w w w . e l s e v ie r . c o m / l o c a t e / b j i d
The
Brazilian
Journal
of
INFECTIOUS
DISEASES
Original
article
Trends
and
predictors
of
HIV-1
acquired
drug
resistance
in
Minas
Gerais,
Brazil:
2002–2012
Helena
Duani
a,b,c,d,∗,
Agdemir
Waleria
Aleixo
c,
Unaí
Tupinambás
a,baUniversidadeFederaldeMinasGerais(UFMG),FaculdadedeMedicina,DepartamentodeClínicaMédica,BeloHorizonte,MG,Brazil bUniversidadeFederaldeMinasGerais(UFMG),FaculdadedeMedicina,Pós-Graduac¸ãoemCiênciasdaSaúde,InfectologiaeMedicina
Tropical,BeloHorizonte,MG,Brazil
cUniversidadeFederaldeMinasGerais(UFMG),LaboratóriodeImunologiaeBiologiaMolecular(DIP-UFMG),BeloHorizonte,MG,Brazil dUniversidadeFederaldeMinasGerais(UFMG),HospitaldasClínicas,BeloHorizonte,MG,Brazil
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received21July2016 Accepted30November2016 Availableonline23December2016
Keywords: HIV Antiretroviral Resistance Genotyping Epidemiology
a
b
s
t
r
a
c
t
Severalstudiesshowthattheprevalenceofmultidrug-resistantHIV-1virusisdecliningover time.Aretrospectivecohortstudywascarriedouttoevaluatethetrendsofdrugresistance inantiretroviraltreatment-exposedindividualsinastateofamiddle-incomecountry,Minas Gerais,southeastregionofBrazil.Weanalyzed2115HIV-1sequencesfrom2002upto2012, from52citiesofMinasGerais.Thegroupswereanalyzedaccordingtothedefinitions:“IAS –3classmutations”,if≥1drugresistancemutationfromIAS2015list(DRM)waspresent ineachclass;“Nofullysusceptibledrugs”astheabsenceofanyfullysusceptibledrugin Stanfordalgorithm;and“GSS≥2,whenamaximumcalculatedGSS(genotypic suscepti-bilityscore)was≥2or≥3,countingonlydrugsavailableinBrazilandUSAatgivencalendar years.TimetrendsofresistancewereanalyzedbyCochran–Armitagetest.Weobserveda decreaseintherateresistancemutationsforPI,NRTI,“IAS–3classmutations”,and“No fullysusceptibledrugs”overthese11years,from69.2%to20.7%,92.3%to90.2%,46.2%to 22.5%,and12.8%to5.7%,respectively(p<0.05).ResistancetoNNRTIincreasedfrom74.4% to81.6%,mainlybecauseofK103Nmutation.TheGSSscore≥2increasedduringtheyears from35.9%to87.3%(p<0.001).WedemonstratethatresistancetoPIandtothethreemain classessimultaneouslyaredeclining,althoughthenumberofpatientsonofantiretroviral therapyhasdoubledinthelasttenyearsinBrazil(125,000in2002to400,000in2014).Broader resistancetestingandtheavailabilityofmoretherapeuticoptionsmighthaveinfluenced thisdecline.TheincreaseinNNRTIresistancecanlimitthisclassasfirstlinetreatmentin Brazilinthefuture.
©2016SociedadeBrasileiradeInfectologia.PublishedbyElsevierEditoraLtda.Thisisan openaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/ by-nc-nd/4.0/).
∗ Correspondingauthor.
E-mailaddress:hduani@gmail.com(H.Duani).
http://dx.doi.org/10.1016/j.bjid.2016.11.009
1413-8670/©2016SociedadeBrasileiradeInfectologia.PublishedbyElsevierEditoraLtda.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
treatment(ART)becameavailable,drasticchangesoccurred in the management of HIV-infected patients. More than 11 million people were on treatment by the end of 2013 withsubsequentimprovementinpatientsqualityoflifeand survival.1However,thisincreaseinuseofARTalsoincreased
acquiredandtransmitteddrugresistance(TDR).The preva-lenceofacquiredHIV-1drugresistance(ADR)inhigh-income countriesfrom Europeand North America,whereART has beenwidelyavailable,isdecliningovertime.2–5DeLucaetal.
