JAIDS Journal of Acquired Immune Deficiency Syndromes:
Epidemiology and Prevention
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
Alencar, Cecília S. MSc*,†; Sabino, Ester C. MD, PhD‡; Carvalho, Silvia M. F. PhD§; Leao, Silvana C. MD‖; Carneiro-Proietti, Anna B. MD, PhD¶; Capuani, Ligia BSc#; Oliveira, Cláudia L. MD, PhD**; Carrick, Danielle PhD, MHS††; Birch, Rebecca J. MPH††; Gonçalez, Thelma T. MD, PhD‡‡; Keating, Sheila PhD‡‡; Swanson, Priscilla A. MSc§§; Hackett, John Jr PhD§§; Busch, Michael P. MD, PhD‡‡; for the NHLBI Retrovirus Epidemiology Donor Study-II (REDS-II), International Component
*Department of Infectious Disease, Federal University of Sao Paulo, Sao Paulo, Brazil;
†LIM03, Laboratório de Medicina Laboratorial/ Departamento de Patologia, Hospital das Clinicas FMUSP, Sao Paulo, Brazil;
‡Department of Infectious Disease/Institute of Tropical Medicine, University of São Paulo, Sao Paulo, Brazil;
§Hemorio, Rio de Janeiro, Brazil;
‖Hemope, Recife, Brazil;
¶Hemominas, Belo Horizonte, Brazil;
#Fundaçao Pro-Sangue/Hemocentro de Sao Paulo, Brazil;
**Federal University of Sao Joao del Rey, Divinopolis, Brazil;
††Westat, Rockville, MD;
‡‡Blood Systems Research Institute, San Francisco, CA;
§§Emerging Pathogens and Virus Discovery, Abbott Diagnostics, Abbott Park, IL.
Correspondence to: Ester C. Sabino, MD, PhD, Department of Infectious Disease/Institute of Tropical Medicine, University of Sao Paulo, Av. Dr Enéas Carvalho de Aguiar, 470, CEP 05403-000 São Paulo, Brazil (e-mail: firstname.lastname@example.org).
Supported by NIIH (HHSN268200417175C/HO/NHLBI NIH HHS).
The authors have no conflicts of interest to disclose.
The Retrovirus Epidemiology Donor Study-II, International Component (Brazil), is the responsibility of the following persons: Blood Centers—Fundação Pró-Sangue/Hemocentro São Paulo (São Paulo): E. C. S., Cesar de A-N, A. M. Jr, L. C., and N. A. S. Hemominas (Belo Horizonte, Minas Gerais): A. B., F. C-P, F. A. P., C. Di L. O., and C. M.; Fundação Hemope (Recife, Pernambuco): D. de A. S., S. A. C. L., and M. I. L.; and H. (Rio de Janeiro): C. L., M. E. L., and S. M. F. C. DataWarehouse—University of São Paulo (São Paulo): J. E. F., M. O., and P. L. T. US Investigators—Blood Systems Research Institute and University of California at San Francisco: M. P. B., E. L. M., B. C. and T. G. Coordinating Center—Westat, Inc: J. S., M. K. and K. K. National Heart, Lung, and Blood Institute, NIH: S. A. G.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.jaids.com).
Received October 18, 2012
Accepted February 18, 2013
Background: There are few surveillance studies analyzing genotypes or primary (transmitted) drug resistance in HIV-infected blood donors in Brazil. The aim of this study was to characterize patterns of HIV genotypes and primary resistance among HIV-seropositive donors identified at 4 geographically dispersed blood centers in Brazil.
Methods: All HIV-infected donors who returned for counseling at the 4 REDS-II Hemocenters in Brazil from January 2007 to March 2011 were invited to participate in a case–control study involving a questionnaire on risk factors. Viral sequencing was also offered to positive cases to assign genotypes and to detect and characterize primary resistance to reverse transcriptase and protease inhibitors according to World Health Organization guidelines.
