infection (over 2 years from diagnosis).

Sequences were obtained from plasma viral RNA, extracted (QIAmp viral RNA mini kit, Qiagen, Hilden, Germany) and amplified with an “in-house” adapted methodology. The amplification of the partial HIV polymerase region (codons 1–235 related to HXB2, accession number K03455) was obtained using primers previously described^5,24 in an adapted nested polymerase chain reaction (PCR).^4,25 The reverse transcriptase (RT)-PCR was performed using 10 μl of viral RNA in a 50 μl volume reagent mix containing 10× PCR buffer High Fidelity, 50 mM MgSO₄, 10 mM dNTP's (Life Technologies, USA), 0.1 M DTT, 10 U RNase inhibitor (New England Biolabs, UK), 50 U RT SuperScript III (Life Technologies, USA), 1.25 U Platinum Taq DNA Polymerase High Fidelity (Life Technologies, USA), and 10 μM of primers (GAG2: 5′-GAGGAAGCTGCAGAATGGG-3′ forward) and (RT137: 5′-TTCTGTATGTCATTGACAGTCCAGC-3′ reverse). The RT-PCR was carried out at 50°C for 30 min and 94°C for 5 min, followed by amplification cycles: 18 cycles at 94°C for 30 s, 50°C for 30 s, 68°C for 2 min 30 s; 17 cycles at 94°C for 30 s, 53°C for 30 s, 68°C for 2 min 30 s, and a final extension at 68°C for 10 min.

A 2.5 μl volume of the RT-PCR product was used in a nested PCR performed with Go Taq Green Master Mix 2X (Promega Biosciences, CA) and 10 μM of primers (PR3: 5′-AGAGCCAACAGCCCCACCA-3' forward) and (RT12: 5′-ATCAGGATGGAGTTCATAACCCATCCA-3' reverse). The nested PCR was carried out at 94°C for 3 min followed by 35 cycles of at 94°C for 30 s, 55°C for 30 s, 72°C for 1 min 30 s, and a final extension at 72°C for 10 min.

An unique fragment of approximately 1,112 bp was sequenced with Big Dye using eight primers to cover the partial pol region (codon 1 to 235). Sequences were manually edited using Sequencher 4.7 software (Gene Codes, USA). Ambiguous DNA bases (mixtures) were considered at sequence edition and were counted as a resistance-associated codon when one or more deduced amino acids were considered resistance mutations.

HIV genotyping resistance test results (Stanford HIV database genotype resistance report, GRI-HIVdb) were reported to the clinical sites to subsidize clinical follow-up. TDRMs were defined according to the Calibrated Population Resistance Version 6.0 (CPR, Stanford Database, SDRM 2009), a specifically designed algorithm for the epidemiologic surveillance of HIV-1 transmitted drug resistance mutations(DRM). DRMs listed in bold in the International Antiviral Society (IAS, 2014) and the Stanford HIV Database Genotyping Resistance Interpretation (GRI-HIVdb) were also described to evaluate additional mutations that might have an impact on ART treatment.

HIV-1 subtyping was initially performed with NCBI Genotyping and REGA HIV Subtyping tools and confirmed with phylogenetic analysis using BEAST v1.7.4, under the GTR+I+G model. Recombinant patterns were evaluated by SimPlot (v. 3.5.1 recombination software).

Statistical analysis

Data were analyzed with Epi Info 6 (CDC, Atlanta, GA), applying a level of statistical significance of 0.05, two tailed. Categorical variables were tested using Yates or Fisher's exact test. Results of continuous variables were expressed as the median and interquartile range (IQR). Mid-p 95% was used to calculate the confidence interval.

Ethical approval

This study was approved by Ethical and Research Committees of Adolfo Lutz Institute, Secretary of Health of São Paulo, Brazil and participating institutions.

Results

From 226 HIV-1 antiretroviral therapy-naive patients enrolled in this study, we could successfully amplify and perform 217 (96%) partial HIV-1 pol sequences, except for one case with only the protease region.