showedthat inWesternEurope countries,three-classdrug resistanceincreasedfrom 0% in1997to apeak of3.0% in 2005,anddeclinedto2%in2008.6AnotherstudyfromPortugal
showedadecreaseintheprevalenceofmultidrugresistance over the last decade, from 6.9% in 2001–2003 to 0.6% in 2009–2012.3However,whenanalyzingNNRTIclass,the
major-ityofthe studiesshowedanincreaseindrugresistance.6–8
There are many studies of ADR from middle-income and resource-limitedcountries;however,longitudinalsystematic dataare scarce.9 Consideringthis scenario,theaimofthis
studywastoanalyzethetrendandidentifythepredictorsof acquiredHIVresistanceinthestateofMinasGerais,southeast regionofBrazil.MinasGeraisisalargestatewith20.6million inhabitantsandmorethan42,000casesofAidsupto2015. Since1996, the Brazilianfederal government provides HIV treatmentfree-of-chargetomorethan400,000patientsanda rationaluseofresourcesisimportanttoensuretheprogram’s sustainability. Moreover, the emergence of viral resistance duringHIV-1infectionmayjeopardizetheBrazilianpolicyfor firstlineandsalvageregimenscurrentlyproposed.Thus,in ordertoassessdrugresistanceprofilesandimprove antiretro-viralstrategiesthesurveillanceofacquireddrugresistanceis substantial.Thisstudycoversawideperiod,fromthemoment thatresistancetestingbecameroutineinclinicalpracticein ourcountryin2002–2012,andcanprovidevaluable informa-tionaboutpredictorsofacquireddrugresistanceinthestate ofMinasGerais.
Material
and
methods
AllHIV-1genotypingtestsofthepublichealthsystemofMinas Geraisstateareperformedinourlaboratory,Laboratóriode ImunologiaeBiologiaMolecularDIPattheSchoolofMedicine ofUFMG.Weconsultedourlaboratorydatabasethatcovers 83clinicslocatedover52citiesinthemainmacroregionsof thestateofMinasGerais.Resistancedatawereretrievedfor 2115(69.4%ofallperformedtests)HIV-1-infectedindividuals (adultsandchildren),whohadundergoneatleastonedrug resistancetestingfrom January2002up toDecember2012. Thegeneticdataresultedfrompopulationsequencingusing anautomatedsequencer(ABIPrism3100,AppliedBiosystems) plusacommerciallyavailableassay(ViroSeqHIV-1 Genotyp-ingSystem,v2.0,Abbott)in2002tothefirsthalfof2008and TrugeneHIV-1genotypingassay(SiemensDiagnostics)from 2008onwards.10Foreachsequence,thecorrespondingpatient
demographics,thelastavailableHIVviralload,andCD4+T-cell
lowlevelofresistance,lowlevelofresistance,intermediate resistance,and highlevelofresistance.Thepredicted drug activity wasscored as1, 0.75,0.5,0.25,and0, respectively. Enfuvirtide and raltegravir scored 0if usedprioror during the genotypingtestand1iftheywere neverused. Maravi-roc,rilpivirineandemtricitabinewerenotconsidered,because theywere notavailableinBrazil untilDecember 2012.The obtainedStanford scorescouldthen becomparedwiththe resultsofacohortstudyinvolvingsevencountriesin West-ernEurope.6TheyalsousedtheRegasubtypingtool.Thelocal
EthicsCommitteeapprovedtheproject.
The mutations analyzed were those included in the 2013 IAS-USA list for nucleoside/tide reverse transcrip-tase inhibitors (NRTI), nonnucleoside reverse transcriptase inhibitors (NNRTI), and major resistance mutations to the protease inhibitors (PIs).13 The mutations were analyzed
accordingtothedefinitions6:“IAS–3classmutations”asat least oneDRM present ineach class;“No fully susceptible drugs”astheabsenceofanyfullysusceptibledrugin Stan-fordalgorithm,and“GSS≥2,whenamaximumcalculated GSS (genotypic susceptibility score)was ≥2, counting only drugsthatwerecommerciallyavailableinBrazilatgiven cal-endaryears.GSScalculationwasdoneconsideringallpossible combinationsof:NRTI±NRTI±NNRTI±PI/r±T20±RAL.For this calculation, at most one agent per drug class was allowedwithinacombination,withtheexceptionofNRTIs, in which unlimitedcombinations were allowed, exceptfor stavudine+zidovudineandemtricitabine+lamivudine.6This
methodofcalculatingGSSandofassigningweights accord-ing to Stanfordalgorithm were followed theway proposed byDeLucaetal.,so thattheresultsofthisstudy couldbe comparedtotheEuropeanstudy,oneofthebiggestpublished dataset.6ThedatawereanalyzedusingSPSSv.15(SPSSInc,
Chicago,IL).Logisticregressionanalysiswasusedtoassess theindependentassociationofthevariablesacrosstime.