Results: Of the 341 HIV-seropositive donors who consented to participate in the risk factor and genetics study, pol sequences were obtained for 331 (97%). Clade B was predominant (76%) followed by F (15%) and C (5%). Primary resistance was present in 36 [12.2%, 95% confidence interval (CI) 8.2 to 15.5] of the 303 individuals not exposed to antiretroviral therapy, varying from 8.2% (95% CI: 2.7 to 13.6) in Recife to 19.4% in São Paulo (95% CI: 9.5 to 29.2); there were no significant correlations with other demographics or risk factors.
Conclusions: Although subtype B remains the most prevalent genotype in all 4 areas, increasing rates of subtype C in Sao Paulo and F in Recife were documented relative to earlier reports. Transmitted drug resistance was relatively frequent, particularly in the city of Sao Paulo which showed an increase compared with previous HIV-seropositive donor data from 10 years ago.
Brazil had 608,230 reported cases of AIDS as of June 2011, the second highest number of cases in the Americas, exceeded only by the United States.1 Monitoring the genetic diversity of HIV-1 in Brazil (and elsewhere) is important to understand the molecular epidemiology and spread of the epidemic. The predominant HIV-1 clade in Brazil is B (∼70%), but unlike in the United States, Brazil also has appreciable numbers of HIV-1 infections caused by clades F, C, and circulating recombinant forms.2,3
As early as 1991, the Brazilian Ministry of Health provided antiretroviral therapy (ART) through its extensive public health system. In 1996, a law was enacted guaranteeing free access of ART to all Brazilians who required treatment, according to Brazilian guidelines. The widespread use of ART has led to a substantial decline in AIDS-related mortality in Brazil.4,5 However, it is expected that the proportion of patients experiencing virologic failure and consequently harboring resistant strains will increase over time. Depending on the behavioral characteristics of these individuals, transmission of drug-resistant strains may occur with increasing frequency.6 The transmission of resistant variants to uninfected individuals, termed primary resistance, raises serious clinical and public health consequences and may dramatically impair the capacity of treating HIV in the near future.7,8 Consequently, monitoring rates of transmission of drug-resistant strains is extremely important in a country like Brazil that provides free ART to approximately 200,000 HIV-infected individuals, many of whom have low levels of education and modest resources making compliance with drug regimens and, hence, resistance a serious concern.
The objective of this study was to characterize HIV strains by genotype and primary resistance in recently identified HIV-infected blood donors for whom detailed demographics, risk factors, and infection staging data were available. This study was designed to further inform Brazilian surveillance efforts and to complement similar monitoring of molecular variants among HIV-infected blood donors in the United States, other International Retrovirus Epidemiological Donor Study (REDS) II/III sites (China and South Africa), and other countries, thus allowing direct comparisons of these parameters in the donor base on a global level.
This study was part of the REDS-II International program.9 A case–control study comparing HIV-seropositive and seronegative blood donors was designed to understand risk exposures and motivations to donate blood among HIV-seropositive relative to matched, seronegative, control donors.10 A blood sample was also collected from all enrolled HIV-seropositive blood donors to characterize HIV genotype and drug resistance.
Subject Recruitment, Enrollment, and Questionnaire
We invited all HIV-seropositive blood donors who had been counseled for HIV from 2007 to 2011 in the 4 Brazilian sites of the REDS-II (Fundação Pro-Sangue, Hemominas, Hemope, and Hemorio).11 HIV-positive donors who were counseled from January 2007 to June 2009 were recruited by a letter, followed by a phone call; subsequent to June 2009, donors were prospectively recruited. If consent was obtained, a questionnaire about HIV risk factors and motivations for donation was administered through an audio computer-assisted self-interview (ACASI) system.10 For consenting HIV-seropositive subjects, we performed genotype and drug resistance testing on plasma samples obtained at the time of counseling, which were rapidly processed and frozen to optimally preserve HIV RNA; in contrast, the index donation sample was used to define recent infection status (see below) because this sample would best represent the stage of seroconversion and hence duration of infection at the time of the original seropositive donation.
Classification of Donors as Recently Infected by Incidence Assay Testing
For all HIV-seropositive donors who had donated during the REDS-II period (2007–2010) in the 3 primary REDS-II centers (Hemope, Hemominas, and Fundação Pro-Sangue), the index donation sample was saved and submitted to BSRI for testing using a less-sensitive (LS) or “detuned” enzyme immunoassay (Vironostika HIV-1 MicroElisa; bioMérieux, Durham, NC) or an LS chemiluminescent immunoassay (Vitros HIV-1/2 Assay; Ortho Diagnostics, Rochester, NY) as previously described.12 For the Hemorio center, LS enzyme immunoassay could only be performed if the index donation sample was available.