Demographic characteristics are depicted in Table 1. Three women enrolled in the study had previous exposure to short-term use of ART for mother-to-children transmission prophylaxis, all of them without detected resistance mutations; one additional male patient (BRSP417-14) with long-term infection was found subsequently to have had a previous 1 month prescription of zidovudine (AZT)+ didanosine (ddI). This case had a secondary non-CPR IP mutation considered by the Stanford Database algorithm to confer potential low-level resistance to tipranavir. Ten women (4.6%) were pregnant, but none of them had been exposed to ART prophylaxis at blood collection. Perinatal infection was found mostly among children (11/17, 64.7%) and then adolescents (2/14, 14.2%), with some adolescents (3/14, 21.4%) reporting homosexual behavior.

Table 1.

Demographic and Laboratory Data by Age Group

	Adults (n=186)	Children and adolescents (n=31)	Total (n=217)
Age (years)	31 (25–38)	9 (4–14)	29 (22–37)
Male gender [n (%)]	141 (75.8)	15 (48.5)	156 (72)
CD4 cell count (cells/mm³)	496 (354–745)	569 (392–1168)	506 (354–778)
Viral load (log₁₀)	4.40 (3.82–4.97)	4.68 (4.43–5.07)	4.52 (3.91–4.99)
HIV risk factor [n (%)]
MSM/bisexual	92 (49.5)	3 (9.7)	95 (43.7)
Heterosexual male	20 (10.8)	0 (0)	20 (9.2)
WSM	28 (15)	0 (0)	28 (12.9)
Perinatal	0 (0)	13 (41.9)	13 (6.0)
No information	43 (24.7)	15 (48.4)	61 (28.2)
HIV-1 subtype [n (%)]
B	132 (70.9)	19 (61.2)	150 (70)
F	14 (7.5)	8 (25.8)	22 (10)
C	14 (7.5)	2 (6.5)	16 (7.5)
Recombinant BF	15 (8.0)	2 (6.5)	16 (7.5)
Recombinant BC	11(6.0)	0 (0)	11 (5.0)

Demographic and laboratory characteristics of patients studied by age group. Values are expressed as median (25th–75th IQR) or number of cases (percentage in parentheses). MSM, men who have sex with men; WSM, woman who have sex with men.

From 217 patients, 78% (170/217) had information about the time of diagnosis; most (141/170, 83%) had been recently diagnosed, with the first positive HIV serology at a median of 58 days (IQR 18–184), with 17% (29/170) classified as long-term infection (median 1998 days, IQR 1281–3523).

As observed in previous studies,^4,7

–10 subtype B predominated in 70% (130/186) of HIV-1 pol sequences among adults, with subtypes C and F in 15/186 (8%) and 14/186 (7.5%), respectively. Recombinant forms of subtype BF were observed in 13/186 (7%) cases and BC in 11/186 (6%). HIV-1 pol sequences from children and adolescents also showed a predominance of subtype B in 64.5% (20/31), but a significantly higher prevalence of subtype F (26%, 8/31) (p=0.005) as compared to adults. Three isolates (9.4%) were subtype BF and one subtype was C.

According to CPR, TDRM was detected in 20/217 (9.2%) isolates, with 65% (16/20) subtype B. TDRM was not observed in subtype C isolates, but was present in subtype F, BC, and BF recombinants. The prevalence of TDRM was higher at the Campinas site (7/44, 16%) compared to the ABCD site (5/99, 5%) (p=0.047; if only adults are considered, p=0.056). At São Paulo metropolitan area TDRM was 11% (8/74). Mutations to nonnucleoside reverse transcriptase inhibitors (NNRTI) were most frequent (4.6%, 10/217), with a higher prevalence of the K103N mutation (43%, 9/21). The prevalence of TDRM to the nucleoside reverse transcriptase inhibitor (NRTI) class was 3.6% (8/217) and to the protease inhibitor (PI) class 1.8% (4/217) (Table 2). Two-class resistance was uncommon, and was observed in only two cases (0.9%, one NRTI+PI and one NNRTI+PI, Table 2). When only children or adolescents were considered, only 1/31 (3%) sequences had a CPR-defined mutation.

Table 2.