Results
Thebaselinedemographiccharacteristics ofparticipantsin the cohortare showninTable1.The2115sequenceswere obtainedfrom 1,887HIV-1-infectedindividualsintreatment failure.Medianageofpatientsatthetimeofgenotypingwas 40.5 years(p25–p7533.4–47.6).Overall,maletofemale ratio was1.6(1311/804).Almost11%ofthepatientshadmorethan onegenotypingtest performed.MedianHIV-RNAviral load was4.3logcopies/mL(p25–p753.9–4.7)andmedianCD4count was226cells/L(p25–p7584–368cells/L).Themedian num-berofprevioustreatmentsexperiencedbyenrolledsubjects was3(p25–p752–3).Theuseofmonoordualtherapywas1.9% and 30.7%,respectively,and63.3%patientsusedunboosted PI. Non-B subtypes were detectedin30.8% of the subjects includedinthestudyandthesecondmostprevalentsubtype wasF. Allsequencesbelongedtopatientswithahistoryof exposuretoNRTI,72.3%toNNRTI,51%toboostedPI,and3.5% tonoveldrugclass(entryinhibitorsandintegraseinhibitors).
b r a z j i n f e c t d i s . 2 0 1 7; 2 1(2) :148–154
Table1–Baselinedemographic,immunovirologicalandARThistoryfeaturesofpatientsintherapeuticfailureinMinasGerais,Brazil:2002–2012.
Characteristic Percent(%)b pb
Allyears 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Analyzedsequences(N) 2115 39 77 171 183 279 120 291 284 252 175 244
Age(years)c 40.5(33.4–47.6)a 35.6 38.6 38.3 40.4 39.6 36.9 39.6 42.6 42.5 41.3 43.2 <0.001
Sexmale(N.%) 1311(62.0) 74.4 75.3 69.0 65.0 63.8 60.8 63.2 57.4 56.3 59.4 58.6 <0.001
NumberofARTregimensbefore genotyping
3(2–4)a 3.0 3.0 2.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 2.0 0.055
Prioruseofmonotherapywith zidovudine(N.%)
40(1.9) 5.1 7.8 5.8 1.6 4.3 5.0 0.3 0.0 0.0 0.0 0.0 <0.001
Prioruseofdualtherapy(N.%) 650(30.7) 51.3 57.1 42.1 42.6 39.4 24.2 27.5 28.2 21.0 25.1 16.4 <0.001
UseofunboostedPI-basedtherapy(N. %)
1339(63.3) 82.1 84.4 73.7 73.2 71.0 71.7 65.3 60.9 57.5 56.0 37.7 <0.001
UseofboostedPI-basedtherapy(N.%) 1079(51.0) 38.5 36.4 36.3 41.5 39.4 45.0 59.5 59.2 58.7 55.4 60.7 <0.001
PrioruseofNNRT(N.%) 1528(72.3) 56.4 49.4 62.0 65.0 59.9 69.2 81.1 77.1 79.0 78.3 82.8 <0.001
PrioruseofT20(N.%) 48(2.3) 0.0 0.0 0.0 0.0 0.4 0.8 7.9 2.5 4.0 1.7 1.2 <0.001
PrioruseofRAL(N.%) 24(1.1) 0.0 0.0 0.0 0.0 0.0 0.0 2.7 1.1 2.0 1.1 2.5 0.014
PrioruseofETR(N.%) 2(0.1) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.8 0.0 0.0 0.140
BaselineCD4+T-cellcount(cells/L)c 226.0(84–368)a 235.0 196.0 220.0 196.0 218.0 259.5 268.0 218.5 210.5 239.0 210.0 0.062
BaselineHIV-1plasmaviralloadc (log10copies/mL) 4.3(3.9–4.7)a 4.7 4.5 4.4 4.4 4.3 4.4 4.2 4.2 4.1 4.0 3.9 <0.001 Subtype(N.%) B 1294(61.2) 64.1 70.1 67.8 71.6 62.7 63.3 55.0 62.0 57.9 57.1 55.3 <0.001 F 463(21.9) 17.9 20.8 22.8 19.1 24.4 20.8 24.1 18.3 19.4 24.6 24.2 0.647 a Values:median(p25–p75). b Cochrane–Armitagetestfortrends. c Pearsoncorrelationfortrends.