HIV-1 Clade Typing and Drug Resistance Testing in Sao Paulo
Subtype and resistance analysis was performed at Fundação Pró-Sangue as previously described.13 We have sequenced a fragment encompassing the protease gene and approximately 700 base pairs of the reverse transcriptase gene.13 The calibrated population resistance tool version 5.0 beta (available through the Stanford University HIV Drug Resistance Database http://cpr.stanford.edu/cpr.cgi) was used to identify transmitted drug-resistant mutations in untreated persons.14 Mutations listed as causing or contributing to resistance are nonpolymorphic in untreated persons and apply to all HIV-1 subtypes in accordance with the World Health Organization guidelines.14 Sequences were submitted to GenBank under the numbers JQ237931–JQ238236 and KC834581-KC834601.
HIV Sequencing and Viral Load Testing at Abbott Diagnostics
All samples that were refractory to polymerase chain reaction (PCR) amplification and sequencing in Sao Paulo, representative, successfully amplified samples that were classified as clade B, and all samples classified as non-B clade were sent to Abbott Diagnostics (Abbott Park, IL) for viral load testing using the Abbott RealTime HIV-1 assay (Abbott Molecular, Des Plaines, IL). In addition, samples that were negative by all PCR methods employed in Sao Paulo were retested at Abbott Diagnostics using the ViroSeq HIV-1 Genotyping System v2.0 (Celera, Alameda, CA).
We investigated the relationships between primary drug resistance and HIV subtype with donation characteristics (site, donation type, and first time/repeat donor), demographic characteristics (gender, age, education, and sexual orientation), HIV infection characteristics (recent or long-standing infection and subtype for primary drug resistance), and HIV risk factors as variables. The HIV risk factors were chosen based on their statistically significant relationship (P < 0.05) with HIV infection in the HIV Risk Factor Study.10 In addition to the unadjusted relationships shown in Tables 2 and 3, we applied stepwise logistic regression modeling, conditioned on hemocenter, with entry into the model set at P < 0.20 and P < 0.05 to stay in the model. None of the candidate predictor variables remained in the stepwise models for either the model of correlates of primary resistance or HIV subtype distributions. All statistical analyses were performed using SAS version 9.2 (SAS Institute, Inc, Cary, NC).
We enrolled 341 HIV-seropositive donors (72 from São Paulo, 110 from Recife, 51 from Belo Horizonte, and 108 from Rio de Janeiro). Detailed results of risk factor interviews for these infected donors relative to a group of matched non–HIV-infected controls are reported elsewhere.10
PCR amplification of HIV pol sequences enabling genotype and resistance analyses was successful in the Brazil Central Laboratory for 302 (88.6%) of the 341 cases; 75% were classified as subtype B. Using the ViroSeq HIV-1 genotyping system v2.0, informative sequence data were obtained for 29 of 39 samples that were initially refractory to amplification in Brazil, resulting in a total of 331 (97%) of the 341 genotype results. Of the 29 samples, only 62% were subtype B, suggesting that the original primers used in Brazil were not as sensitive as Abbott's assay for amplification of pol sequences and genotyping non-B subtypes.
Figure 1 compares the viral load distributions of the 331 cases that were successfully amplified and genotyped with the 10 refractory samples. Of the 10 samples that were not amplifiable by both the Brazil Central Laboratory and Abbott, 7 were below the level of RNA detection in the Abbott RealTime HIV-1 assay and the remaining 3 samples had viral loads of 1.7, 2.8, and 4.3 log10 copies per milliliter.
Table 1 summarizes the HIV-1 subtype results by collection site. Subtype B was predominant (76%), followed by subtype F (15%) and subtype C (5%). Subtype F had a clear predominance in relation to C in Recife and Rio de Janeiro, whereas in Sao Paulo and Belo Horizonte, the proportion of C was similar to F.
