bDepartmentofGynecologyandObstetricsofthePotiguarUniversity,Natal,Brazil cDepartmentofBiosciencesoftheFederalUniversityofRioGrandedoNorte,Natal, Brazil
KEYWORDS Acquired immunodeﬁciency syndrome; Highlyactive antiretroviraltherapy; Carcinoma; Incidence
Summary Afterhighlyactiveantiretroviraltherapy(HAART)becamewidespread, severalstudiesdemonstratedchangesintheincidenceofdeﬁningandnon-deﬁning AIDScancersamongHIV/AIDSpatients.Weconductedasystematicreviewof obser-vational studies evaluating the incidence of malignancies before and after the introductionofHAARTinpeoplewithHIV/AIDS.Eligiblestudiesweresearchedup toDecember2012inthefollowing databases:Pubmed, Embase,Scielo, Cancerlit andGoogleScholar.Inthisstudy,wedeterminedthecancerriskratiobycomparing thepre-andpost-HAARTeras.Twenty-one relevantarticleswerefound,involving more than 600,000 peoplewith HIV/AIDS and10,891 newcases of cancers. The riskfor thedevelopmentofanAIDS-deﬁningcancerdecreasedafterthe introduc-tionofHAART:Kaposi’s sarcoma(RR=0.30,95% CI:0.28—0.33)andnon-Hodgkin’s lymphoma (RR=0.52, 95% CI:0.48—0.56), in contrasttoinvasive cervical cancer (RR=1.46, 95% CI:1.09—1.94).Among the non-AIDS-deﬁningcancers, theoverall riskincreasedaftertheintroductionofHAART(RR=2.00,95%CI:1.79—2.23).The
incidenceofAIDS-deﬁningcancersdecreasedandtheincidenceofnon-AIDS-deﬁning cancersincreasedaftertheearlyuseofHAART,probablyduetobettercontrolofviral replication,increasedimmunityandincreasedsurvivalprovidedbynewdrugs. ©2014KingSaudBinAbdulazizUniversityforHealthSciences.PublishedbyElsevier Limited.Allrightsreserved.
Introduction ... 2
Methods ... 3
Dataextraction ... 3
Analcancer ... 6
Hodgkin’slymphoma ... 6
Prostatecancer ... 7
Conclusion... 8 Funding... 8 Competinginterests ... 8 Ethicalapproval ... 8 References... 8
Since1996,whenhighlyactiveantiretroviral ther-apy(HAART)becamewidespreadinNorthAmerica, Europe and Australia, the mortality rate in HIV-infected patients dropped dramatically, mainly becauseofthedecreaseintheincidenceof oppor-tunistic infections. In addition to life expectancy, the therapy also affected the epidemiology of non-AIDS-deﬁning cancers (NADCs) and has an importantimpactontheevolutionofthesetumors
BeforeHAART,cancerswereresponsibleforless than10%ofdeathsamongHIV-infectedpatients. AfterHAART,28%ofdeathsinthispopulationhave been attributed to neoplastic causes , despite the substantial decline in the risk of acquiring AIDS-deﬁning cancers (ADCs), especially Kaposi’s sarcoma (KS) and non-Hodgkin’s lymphoma (NHL)
. It is believedthat this phenomenonoccursas
a function of the increase in the incidence and mortalityofNADCs.However,itisstillnotclear in the literature whether the higher incidence of carcinomas in the post-HAART era is only due to the larger number of new cases of NADCs or if thereareotherfactorsinvolved,suchasthelonger lifeexpectancyaffordedbyHAART.
SomestudiesspeculatethatasHIVpatientslive longer after the introduction of HAART, there is a greater potential to developcancers . Other authors believe that the association between dif-ferentriskfactors,suchasHIV-infectionchronicity and its probable oncogenic role, maybe involved
In fact, despitethe advancesin medicine, can-cerisnowoneoftheleadingcausesofdeathamong patientswho livewith HIV.Thisarticleaimsto provideasystematicreviewtoevaluatetheimpact of HAART on the incidence of deﬁning and non-deﬁningAIDScancers.