80 70 60 50 40 30 20 10 0 2002 2003 2004 2005 2006 2007 PI NRTI NNRTI %
IAS 3 class mutations No fully susceptible drug GSS≥2
2008 2009 2010 2011 2012
Fig.1–TrendsofresistancemutationsamongART-experiencedpatientsinMinasGerais,Brazil:2002–2012(p<0.05forall trends).
UseofanyPI,NNRTI,DRV,ETR,T20,andRALinthelast reg-imenwaspresentin59.9%,36.3%,1.1%,0.1%,1.6%,and0.9% ofthe patients, respectively. Over the years, therewas an increaseofpatientage,offemales,useofNNRTI,non-B sub-types,andadecreaseofviralloadatthetimeofgenotyping. CD4countswerestable.
Drug
resistance
mutations
over
time
Fig.1showstheDRMstrendsovertimetodifferentdrugclass: NRTIassociatedDRMshavedecreasedslightlyfrom92.3%to 90.2%.PImutationswerefoundat69.2%ofpatientsin2002 andat34.8%in2012.NNRTImutationsareincreasing,74.4% in2002to81.6%in2012.“IAS–3classmutations”were46.4% anddecreasedto22.5%in2012.
Thethymidineanalogmutations(TAM):M41L,D67N,K70R, L210W,T215YF,andK219EQdroppedfrom56.4%,41%,20.5%, 41%,69.2%,17.2%to27.9%,19.3%,14.8%,15.6%,38.5%,and 16%,respectively.Asignificantdeclinewasobservedforboth type1TAM(M41L,D67N,L210W,andT215Y)andtype2TAM (K70R,T215F, and K219E/Q) profiles, ranging from 74.4% to 48.4%andfrom71.8%to47.1%,respectively.The69-insertion complexand151complexdisappearedalongtheyears.The K65RandM184Vmutationsincreasedfrom2.6%to7.4%and from69.2%to82%,respectively.AmongNNRTIDRMs,a sig-nificant declinefrom 20.5% to 11.9% was found in Y118IC mutation,andanincreasewasfoundtoK103Nmutation,from 43.6%to48%.ForthePImutations,anincreaseoffrequency
wasfoundforI54ML,L76V,andN88S,withapeakintheyears between2008and2009,subsidingsubsequently.ThePI muta-tions L90MandV82AFTLSshowedasignificant decreasein thisperiod.ThedataareshowninTable2.Themostfrequent mutationsinall years forNRTIwere:M184V(77.4%), M41L (43.5%), and T215Y(39.9%). For NNRTI, G190A, K103N, and V108Iwerepresentin10.7%,42.9%,and11.6%ofsequences, respectively.L90M(24.2%),M46I(20.5%)andV82A(21.1%)were themostcommonDRMstoPI.
Thepercentageofsequencesshowing“Nofullysusceptible drugs”declinedfrom12.8%to5.7%.ThecalculatedGSSofthe salvageregimenaccordingtotheavailabledrugsineachyear increasedfrom35.9%to87.3%,asshowninFig.2.