Table 2 presents the subtype results, grouped as B or non-B, according to the donor demographics, stage of infection, and major behavioral risk factors. The distribution was similar for most variables, although men overall and men who acknowledged homosexual risk behaviors had a higher proportion of subtype B relative to non-B infections, with P values of 0.06. Recent infection status could be determined for 246 index donation samples from donors who had not received ART, with 43 cases (17.5%) classified as having acquired infection within approximately 6 months before the seropositive donation; the frequency of non-B HIV-1 subtypes was similar for donors with recent and long-standing infections (28% and 24%, respectively). The distributions of HIV viral loads were similar among donors with subtypes B and non-B infections (Fig. 1).
Of the 331 samples successfully amplified and sequenced, 303 had never been exposed to ART; of these, 241 had the index donation plasma sample available for recent infection status. Primary drug resistance could be detected in HIV strains from 36 (11.8%) of the 303 characterized individuals. The majority of the samples (31) were resistant to 1 drug class (17 to nonnucleoside reverse transcriptase inhibitor, 8 to protease inhibitors, and 6 to nucleoside reverse transcriptase inhibitor), 4 samples to 2 drug classes (nonnucleoside reverse transcriptase inhibitor and nucleoside reverse transcriptase inhibitor), and 1 to all 3 drug classes. Supplemental Digital Content (see Table S1, http://links.lww.com/QAI/A408) describes the relevant mutations detected in this population and the impact of these mutations on drug response. Resistance mutations to Efavirenz that is used in most first-line treatment regimens in Brazil were detected in 20 (6.6%) cases.
Table 3 presents the drug-resistant status of HIV-1 sequences according to the demographic and risk factor characteristics of the donors. Drug resistance could not be associated with any risk factors. There was no significant difference in the rates of resistance relative to duration of infection at the time of donation based on incidence assay testing (15.4% for recently acquired vs. 8.9% for long-standing infections).
We analyzed 341 HIV-seropositive blood donors from 4 geographically disbursed blood centers in Brazil. This represents successful recruitment and blood sampling for molecular studies of approximately 50% of all HIV-seropositive donors identified by these blood centers during the eligibility period. Use of residual plasma samples from the index donations to determine genotypes and particularly primary resistance would be the best study design to avoid bias related to donor return; however, Brazilian institutional review boards do not consistently approve studies using such samples because of lack of specific consent for genetic studies. Consequently, our study only generated sequence data on samples from individuals who returned for counseling and consented to participate in the study. No differences were observed in accrual rates by known demographic parameters (not shown), suggesting no ascertainment bias, although this does represent a potential limitation of our study.
The reverse transcriptase and protease gene could be amplified and sequenced in 97% of the samples after supplementing the original PCR assay performed in the Central Laboratory in Brazil with results from the ViroSeq HIV-1 Genotyping System v2.0. Subtype B was found to be the most prevalent in all 4 sites, varying from 73% in Recife to 80% in Sao Paulo.
Subtype F was present in 24% [95% confidence interval (CI): 15.9 to 31.9] of HIV-infected donors in Recife, confirming recent results from a study from Cavalcanti.15 Subtype C infections are mainly detected in the South of Brazil, where they represent 50% of recent infections and are associated with heterosexual transmission.16–19 In the early 1990s, this subtype was not detected in the city of Sao Paulo.20,21 Subsequently, we detected clade C in 3.8% of HIV-seropositive donations from 1998 to 2002 in the same Sao Paulo center that participated in the current study.13 Prevalence of subtype C of less than 5% of successfully genotyped infections was also described by others.22,23 In the current study, the prevalence of subtype C has increased to 9% (95% CI: 2.1 to 15.3), suggesting that the proportion of infections because of this clade may be slowly increasing in the city of Sao Paulo.18 Recently, a similar trend was observed by Brigido et al18 who detected subtype C sequences in 11% of recently diagnosed patient samples from the city of Sao Paulo.