Patients' Sequences with Drug Resistance Mutations

			DRM mutation found
Patient ID	Age/gender	Site	NRTI	NNRTI	PI	Subtype
BRSP098-12	23/M	ABCD	V106I			F
BRSP147-12	31/F	ABCD	T215I, V106IV			B
BRSP237-12	25/M	Campinas	M184I			B
BRSP249-12	37/M	Campinas		K103N		B
BRSP252-12	32/F	Campinas		V108I		B
BRSP260-12	62/F	São Paulo	K70N			B
BRSP308-12	32/M	Campinas	N/A	N/A	L76V	B
BRSP310-12	31/F	Campinas		K103N		BF
BRSP316-12	31/M	Campinas		V179DE		B
BRSP319-12	29/M	Campinas		K103N		B
BRSP323-12	33/M	Campinas			Q58E	B
BRSP326-12	26/F	Campinas			Q58E	B
BRSP431-12	52/M	Campinas	V106I			F
BRSP433-12	36/F	Campinas	V106I			F
BRSP434-12	51/M	Campinas	V106I			B
BRSP539-12	31/M	São Paulo		Y188H , G190A ,H221Y		B
BRSP556-12	19/M	ABCD	T215S			B
BRSP001-13	2/F	ABCD	T69NT			B
BRSP034-13	36/M	São Paulo	V106IV			B
BRSP070-13	21/M	ABCD		K103N , P225H ,V179LV		B
BRSP162-13	49/M	ABCD		E138A		B
BRSP257-13	39/M	ABCD		E138A		B
BRSP358-13	11/M	São Paulo		K103N	I85V	F
BRSP489-13	38/M	São Paulo			I85V	B
BRSP574-13	23/M	ABCD		E138A		B
BRSP588-13	0/F	São Paulo	V106IV			F
BRSP616-13	29/M	São Paulo		F227FL		B
BRSP630-13	16/F	São Paulo		E138AE		C
BRSP718-13	32/M	São Paulo		V108I		B
BRSP728-13	20/M	São Paulo		K103N		B
BRSP788-13	39/M	ABCD		E138A		B
BRSP835-13	24/M	ABCD		K103N		B
BRSP864-13	52/M	ABCD		V179D		B
BRSP001-14	6/M	São Paulo	T69N			B
BRSP050-14	42/M	São Paulo		V179D		B
BRSP054-14	32/M	ABCD		E138AE		B
BRSP078-14	25/M	ABCD		E138A		B
BRSP158-14	38/F	São Paulo	M41L , T215D			B
BRSP185-14	28/M	ABCD		E138A		B
BRSP310-14	42/M	São Paulo		V90IV, K101E	I50L , Q58E, V82M, L90M	BC
BRSP339-14	26/M	São Paulo	K219R			BF
BRSP354-14	33/M	ABCD	T69N			B
BRSP395-14	56/F	ABCD			M46IMV ^*	B
BRSP417-14	35/M	São Paulo			K43T	B
BRSP536-14	27/M	São Paulo		K103N , E138A		B
BRSP554-14	18/M	ABCD			T74S	C
BRSP618-14	54/M	Campinas	M41L , T215V			BF
BRSP627-14	30/M	ABCD		V179D		BF
BRSP648-14	37/M	ABCD		E138A		B
BRSP796-14	38/M	ABCD		K103N		BF
BRSP892-14	56/F	Campinas	M184V	V179T		B

Patients' HIV-1 sequences with one or more drug resistance mutations (DRM) by drug class as NRTI, nucleoside reverse transcriptase inhibitors; NNRTI, nonnucleoside reverse transcriptase inhibitors; PI, protease gene. N/A, not available; M, male; F, female.

Clinical sites are grouped by geographic areas as São Paulo, ABCD, and Campinas. Aminoacid substitutions listed in CPR (TDRM) as drug resistance mutations are in bold, additional DRM listed in the IAS 2014 list are in italic (bold and italic for those listed in both lists), and additional mutations considered at the GRI-HIVdb are underlined.

The mutation M46I is a surveillance TDRM according to the CPR list, but the mixture observed (M46IMV) was not computed as TDRM when the study sequences were evaluated at the CPR site.

When we compared previous reported estimates⁴ among adults from sites evaluated in this study, we found a nonsignificant trend in Campinas (15/156, 9.6% in 2009 and 7/44, 15.9% in 2013; p=0.3). In the city of São Paulo a significant increase in TDRM prevalence was observed, from 3% (2/69) to 14% (7/50) (p=0.03). The ABCD sites had no previous evaluation.

DRMs listed in bold in the IAS list (2014) were found in 37/217 (17%) of HIV-1 sequences, with 3/31 (9.6%) among children and adolescents. Some of these sequences also had other CPR mutations. Most of these non-CPR mutations (60%, 22/37) were associated with newer drugs of the NNRTI class, such as E138A (4.6%, 10/217), V179D/E/L (2.3%, 5/217), and H221Y (0.46%, 1/217). Ten cases presented at least one additional amino acid substitution considered by the GRI-HIVdb.