Themultivariatelogisticregressionanalysistoinvestigate thepredictorsofclassresistancedevelopmentontreatmentis showninTable2.IndependentpredictorsofPIresistancewere genotypingyear ofstudy,gender,CD4count, previousZDV monotherapy,ZDV+ddIdualtherapy,unboostedPI,boosted PI,NNRTI,T20,RAL,andanyPIinthelastregimen.Predictors ofNRTIresistanceweregender,age,CD4count,viralloadat failure,dualZDV+ddItherapy,unboostedandboostedPI,and previousNNRTIuse.NNRTIresistancewasfoundin associa-tionwithsomeindependentpredictors,includinggender,CD4 count,previousunboostedPI,andNNRTIand NNRTI+NRTI useinthelastregimen.Femalegenderwasaprotectivefactor toPI,NRTIandIAS–3classmutations,butnottoNNRTI.The usageofunboostedratherthanboostedPIswasthreetimes morelikelytobeassociatedwithevolutiononPIresistance. Thepredictiveabilityofthelogisticregressionmodelsused
b r a z j i n f e c t d i s . 2 0 1 7; 2 1(2) :148–154
Table2–Predictorsofproteaseinhibitor(PI),nucleosidereverse-transcriptaseinhibitor(NRTI),non-nucleosidereverse-transcriptaseinhibitor(NNRTI),IAS3class resistance,nofullysusceptibledrug,andgenotypicsusceptibilityscore≥2intreatmentfailingHIV-1infectedpatients.MinasGerais,Brazil,2002–2012.
Multivariateanalysis:oddsratio(95%confidenceinterval)
Variable(reference) PI NRTI NNRTI IAS–3class
resistance Nofullysusceptible drug GSS≥2 Year 0.88(0.84–0.92) – – 0.92(0.89–0.96) – 1.24(1.19–1.30) Gender(male) 0.55(0.44–0.69) 0.61(0.44–0.83) 1.37(1.08–1.74) 0.79(0.64–0.97) – 1.61(1.3–1.99) Ageinyears(>50) <13 – 0.51(0.23–1.13) – – – 0.53(0.35–0.82) 13–30 – 0.43(0.23–0.79) – – – – 31–50 – – – – – – CD4count/cells/L(>500) <201 0.5(0.36–0.71) 0.48(0.26–0.9) 1.47(1.05–2.06) – – – 201–500 – 0.79(0.42–1.47) 1.23(0.88–1.71) – – – VLatfailure/copies/mL(>100,000) <10,000 – 1.87(1.14–3.09) – – 0.53(0.34–0.84) – 10,000–50,000 – – – – 0.6(0.39–0.91) – 50,000–100,000 – – – – – – Numberofregimens – – – – – 0.73(0.54–0.99) MonoZDV 2.51(1.01–6.27) – – 2.16(1.08–4.29) 2.83(1.22–6.56) – ZDV+ddI 1.32(1.01–1.73) 1.58(1.02–2.46) – 1.38(1.09–1.75) 1.54(1.08–2.2) 0.52(0.41–0.67)
Otherdualtherapy – – – – – 0.56(0.39–0.82)
UnboostedPI 5.63(4.41–7.2) 1.63(1.16–2.29) 0.72(0.55–0.95) 4.18(3.24–5.4) 2.47(1.6–3.83) 0.64(0.49–0.83) BoostedPI 1.75(1.33–2.31) 0.44(0.31–0.63) – 1.58(1.24–2) 4.4(2.72–7.12) 0.63(0.49–0.81) PrioruseofNNRTI 0.52(0.4–0.68) 0.61(0.4–0.93) 5.57(4.39–7.08) 2.59(2.01–3.33) 1.96(1.33–2.9) 0.53(0.41–0.68)
PrioruseofT20 3.92(1.47–10.43) – – 3.6(1.79–7.2) 3.06(1.61–5.8) 0.14(0.064–0.28)
PrioruseofRAL 3.69(1.05–13.01) – – – – –
NRTI+NNRTIinlast regimen
– – 5.99(4.03–8.92) 0.53(0.4–0.7) 0.44(0.26–0.75) –
AnyPIinthelast regimen
60 50 40 30 20 10 0 2001200220032004200520062007200820092010201120122013 110 100 90 80 60 70 50 40 30 20 10 0 2001200220032004 TDF GSS≥2 (Brazil) GSS≥3 (Brazil) GSS≥2 (USA) GSS≥3 (USA) TDF P ercent, % P ercent, % DRV DRV 200520062007200820092010201120122013
Fig.2–Drugresistanceprofileovertime.Percentageof sequenceswithsalvageregimenwithGSS≥2(blueand gray)andGSS≥3(yellowandorange)accordingtothe availabledrugsinBrazilandUSAeachyear(redline),
p<0.001forallGSStrends.DrugsavailableinBrazil/USA eachyear:1991/1987:ZDV;1993/1991:ddI;1996/1995:3TC, SQV;1996/1996:RTV,IDV;1997/1994:D4T;1998/1996:NVP; 1998/1997:NFV;1999/1998:EFV;1999/1997:DLV;2001/1998: ABC;2002/2000:LPV/r;2003/2001:TDF;–/2003:FTC; 2004/2003:ATV;2005/2003:FPV,T20;2008/2006:DRV; 2009/2007:RAL;2009/2005:TPV;2010/2008:ETR;–/2011: RPV;2013/2007:MVC.