There are few studies that have evaluated rates of transmitted drug resistance in large sample sets of newly diagnosed subjects in Brazil.3,13,24–26 Only 3 of these studies characterized more than 100 samples from different regions of Brazil. The rates of primary drug resistance in these 5 large studies varied from 2.2% to 8.1%.3,13,24–26 Smaller studies, however, have called attention to the possibility of higher rates of primary resistance in Brazil. Sucupira et al28 analyzed 75 drug-naive, HIV-positive individuals in the city of Santos, a port city in southern Brazil, and found 21 (28%) harboring resistant strains, whereas Bermudez-Aza et al27 evaluated 99 ART-naive, homosexual men and found 21% with drug resistance, raising concern that in certain areas or population groups transmission of drug-resistant strains may be rapidly increasing.
Our data show an overall prevalence of primary ART resistance of 12.2% (95% CI: 8.2 to 15.5), a rate considered intermediate by World Health Organization. In Sao Paulo, however, the rate was 19.4% (95% CI: 9.5 to 29.2), significantly higher (P = 0.02) than our previous study that detected a prevalence of 6.3% among blood donors from 1998 to 2002.13
Although we have tried to understand the correlates of drug resistance by performing a detailed questionnaire on risk factors including sexual behavior, no clear associations could be found. Recently, Bermudez-Aza et al27 analyzed homosexual men from 10 cities in Brazil using respondent-driven sampling and found 21% acquired ART-resistant viruses,28 similar to the prevalence found among homosexual men in our study (17.7%). It is possible that this group has a higher rate of resistance, but our sample size was not large enough to detect a significant difference in relation to other sexual orientation groups.
In conclusion, our data show an increasing diversity of HIV clades and a moderate rate of primary resistance in 4 sites in Brazil and in particular an increase in clade diversity and transmitted ART resistance in the city of Sao Paulo. These results further illustrate the important contribution of studies of infected blood donors as sentinel populations for molecular characterization of HIV and other transfusion-transmitted infectious agents.
1. Secretaria de Vigilância em Saúde, Dep de DST AeHV, ed. Boletim Epidemiológico—Aids e DST. Brasilia, Brazil: Ministério da Saúde; 2011.
2. Morgado MG, Sabino EC, Shpaer EG, et al.. V3 region polymorphisms in HIV-1 from Brazil: prevalence of subtype B strains divergent from North American/European prototype and detection of subtype F. AIDS Res Hum Retroviruses. 1994;10:569–576.
3. Brindeiro RM, Diaz RS, Sabino EC, et al.. Brazilian Network for HIV Drug Resistance Surveillance (HIV-BResNet): a survey of chronically infected individuals. AIDS. 2003;17:1063–1069.
4. Teixeira PR, Vitoria MA, Barcarolo J. Antiretroviral treatment in resource-poor settings: the Brazilian experience. AIDS. 2004;18(suppl 3):S5–S7.
5. Marins JR, Jamal LF, Chen SY, et al.. Dramatic improvement in survival among adult Brazilian AIDS patients. AIDS. 2003;17:1675–1682.
6. Sethi AK, Celentano DD, Gange SJ, et al.. High-risk behavior and potential transmission of drug-resistant HIV among injection drug users. J Acquir Immune Defic Syndr. 2004;35:503–510.
7. Wensing AM, Boucher CA. Worldwide transmission of drug-resistant HIV. AIDS Rev. 2003;5:140–155.
8. Tang JW, Pillay D. Transmission of HIV-1 drug resistance. J Clin Virol. 2004;30:1–10.
9. Kleinman S, King MR, Busch MP, et al.. The National Heart, Lung, and Blood Institute retrovirus epidemiology donor studies (Retrovirus Epidemiology Donor Study and Retrovirus Epidemiology Donor Study-II): twenty years of research to advance blood product safety and availability. Transfus Med Rev. 2012;26:281–304; e1–e2.
10. Almeida-Neto C, Goncalez TT, Birch R, et al.. Risk factors for human immunodeficiency virus infection among Brazilian blood donors: a multicenter case-control study using audio computer-assisted structured interviews. Vox Sang. Mar 20, 2013. [Epub ahead of print] Available at: http://onlinelibrary.wiley.com/doi/10.1111/vox.12028/abstract
. Accessed May 20, 2013.
11. Carneiro-Proietti AB, Sabino EC, Sampaio D, et al.. Demographic profile of blood donors at three major Brazilian blood centers: results from the International REDS-II study, 2007 to 2008. Transfusion. 2010;50:918–925.