Regarding the NRTI drug class, a limited impact in susceptibility is expected as we observed an important loss of susceptibility only when two or more mutations were present, with the exception of mutation M184V, which itself confers high resistance to lamivudine. In the protease region some polymorphisms were found (Q58E and K43T) that are associated with low/potential-low resistance to tipranavir, a new generation PI drug. One case exhibited high resistance to lopinavir with the L76V mutation and another one showed three PI resistance mutations, which conferred resistance to all first generation PI drugs.

Discussion

Monitoring ART resistance among untreated individuals is an important public health tool, but comparability among studies is subject to different bias that may influence results, as the sampling in geographic areas with an uneventful distribution of TDRM and, equally important, potential heterogeneity among patients' subpopulations.³ One major factor is the stringency in the exclusion of previously exposed individuals, a problem that does not seems to be trivial when this issue is properly evaluated.⁶ We reevaluated clinical information from all cases with detectable resistance mutations to exclude ARV exposure, but we included few cases in which short-term ARV exposure was documented, with none presenting mutations associated with regimens used.

We tried to expand our monitoring by the inclusion of new sites, but we reevaluated previous studied sites, maintaining the recruitment and other inclusion and exclusion criteria as close as possible to a former study from 2009⁴ to allow a follow-up evaluation. The new geographic area included here comprised four middle-sized cities (ABCD). This work is the first study including this specific area, which showed a lower TDRM estimate.

Our study demonstrates a TDRM prevalence of 9.2%. Overall, our estimates are similar to those reported from many other studies in Brazil, with intermediate estimates ranging from 5% to 15%.^4,7
–9 A study among HIV-infected blood donors in Brazil found a TDRM prevalence of 19.4%, the highest reported in São Paulo,¹⁰ but the confidence interval (CI 95%) for this estimate ranged from 9.5% to 29.2%. TDRM levels higher than 15%, as observed in our study for Campinas, but also with large CI 95%, have been reported in some Brazilian cities, such as Santos and Salvador.^11,12 One of these studies¹¹ was based on a small sample size (only 47 cases per site) and the other¹² included non-CPR mutations in the estimates. However, these cities deserve further attention, especially due to the fact that NNRTI mutations are the most frequently observed, potentially compromising the recommended NNRTI-based first line ART regimen.¹⁸

Regarding children and adolescents, risk exposure was underestimated due to the lack of proper information, but perinatal exposure was the main route of transmission reported among children. A single patient presented TDRM considered by CPR, which was the K103N mutation, contrasting with studies that show a prevalence varying from 8.7% to 12.8%.^13

–16 However, a Brazilian study conducted between 2000 and 2004 evaluated a cohort of 24 children and did not found any TDR.¹⁷ These studies with children are still limited to properly evaluate the prevalence of TDRM in this population.

As expected, subtype B remains predominant while non-B subtypes and recombinants were found at a lower frequency, similar to other observations in the area.^4,7,10 In cities located in the southern part of Brazil the scenario for the epidemic is different, with a predominance of subtype C and CB recombinants. Some studies observed that TDRM seems to be less frequently found among subtype C,^19,20 but the small sample size of non-B subtypes does not allow a proper evaluation of this issue. It is worth noting that subtype C and CB recombinants have attained proportions similar to subtype F and BF recombinants, an increase in the subtype C already observed ²¹ in other studies.

We reported our TDRM results using CPR to determine prevalence rates, enabling a comparison to other molecular epidemiology studies, but we also evaluated additional DRMs that are considered only by IAS or GRI-Hdb. Some of those mutations may compromise new NNRTIs such as rilpivirine or etravirine, particularly in the E138A mutation. Although not described in previous studies in the country, when we reanalyzed our dataset,⁴ we observed that this mutation was already present in some samples, with a similar proportion (E138A 4%, 9/225).The presence of E138A among subtype B sequences in our study, 6.9% (9/130, CI 95% 3.4–12), is somewhat higher than that observed by Sluis-Cremer²² among subtype B sequences from antiretroviral therapy-naive patients at large databases (2.3% in the Stanford database and 2% in the British Columbia Centre database), and our findings are actually similar to findings observed in subtype C in these databases (6.1% and 5.9%, respectively).