Studies meeting the following criteria were included:(1)cohortstudiesofpeoplewithHIV/AIDS thatassessedcancerincidencebeforeandafterthe introduction of HAART, (2) studies including adult subjects, (3) studies published between 2001 and 2012 and (4) studiesincludingdata on the cancer incidenceinperson-years.
Eligible studies were identiﬁed by searching the following databases: Pubmed, Embase, Scielo, Cancerlit and Google Scholar. The studies were identiﬁed by a literature search of the databases usingthefollowing medicalsubject headingterms and/or text words: ‘‘AIDS’’, ‘‘HIV’’, ‘‘cancer’’, ‘‘HAART’’, ‘‘incidence’’ and ‘‘cohort’’. The ref-erence lists of the identiﬁed publications were reviewed for additional pertinentstudies. No lan-guagerestrictionswereimposed.
Three researchers (PHL, PCS and MCMC) searchedfor articles published prior toDecember 2012. After searching the databases, 2392 poten-tially relevant HIV/AIDS papers were identiﬁed, 2333ofwhichwereexcluded:1929afterreviewing the title and 404 after reviewing the abstract. Three studies were included after reviewing the References.
Thus,62papersmetthesearchcriteriaandwere evaluatedusingtheNewcastle-Ottawascale;35 were excluded.TheNewcastle-Ottawascale eval-uates the quality of cohort studies; articles that scorehigherthanﬁveareconsideredtobestudies of ‘‘high methodological quality’’. The scale was independently applied by two researchers (RNOC and MCMC), and any discordance was solved by a thirdresearcher(AKSG).Sixarticleswereexcluded forrepeatedreporting(Fig.1).
Variousstudy characteristics were extracted from theoriginalstudiesandincludedinthesystematic review. The data included the ﬁrst authors’ last names, the year of publication, the country, the studydesign,the periodoffollowup,the number of patients, the total person-years and the num-ber ofcancers (Table1). Cancer incidence among HIV/AIDS patients in the pre- and post-HAART
2392 papers indentified
62papers met search criteria
2333excluded: 1929 after review of title and 404 after review of abstract
41 excluded: 35 low Newcastle-Ottawa and 6
repeated reporting 3 included: searched on
periods were abstracted from each study. Three blind reviewers (RNOC, MCMC and AKSG) used the inclusion criteria to choose eligible articles. Disagreements were solved by means of mutual consensus.
DatawereenteredinReviewManager(RevMan)4.2. Thissoftwareallowstheusertoenterprotocols,to completereviews,includingtext,characteristicsof thestudies,comparisontables,andstudydata,and toperformmeta-analysesofthedataentered.
With the incidence in person-years, we deter-mined the risk ratio for each studied malignancy usingﬁxedand randomeffectsmodelsandtesting for the heterogeneity of effects using the Chi-squaredtestinRevMan4.2.
Wefound 21relevantarticlesinvolving morethan sixhundredthousandHIV/AIDSpatients,who con-tributed more than two million person-years of follow-upandwhowerediagnosedwith10,891new casesofcancers.Thedesignfeaturesofthechosen studiesareindicatedinTable1.