showedthatmutationsofresistancetoNNRTIandPIwere bet-terpredictedbytheconsideredvariables,astheyhadgreater sensitivityandspecificity.
Discussion
This study describes the results of the HIV genotyped sequencesofARTfailingpatientsreceivingcareinthestate ofMinasGerais,Brazilduring11years.Thedemographicdata arerepresentativeofHIV-infectedpatientsinthewhole coun-try,withmedianageof40yearsandapredominanceofmale sex.14Anincreasinginageofthefailingsubjectswasobserved
acrossthe years showingthat patientsattendingHIV clin-icsaregettingolderduetolongersurvivalasresultofmore potentandwelltoleratedfirst-lineregimens.Asamatterof factthe incidence ofHIVinfection ishigher amongyoung peopleinBrazil.15Ofnote,thedecreaseinviralloadatthe
timeofgenotypingcouldbeassociatedtoearlierdetection ofthetherapeuticfailure.16Themultivariateanalysisshowed
ofBrazilaremainlymenwhohavesexwithmen(43%).This subsetofpatientsishistoricallymoreexposedtosuboptimal therapythantheoverallpopulation.17,18Datafromalarge
Ital-ianstudyshowsadifferentpattern,withfemalegenderbeing atahigherrisktodevelopresistancemutationstoPI,NRTI, andNNRTIclasses.4
The prevalence ofF subtype is increasing over time in ourstate.TheHIVepidemicinSouthAmericaandBrazilis mainlycausedbysubtypeB.ThesubtypeCpredominatesin the southofthecountry andneighboringcountries, reach-ing 44%insomecities.19–22 ThesubtypeFpredominatesin
Northeastofthecountry.23Phylogeneticanalysisisnecessary
tounderstandthedynamicsoftheFsubtypeincreaseinMinas Gerais.
Wefoundthatinthis cohortofART-exposedgenotyped patientsinamiddle-incomecountry,thePI,NRTI,andIAS–3 classresistanceisdecliningovertime,whereasNNRTI resis-tanceisincreasing.ThesedatawereconfirmedtoPI,IAS–3 classmutations,andGSS≥2bymultivariateanalysis show-inganindependentassociationofcalendaryearwithreduced probabilityofresistance.National dataconfirmthistrend24
and thisisinlinewiththe NationalHIVTreatment Guide-lineschangesduringtheseyears,whichencouragedtheuse ofNRTI+NNRTI(EFV)asthefirsttreatmentoption.14,25 The NNRTI resistanceincreasewasalsoassociatedtothelower geneticbarrierofthisdrugclass.From2002to2005,themost commonNNRTI mutation,K103N,decreased,but startedto increase after 2005 through to 2012. Our state data differ from theNationaldata: whiletheof3classresistancerate inMinasGeraistendstodecrease,thisrateremainedstable countrywide.16
Data from high-income countries from Europe show similar trends to PI, NRTI, and IAS – 3 class resistance mutations.3,6,26,27 Somehigh-incomecountriesalsoshowed
thattherateofNNRTIresistancetendstoincrease.3,6,27,28For
thePIclass,thisdeclinewasofalmost50%,anditcanbethe reflexofthemorewidespreaduseofboostedPI.
ThereductionofADRatthemomentoftherapeuticfailure canalsobeassociatedtothereductionofTDR,29,30although
inaGermanstudytheproportionofTDRwasstabledespite asignificantADRdeclineovertime.27TheTDRinBrazilhas
beenstudiedbutnotsystematically,sothereareno longitu-dinaldatatocompareTDRandADR.17,23Arecentstudyhas
shownthattheremaybesomeassociationbetweenresistance mutationsthatappearamongpatientsfailingtherapy(ADR) andTDRinpatientswitharecentinfection,suggestingthat theADRcanalsobeusedtomonitorTDR.16
Themainlimitationofthisstudyisitsretrospectivedesign, basedonthesampleofasinglestateofBrazil.Thenational databanksandmedicalformsusedtoextractbaselinedatadid notcontainallrequestedinformationandmissingdatawere common,thuslimitingthesizeofthesample.Adherencewas animportantvariablethatcouldnotbeevaluatedhere.