12. Sabino EC, Goncalez TT, Carneiro-Proietti AB, et al.. Human immunodeficiency virus prevalence, incidence, and residual risk of transmission by transfusions at Retrovirus Epidemiology Donor Study-II blood centers in Brazil. Transfusion. 2012;52:870–879.
13. Barreto CC, Nishyia A, Araujo LV, et al.. Trends in antiretroviral drug resistance and clade distributions among HIV-1–infected blood donors in Sao Paulo, Brazil. J Acquir Immune Defic Syndr. 2006;41:338–341.
14. Bennett DE, Camacho RJ, Otelea D, et al.. Drug resistance mutations for surveillance of transmitted HIV-1 drug-resistance: 2009 update. PLoS One. 2009;4:e4724.
15. Cavalcanti AM, Brito AM, Salustiano DM, et al.. Primary resistance of HIV to antiretrovirals among individuals recently diagnosed at voluntary counselling and testing centres in the metropolitan region of Recife, Pernambuco. Mem Inst Oswaldo Cruz. 2012;107:450–457.
16. Brigido LF, Nunes CC, Oliveira CM, et al.. HIV type 1 subtype C and CB Pol recombinants prevail at the cities with the highest AIDS prevalence rate in Brazil. AIDS Res Hum Retroviruses. 2007;23:1579–1586.
17. Gonzalez AM, Filho GJ, Pestana JO, et al.. Effects of Eurocollins solution as aortic flush for the procurement of human pancreas. Transplantation. 2005;80:1269–1274.
18. Brigido LF, Ferreira JL, Almeida VC, et al.. Southern Brazil HIV type 1 C expansion into the state of Sao Paulo, Brazil. AIDS Res Hum Retroviruses. 2011;27:339–344.
19. Silveira J, Santos AF, Martínez AM, et al.. Heterosexual transmission of human immunodeficiency virus type 1 subtype C in southern Brazil. J Clin Virol. 2012;54:36–41.
20. Sabino EC, Diaz RS, Brigido LF, et al.. Distribution of HIV-1 subtypes seen in an AIDS clinic in Sao Paulo City, Brazil. AIDS. 1996;10:1579–1584.
21. Luppi CG, Eluf-Neto J, Sabino E, et al.. Late diagnosis of HIV infection in women seeking counseling and testing services in Sao Paulo, Brazil. AIDS Patient Care STDS. 2001;15:391–397.
22. Turchi MD, Diaz RS, Martelli CM, et al.. Genetic diversity and HIV-1 incidence estimation among cocaine users in Sao Paulo, Brazil. J Acquir Immune Defic Syndr. 2002;30:527–532.
23. Brigido LF, Franco HM, Custodio RM, et al.. Molecular characteristics of HIV type 1 circulating in Sao Paulo, Brazil. AIDS Res Hum Retroviruses. 2005;21:673–682.
24. Sprinz E, Netto EM, Patelli M, et al.. Primary antiretroviral drug resistance among HIV type 1-infected individuals in Brazil. AIDS Res Hum Retroviruses. 2009;25:861–867.
25. Inocencio LA, Pereira AA, Sucupira MC, et al.. Brazilian network for HIV drug resistance surveillance: a survey of individuals recently diagnosed with HIV. J Int AIDS Soc. 2009;12:20.
26. Gonsalez CR, Alcalde R, Nishiya A, et al.. Drug resistance among chronic HIV-1-infected patients naive for use of anti-retroviral therapy in Sao Paulo city. Virus Res. 2007;129:87–90.
27. Bermudez-Aza EH, Kerr LR, Kendall C, et al.. Antiretroviral drug resistance in a respondent-driven sample of HIV-infected men who have sex with men in Brazil. J Acquir Immune Defic Syndr. 2011;57(suppl 3): S186–S192.
28. Sucupira MC, Caseiro MM, Alves K, et al.. High levels of primary antiretroviral resistance genotypic mutations and B/F recombinants in Santos, Brazil. AIDS Patient Care STDS. 2007;21:116–128.
HIV; genotypes; blood donors; Brazil
Supplemental Digital Content
© 2013 Lippincott Williams & Wilkins, Inc.
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.