Regarding the ADCs, the beginning of HAART decreased the risk to develop KS (RR=0.30, 95% CI: 0.28—0.33) and NHL (RR=0.52, 95% CI: 0.48—0.56) when compared with the pre-HAART period(Figs.2and3).IncontrasttoKSandNHL,we
Table1 CohortstudydesignfeaturesofcancerriskinHIV/AIDSpatients. Study,year(reference) Country Studytype Periodof
follow-up Numberof patients Total numberof person-years Numberof cancers
Dorruccietal.,2001 Italy Cohort 1981—1998 483 3364 6
Boweretal.,2003 UK RLa 1986—2001 8400 22,694 11
Boweretal.,2004 UK Cohort 1984—2003 8640 40,126 26 Hessoletal.,2004 USA RLa 1994—2001 1554 7909 41
Biggaretal.,2006 USA RLa 1980—2002 317,428 477,368 173
D’Souzaetal.,2008 USA Cohort 1984—2006 6972 100,000 69 Franceschietal.,2008 Switzerland Cohort 1984—2006 12,959 73,412 597 Poleseletal.,2008 Switzerland Cohort 1984—2006 12,959 75,222 429 Engelsetal.,2008 USA Cohort 1991—2002 57,350 186,157 871 Crum-Cianﬂoneetal.,2009 USA Cohort 1984—2006 4498 33,486 446 Cliffordetal.,2009 Switzerland Cohort 1984—2007 14,606 84,611 47 DalMasoetal.,2009 Italy RLa 1986—2004 21,951 101,668 1995
Leewenetal.,2009 Australia Cohort 1982—2004 37,407 311,496 2283 Powlesetal.,2009 UK Cohort 1983—2007 11,112 71,687 156 Buchaczetal.,2010 USA Cohort 1994—2007 8070 3000 16 Crum-Cianﬂone,2010 USA Cohort 1985—2008 4506 37,806 66 Franceschietal.,2010 Switzerland RLa 1985—2006 9429 53,715 1460
Seabergetal.,2010 USA Cohort 1984—2007 6972 77,320 933 Bessonetal.,2011 France RLa 1993—1998 80,000 230,173 900
Kirketal.,2001 Europe Cohort 1994—2000 8471 26,764 222 Franzettietal.,2012 Italy Cohort 1985—2011 5924 50,990 144
found thattheriskofdevelopinginvasive cervical cancer (RR=1.46, 95% CI: 1.09—1.94) increased aftertheintroductionofHAART(Fig.4).
In the articles that included data on non-Hodgkin’s lymphoma, we found thatthe introduc-tionofHAARTdecreasedtheratesofnon-Hodgkin’s lymphoma, especially the risk of primary brain lymphoma (RR=0.24, 95% CI: 0.20—0.30). The risk of acquiring diffuse large B-cell lymphoma and Burkitt’s lymphoma, the two most com-mons systemicNHL subtypes among patients with HIV/AIDS, also decreased signiﬁcantly (RR=0.44,
The overall risk ofdevelopingNADCs increased aftertheintroductionofHAART(RR=2.00,95%CI: 1.79—2.23),asshowninFig.5.However,amongthe cancersstudied,wefoundasigniﬁcantriskincrease in a group of six malignancies: anus (RR=4.28, 95% CI: 3.25—5.64), colorectal (RR=1.91, 95% CI: 1.03—3.52),Hodgkin’slymphoma(RR=1.40,95%CI: 1.16—1.69), liver (RR=3.58, 95% CI: 2.57—4.99), lung (RR=2.19, 95% CI: 1.54—3.11) and prostate (RR=2.57,95%CI:1.49—4.44).
Figure2 IncidenceratesforKaposi’ssarcomainthepre-andpost-HAARTerasandtheriskratiosofincidencerates inthepre-HAARTcomparedwiththepost-HAARTera.
Figure3 Incidenceratesfor non-Hodgkin’slymphomainthe pre-andpost-HAARTandtheriskratiosof incidence ratesinthepre-HAARTcomparedwiththepost-HAARTera.
Figure4 Incidenceratesforinvasivecervicalcancerinthepre-andpost-HAARTerasandtheriskratiosofincidence ratesinthepre-HAARTcomparedwiththepost-HAARTera.
Figure5 Incidenceratesfornon-AIDS-deﬁningcancersinthepre-andpost-HAARTerasandtheriskratiosofincidence ratesinthepre-HAARTcomparedwiththepost-HAARTera.
However, we did not ﬁnd a statistically signiﬁ-cant increase or decrease in the development of the following NADC types after the initiation of HAART:bladder(RR=1.88,95%CI:0.59—6.00),CNS (RR=1.81, 95% CI: 0.86—3.85), kidney (RR=1.79, 95% CI: 0.72—4.44), leukemia (RR=0.84, 95% CI: 0.44—1.60), melanoma (RR=1.08, 95% CI: 0.71—1.65) and testis (RR=0.84, 95% CI: 0.38—1.89).
Several theories attempt to explain the increas-ing importance of NADCs among HIV-infected patients. Although some studies have demon-strated a direct relationship between HAART and increased incidenceofNADCs, itis morelikely that antiretroviral therapy indirectly affects the epidemiology of these cancers because HAART increasesthelifeexpectancyofpeoplelivingwith HIV,thusallowingtheemergenceofthese carcino-mas.