Lambert-Niclot and colleagues evaluated the prevalence of rilpivirine-associated mutations in a French cohort of 1,729 antiretroviral therapy-naive patients and found mostly the E138A mutation among non-B subtypes (4.1%, 35/867, versus 2%, 17/862, in subtype B).²³ This mutation seems to be relevant only to the new NNRTI. Etravirine is used in the country only in salvage therapy,¹⁸ a situation in which drug activity may be limited due to class recycling, and the other, rilpivirine, is not yet available through public access. However, the latter is being recommended for first line therapy and evaluated as long-acting nanosuspension formulations that might provide new preexposure prophylaxis strategies (PrEP),^26,27 situations in which our findings may become more worrisome.

Conclusions

Overall, we observed an increase in TDRM among adults when compared to data from a previous study, but rates are within the range observed in other studies. Children and adolescents show a lower prevalence of TDRM. Continuous monitoring of the situation as well as pretreatment genotyping assessment are necessary, especially taking into account new policies¹⁸ regarding treatment as prevention that are being implemented in the country.

Sequence Data

Sequences are available at GenBank with accession numbers KP115360 to KP115579.

Footnotes

Acknowledgments

Participants in the São Paulo Salvage Workgroup who contributed to this work included Max Igor Banks Ferreira Lopes, Margareth da Eira, Denise Peluso Pacola, Yu Ching Lian, Roberio Carneiro, Alina Bernardes Habert, Fabio AB Cabral, Lucy Vasconcelos, Paulo Friggi de Carvalho, Rafaella Shimada, Marcina Garcia, Maria Roberta Lopes, Adriane Maira Delicio, and Ana Cláudia Diaz.

This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) [grant 2011/21958-2].

Author Disclosure Statement

No competing financial interests exist.

References

Di Biagio

, Rusconi

, Marzocchetti

, et al.: The role of baseline HIV-1 RNA, drug resistance, and regimen type as determinants of response to first-line antiretroviral therapy. J Med Virol, 2014; 86(10):1648–1655.

Cambiano

, Bertagnolio

, Jordan

, et al.: Transmission of drug resistant HIV and its potential impact on mortality and treatment outcomes in resource-limited settings. J Infect Dis, 2013; 207(Suppl 2):S57–62.

Zu Knyphausen

, Scheufele

, Kücherer

, et al.: First line treatment response in patients with transmitted HIV drug resistance and well defined time point of HIV infection: Updated results from the German HIV-1 seroconverter study. PLoS One, 2014; 9(5):e95956.

Ferreira

JLP

, Rodrigues

, Lança

, et al.: Transmitted drug resistance among people living with HIV/Aids at major cities of São Paulo State, Brazil. Adv Virol, 2013; 2013:878237.

Steegen

, Demecheleer

, De Cabooter

, et al.: A sensitive in-house RT-PCR genotyping system for combined detection of plasma HIV-1 and assessment of drug resistance. J Virol Methods, 2006; 133(2):137–145.

Fogel

, Wang

, Parsons

, et al.: Undisclosed antiretroviral drug use in a multinational clinical trial (HIV Prevention Trials Network 052). J Infect Dis, 2013; 208(10):1624–1628.

Gonsalez

, Alcalde

, Nishiya

, et al.: Drug resistance among chronic HIV-1-infected naïve for use of anti-retroviral therapy in São Paulo city. 2007; 129(2):87–90.

Inocencio

, Pereira

, Sucupira

MCA

, et al.: Brazilian Network for HIV Drug Resistance Surveillance: A survey of individuals recently diagnosed with HIV. AIDS Soc, 2009; 12:20.

Sanabani

, Pastena

ÉR

, da Costa

, et al.: Characterization of partial and near full-length genomes of HIV-1 strains sampled from recently infected individuals in São Paulo, Brazil. PLoS One, 2011; 6(10):e25869.

10.

Alencar

, Sabino

, Carvalho

, et al.: HIV genotypes and primary drug resistance among HIV-seropositive blood donors in Brazil: Role of infected blood donors as sentinel populations for molecular surveillance of HIV. J Acquir Immune Defic Syndr, 2013; 63(3):387–392.