In summary, we have demonstrated that antiretroviral resistanceisdeclining inMinasGerais,Brazil,althoughthe ARVavailabilityhasmorethantripledinthelast10yearsin
thecountry(125,000in2002to400,000in2014).Broader resis-tancetestingmighthaveinfluencedthisdecline,aswell as resistancetestingguidedtherapyandtheavailabilityofmore therapeuticoptions.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
Acknowledgements
The Projeto G.E.R.A. is team: Ronaldo Batista Melo Junior, LuanaAssis Torres,LuisaCinque de Melo,IsabelaVianello Valle,LigiaIasminePereiradosSantosGualberto,JúliaTeixeira Tupinambás,andMarianaFausterFonsecaAraújodeOliveira.
r
e
f
e
r
e
n
c
e
s
1. WHO.GlobalupdateonthehealthsectorresponsetoHIV; 2014.Availablefromhttp://appswhoint/iris/bitstream/10665/ 128494/1/9789241507585engpdf?ua=12014[accessed 24.12.14].
2. CharestH,Doualla-BellF,CantinR,etal.Asignificant reductioninthefrequencyofHIV-1drugresistanceinQuebec from2001to2011isassociatedwithadecreaseinthe monitoredviralload.PLOSONE.2014;9:e109420.
3. VercauterenJ,TheysK,CarvalhoAP,etal.Thedemiseof multidrug-resistantHIV-1:thenationaltimetrendin Portugal.JAntimicrobChemother.2013;68:911–4.
4. FranzettiM,ViolinM,AntinoriA,etal.Trendsandcorrelates ofHIV-1resistanceamongsubjectsfailinganantiretroviral treatmentoverthe2003–2012decadeinItaly.BMCInfectDis. 2014;14:398.
5. AbrahamAG,LauB,DeeksS,etal.Missingdataonthe estimationoftheprevalenceofaccumulatedhuman immunodeficiencyvirusdrugresistanceinpatientstreated withantiretroviraldrugsinNorthAmerica.AmJEpidemiol. 2011;174:727–35.
6. DeLucaA,DunnD,ZazziM,etal.Decliningprevalenceof HIV-1drugresistanceinantiretroviraltreatment-exposed individualsinWesternEurope.JInfectDis.2013;207:1216–20.
7. vonWylV,YerlyS,BurgisserP,etal.Long-termtrendsofHIV type1drugresistanceprevalenceamongantiretroviral treatment-experiencedpatientsinSwitzerland.ClinInfect Dis.2009;48:979–87.
8. VercauterenJ,WensingAM,vandeVijverDA,etal.
Transmissionofdrug-resistantHIV-1isstabilizinginEurope.J InfectDis.2009;200:1503–8.
9. WHO.SurveillanceofHIVdrugresistanceinadultsreceiving ART.Geneva,Switzerland:HIV/AIDSDo;2014.
10.GrantRM,KuritzkesDR,JohnsonVA,etal.Accuracyofthe TRUGENEHIV-1genotypingkit.JClinMicrobiol.
2003;41:1586–93.
11.AlcantaraLC,CassolS,LibinP,etal.Astandardized frameworkforaccurate,high-throughputgenotypingof recombinantandnon-recombinantviralsequences.Nucleic AcidsRes.2009;37:W634–42.
12.deOliveiraT,DeforcheK,CassolS,etal.Anautomated genotypingsystemforanalysisofHIV-1andothermicrobial sequences.Bioinformatics.2005;21:3797–800.
13.JohnsonVA,CalvezV,GunthardHF,etal.Updateofthedrug resistancemutationsinHIV-1:March2013.TopAntivirMed. 2013;21:6–14.
14.MinistériodaSaúde.In:AidsPNdDe,editor.Protocoloclínico ediretrizesterapêuticasparamanejodainfecc¸ãopelohivem adultos.Brasília,Brasil:MinistériodaSaúde.Secretariade VigilânciaemSaúde;2013.