Anotherpossibilityis thatthegenomic instabil-ity caused by zidovudine may contribute to the increased incidence of certain types of NADCs. Theﬁrstnucleosidereversetranscriptaseinhibitors (NRTIs) approved for the therapy of HIV is incorporated into DNA, causes mutations in the hypoxanthine-guanine phosphoribosyl-transferase (HPRT) and thymidine kinase (TK) genes, and induces micronuclei, chromosomal aberrations, sister chromatid exchange, shortened telomeres, andothergenotoxiceffectsinculturedcells.
There is aproven association between anogenital cancers and infection by human papilloma virus (HPV), and we also know thatpatientsliving with HIVhaveahigherincidenceofHPVinfection, espe-ciallyofhigh-riskstrains.Wefoundafourfold increase in the risk for developinganal cancer in thepost-HAARTera(RR:4.28,95%CI:3.25—5.64). Our ﬁnding was similar to the results found by Legarthetal.in2013(RR:77.9,95%CI:36.2—167.7)
. Thisfactcanbe explainedpartlybythefact that the use of HAART does not reduce the inci-denceofanogenitalcancerprecursorlesionsandit doesnotseemabletoeliminateHPVinfection. Furthermore, therecommendations ofperforming screeningforanal cancerand itsprecursorsinthe population withHIVcanleadtoanincreaseinthe
numberofdiagnoses .Recently, amulticenter studyshowedthattheuseofHAARTwasassociated withalowerprevalenceofanalintraepithelial neo-plasia(AIN)andalowerprevalenceofHPVinfection
.However,thisstudyhassomelimitationsthat shouldbe highlighted,includingitscross-sectional design. Additionally, the article does not address theeffectofHAARTonhigh-gradeanal intraepithe-lialneoplasia(HGAIN)speciﬁcally,asthesizeofthe study was limited and only a small proportion of thosewithAINhadHGAIN.
Hodgkin’slymphoma(HL)isacommonmalignancy among patients with AIDS, congenital immunode-ﬁciency syndromes or those under immunosup-pressive drug regimens. This fact suggests that immunosuppressionisdirectlyrelatedtothe patho-genesis of HL; however, the relationship is much more complex and multifactorial . We found an increased risk for developing HL after HAART (RR: 1.40, 95% CI:1.16—1.69) that was similar to the standardized incidence ratio (SIR) found by Clifford et al. in 2005 (pre-HAART SIR=17.3 com-pared to post-HAART SIR=36.2)  and Herida et al. in 2003 (pre-HAART SIR 22.75 compared to post-HAARTSIR=31.66).Inadditiontothe sub-stantialincreaseintheageofthepopulationliving withHIVaftertheintroductionofHAART(from34.2 years to42.8 years), the increased incidence is related to immune reconstitution: it is known that the incidence of HL decreases with severe immunosuppression and increases with moderate immunosuppression . It is also known that HL isstronglyassociatedwithEpstein—Barrvirus(EBV) infectionandthatimmunereconstitutionincreases the stimulation of B cells, thereby increasing the numberoflymphocytesinfectedwithEBV.
Ingeneral,theincidenceoflivercancerisincreased in patients with HIV/AIDS. Some studies suggest that this difference may result from a greater prevalenceofinfectionbyhepatitisB virus(HBV), andespeciallyhepatitisCvirus(HCV).The co-infectionofHIVandHBVorHCVseemstoincrease the risk of cirrhosis, end-stage renal disease and hepatocellular carcinoma. Our statistical analysis foundthattheriskfordevelopinglivercancerafter HAARTincreasedmorethanthreetimescompared tothepre-HAARTrisk(RR:3.58,95%CI:2.57—4.99), similartowhatwasfoundbySahasrabuddheetal. in 2012, (RR=2.50, 95% CI: 1.70—3.70) . The main reason for the higher risk is the increase in
survival of people living with HIV/AIDS after the introduction ofHAART; inaddition,it issuspected thatthe hepatotoxicity ofHAARTmay worsenthe carcinogeniceffectofhepatitisBandC.