11.

de Moraes Soares

, Vergara

, Brites

, et al.: Prevalence of transmitted HIV-1 antiretroviral resistance among patients initiating antiretroviral therapy in Brazil: A surveillance study using dried blood spots. J Int AIDS Soc, 2014; 17(1):19042.

12.

Sucupira

, Caseiro

, Alves

, et al.: High levels of primary antiretroviral resistance genotypic mutations and B/F recombinants in Santos, Brazil. AIDS Patient Care STDS, 2007; 21(2):116–128.

13.

Soto-Ramirez

, Rodriguez-Diaz

, Harris

, and Hazra

: HIV drug resistance-associated mutations in antiretroviral naïve HIV-1-infected Latin American children. Adv Virol, 2010; 2010:407476.

14.

Ferreira

FGF

, Pinto

, Kakehasi

, et al.: Prevalence of primary drug resistance-associated mutations among HIV type 1 vertically infected children in Belo Horizonte, Brazil. AIDS Res Hum Retroviruses, 2010; 26(2):229–232.

15.

Arruda

, Simões

, Sucupira

, et al.: Short communication: Intermediate prevalence of HIV type 1 primary antiretroviral resistance in Ceará State, Northeast Brazil. AIDS Res Hum Retroviruses, 2011; 27(2):153–156.

16.

Almeida

, Rodrigues

, Zaparoli

, et al.: Prevalence of transmitted HIV-1 drug resistance mutations in children and adolescents in São Paulo, Brazil. Pediatr Infect Dis J, 2012; 31(12):e255–257.

17.

Almeida

, Berezin

, Rodrigues

, et al.: Diversity and prevalence of antiretroviral genotypic resistance mutations among HIV-1-infected children. J Pediatr (Rio J), 2009; 85(2):104–109.

18.

Brazilian protocol for ART treatment, accessed in November 5, 2014, available at www.aids.gov.br/publicacao/2013/protocolo-clinico-e-diretrizes-terapeuticas para-manejo-da-infeccao-pelo-hiv-em-adultos.

19.

de Medeiros

, Junqueira

, Matte

MCC

, et al.: Co-circulation HIV-1 subtypes B, C, and CRF31_BC in a drug-naïve population from Southernmost Brazil: Analysis of primary resistance mutations. J Med Virol, 2011; 83:1682–1688.

20.

Rodrigues

, Scherer

, Oliveira

, et al.: Low prevalence of primary antiretroviral resistance mutations and predominance of HIV-1 clade C at polymerase gene in newly diagnosed individuals from south Brazil. Virus Res., 2006; 116(1–2):201–207.

21.

Brígido

, Ferreira

, Almeida

, et al.: Southern Brazil HIV type 1C expansion into the state of São Paulo, Brazil. AIDS Res Hum Retroviruses, 2011; 27(3):339–344.

22.

Sluis-Cremer

, Jordan

, Huber

, et al.: E138A in HIV-1 reverse transcriptase is more common in subtype C than B: Implications for rilpivirine use in resource-limited settings. Antiviral Res, 2014; 107:31–34.

23.

Lambert-Niclot

, Charpentier

, Storto

, et al.: Prevalence of pre-existing resistance-associated mutations to rilpivirine, emtricitabine and tenofovir in antiretroviral-naive patients infected with B and non-B subtype HIV-1 viruses. J Antimicrob Chemother, 2013; 68(6):1237–1242.

24.

Stuyver

, Wyseur

, Rombout

, et al.: Line probe assay for rapid detection of drug-selected mutations in the human immunodeficiency virus type 1 reverse transcriptase gene. Antimicrob Agents Chemother, 1997; 41(2):284–291.

25.

Ferreira

JLP

, Thomaz

, Rodrigues

, et al.: Molecular characterisation of newly identified HIV-1 infections in Curitiba, Brazil: Preponderance of clade C among males with recent infections. Mem Inst Oswaldo Cruz, 2008; 103(8):800–808.

26.

Klooster

, Hoeben

, Borghys

, et al.: Pharmacokinetics and disposition of rilpivirine (TMC278) nanosuspension as a long-acting injectable antiretroviral formulation. Antimicrob Agents Chemother, 2010; 54(5):2042–2050.

27.

Rajoli

RKR

, Back

, Rannard

, et al.: Optimisation of intramuscular sustained release nano-formulations using in silico modelling. 21st Conference on Retroviruses and Opportunistic Infections (CROI), Boston, MA, March 3–6, 2014. Abstract #518.