15.MinistériodaSaúde.In:ViraisDdDAeH,editor.Boletim Epidemiológico–AidseDSTAnoII–no1–atésemana epidemiológica26a.Brasília,DF:MinistériodaSaúde, SecretariadeVigilânciaemSaúde;2013.
16.TilghmanMW,Perez-SantiagoJ,OsorioG,etal.Community HIV-1drugresistanceisassociatedwithtransmitteddrug resistance.HIVMed.2014;15:339–46.
17.TupinambasU,DuaniH,MartinsAV,AleixoAW,GrecoDB. Transmittedhumanimmunodeficiencyvirus-1drug resistanceinacohortofmenwhohavesexwithmeninBelo Horizonte,Brazil–1996–2012.MemInstOswaldoCruz. 2013;108:470–5.
18.GoldsamtLA,ClattsMC,ParkerMM,ColonV,HallackR, MessinaMG.Prevalenceofsexuallyacquiredantiretroviral drugresistanceinacommunitysampleofHIV-positivemen whohavesexwithmeninNewYorkcity.AIDSPatientCare STDs.2011;25:287–93.
19.BrindeiroRM,DiazRS,SabinoEC,etal.BrazilianNetworkfor HIVDrugResistanceSurveillance(HIV-BResNet):asurveyof chronicallyinfectedindividuals.AIDS.2003;17:1063–9.
20.BrennanCA,BritesC,BodelleP,etal.HIV-1strainsidentified inBrazilianblooddonors:significantprevalenceofB/F1 recombinants.AIDSResHumRetroviruses.2007;23:1434–41.
21.MachadoLF,IshakMO,VallinotoAC,etal.Molecular epidemiologyofHIVtype1innorthernBrazil:identification ofsubtypesCandDandtheintroductionofCRF02AGinthe AmazonregionofBrazil.AIDSResHumRetroviruses. 2009;25:961–6.
22.GrafT,PintoAR.TheincreasingprevalenceofHIV-1subtype CinSouthernBrazilanditsdispersionthroughthecontinent. Virology.2013;435:170–8.
23.deMoraesSoaresCM,VergaraTR,BritesC,etal.Prevalenceof transmittedHIV-1antiretroviralresistanceamongpatients initiatingantiretroviraltherapyinBrazil:asurveillancestudy usingdriedbloodspots.JIntAIDSSoc.2014;17:19042.
24.DiazRS,InocencioLA,SucupiraMC,etal.Thevirologicaland immunologicalcharacteristicsoftheHIV-1-infected
populationinBrazil:frominitialdiagnosistoimpactof antiretroviraluse.PLOSONE.2015;10:e0139677.
25.MinistériodaSaúde.In:AidsPNdDe,editor.Recomendac¸ões paraTerapiaAnti-retroviralemAdultosInfectadospeloHIV 2008.7aEdic¸ão.ed.Brasília,Brasil:MinistériodaSaúde. SecretariadeVigilânciaemSaúde;2008.
26.VercauterenJ,DeforcheK,TheysK,etal.Theincidenceof multidrugandfullclassresistanceinHIV-1infectedpatients isdecreasingovertime(2001–2006)inPortugal.Retrovirology. 2008;5:12.
27.SchmidtD,KollanC,FatkenheuerG,etal.Estimatingtrends intheproportionoftransmittedandacquiredHIVdrug resistanceinalongtermobservationalcohortinGermany. PLOSONE.2014;9:e104474.
28.PaquetAC,SolbergOD,NapolitanoLA,etal.Adecadeof HIV-1drugresistanceintheUnitedStates:trendsand characteristicsinalargeprotease/reversetranscriptaseand co-receptortropismdatabasefrom2003to2012.AntivirTher. 2014;19:435–41.
29.YangWL,KouyosR,ScherrerAU,etal.Assessingtheparadox betweentransmittedandacquiredHIVType1drugresistance mutationsintheSwissHIVcohortstudyfrom1998to2012.J InfectDis.2015.
30.PhamQD,WilsonDP,LawMG,KelleherAD,ZhangL.Global burdenoftransmittedHIVdrugresistanceandHIV-exposure categories:asystematicreviewandmeta-analysis.AIDS. 2014;28:2751–62.