Manystudiesindicatethatlungcanceristhemost common NADC in the HIV-infected population, in both the pre-and post-HAART eras. As smokingis the main risk factor for the development oflung cancer, we established the relationship between the increased incidence of this tumor with the highestrateofsmokingintheHIV-infected popula-tion;infact,35—70%ofpeoplelivingwithHIV/AIDS smoke,whileonly20%ofthegeneralpopulationare smokers.However,otherfactorsalsoappearto berelatedtotheincreasedincidenceoflungcancer inthesepatientsbecausetheriskremainshigheven after correcting for smoking status . Although somestudiesdonotshowastatisticallysigniﬁcant increaseintheriskofdevelopinglungcancerafter HAART [10,20], our studyfound a doubling ofthe pre-HAARTerariskinthepost-HAARTera(RR:2.19, 95%CI:1.54—3.11),andthisresultissimilartowhat wasfoundbyPoleseletal.in2010(RR:1.8,95%CI: 1.00—3.2).Theincreasedriskmaybeexplained in part bythe long periodfrom the acquisitionof HIV until the diagnosis for lung cancer (11 years on average in the post-HAART era). Thus, studies inthepre-HAARTera,whenthelifeexpectancyof the HIV-infected population was much lower than thatofthepost-HAARTera,madeno diagnosesof lungcancersimplybecausethepatientsdiedfrom other causes ﬁrst, mainly opportunistic infections andAIDS-deﬁningcancers.
Although we found an increased incidence of prostate cancer in the post-HAART era (RR: 2.57, 95%CI:1.49—4.44),veryfewofthepublished stud-ies included this outcome, so we cannot clearly determine the direction of this neoplasia in HIV-positivepatients.Moreover,amongthefewarticles thataddress prostatecancer, the samples studied are small and thus not able to detect signiﬁcant differences . In 2009, Silberstein et al. 
suggested thatHAART may be a protective factor against prostate cancer. However, in 2008, Pan-tanowitzetal.showedthatinacohortstudy, 82% of men diagnosedwith prostate cancer were receiving HAART. We know that the riskof devel-oping prostate cancer is associated with the age andoriginofthepatients;moreover,thepresence or absence of a screening program substantially
[10,22,23] showed that the incidence of prostate cancer in patients living with HIV/AIDS may be equal,lowerorevenhigherthanthatofthegeneral population, which highlights the need to perform more studies before deﬁnitive conclusions can be reached.
Sincethe advent ofHAART,there has beena dra-matic and continuing decline in the incidence of KS , especially in the late post-HAART period comparedto the earlypost-HAART . Asin the studies of Cianﬂone et al. (RR: 12.32, 95% CI: 0.24—12.44) and van Leewen et al. (RR: 0.21, 95%CI: 0.18—0.24) , we found a reduction in the risk for developing KS after the initiation of HAART(RR:0.30,95%CI:0.28—12.33).Despitethe fact that the restoration of the immune system can lead to a transient increase in KS lesions, a manifestationofimmunereconstitutionsyndrome, HAARTreducestheincidenceofKSbyvarious fac-tors, including the reduction of HIV replication, thusimprovingtheimmuneresponsesagainstHHV8 andthedirectanti-angiogenicactivityofsome pro-tease inhibitors. Recent studies suggest that the incidence of KS in people with HIV is now stabi-lized at a much lower rate, although the rate is still substantially high when compared to that of thegeneralpopulation.
SimilartotheratesofKS,wefoundareductionin theriskfor developingNHL afterthe introduction ofHAART (RR=0.52, 95%CI:0.48—0.56).Although somestudiesshowthatthedeclineintheincidence is stabilizing, others show that the decline con-tinuesevenwhencomparingtheearlypost-HAART periodstothelatepost-HAARTera,suggestingthat in the future, the difference in risk between the generalpopulationandHIV-infectedpopulationwill disappear[12,20].ThereductionintheriskofNHL is directly related to the increased CD4 count; however, for Burkitt’s lymphoma, that is not true becausethistypeofNHLoccurswhentheCD4count isrelativelyhigh(>200cells/mm3).Evenso,we foundareductionintheriskforBurkitt’slymphoma sincetheintroduction ofHAART(RR:0.44,95%CI: 0.27—0.73),aswellasintherisksforprimary lym-phomaofthecentralnervoussystem(RR:0.24,95% CI:0.20—0.30) and diffuse large B-cell lymphoma (RR:0.44,95%CI:0.27—0.73).
Unlike other AIDS-deﬁning malignancies, the inci-dence of invasive cervical cancer does not seem to have diminished after the advent of HAART
. Cohort studies with individual patient data on HAART use have produced conﬂicting results withareducedriskassociatedwithHAARTreported in some but not all studies . We found an increasedriskfordevelopinginvasivecervical can-cer after the introduction of HAART (RR: 1.46, 95% CI: 1.09—1.94), similar to what was found by Gingues and Gill  in 2006. The increased incidence for developing invasive cervical cancer, evenafter theadventofHAART,mayhaveseveral explanations, including the fact that the female populationwithHIVislivinglonger,that antiretrovi-raltherapydoesnotpreventtheappearanceofICC asitwasthoughttopreviously,oreventhatthereis aninadequatescreeningprogramforwomenliving with HIV.Thus,itis unclearhowthe immune reconstitutioninducedbyHAARTaffectsthelesions inducedbyHPV;moreresearchisneededto estab-lishasecureassociation.
When comparingthe data between these stud-ies, some limitations in the source data must be considered. In the included studies, most reports included person-years from the 5 years before the onset of AIDS, but others included only the post-AIDS period, and the duration of the post-AIDS follow-up period varied. This may explain the heterogeneity among the studies included in thissystematicreview.Becauseimmunedeﬁciency variesconsiderablyduringthenaturalhistoryofHIV infection and is also affected by treatment, the degreeofimmunedeﬁciencyprobablyvariedacross theHIV/AIDSstudies.Finally,somestudiesdidnot reportdatacoveringallcancertypes.Nonetheless, therewasnodemonstrablepublicationbias.
Arguably,HAART contributedtotheincreased sur-vivalofpeoplelivingwithHIVbyprovidinggreater controlofviralreplicationandincreasedimmunity. Thissomewhatexplainsthereductionofnewcases ofAIDS-deﬁningcarcinomas suchasKSand NHLin thepost-HAARTerabecausetheincidenceofthese cancersdecreasesin theimmunocompetent popu-lation.
In addition, the increased survival allowed the detectionofmalignanttumors,asNADCs(liver,lung andprostate)weregreatlyundiagnosedbeforethe introduction of HAART because these carcinomas arisinginHIV-infectedindividualswereuncommon inthepre-HAARTera.
Moreover, other factors are associated with the lower or higher incidence of cancers follow-ing the introduction of HAART. Some are already welldeﬁned, suchasthe hepatotoxicityofHAART that may predispose patients to the emergence of hepatic carcinoma and immune reconstitu-tion, which increases the number of lymphocytes infected with EBV. However, it is not completely knownhowthesefactorsincreasethecarcinomasof thelung,prostateandcervixinthepost-HAARTera, and larger studies are needed to elucidate these interactions.
BowerM,PalmieriC,DhillonT.AIDS-relatedmalignancies: changing epidemiology and the impact of highly active antiretroviraltherapy.CurrOpinInfectDis2006;19:14—9. BonnetF,LewdenC,MayT,HeripretL,JouglaE,Bevilacqua
S,etal.Malignancy-relatedcausesofdeathinhuman immu-nodeﬁciency virus-infected patients in the era of highly activeantiretroviraltherapy.Cancer2004;101:317—24. Silverberg MJ, Abrams DI. AIDS-deﬁning and
non-AIDS-deﬁningmalignancies:canceroccurrenceinthe antiretro-viraltherapyera.CurrOpinOncol2007;19:446—51. Crum-Cianﬂone N, Hullsiek KH, Marconi V, Weintrob A,
Ganesan A, Barthel RV, et al. Trends in the incidence of cancers among HIV-infected persons and the impact of antiretroviral therapy: a 20-year cohort study. AIDS 2009;23:41—50.
InternationalCollaborationonHIVandCancer.Highlyactive antiretroviraltherapy and incidence ofcancer inhuman immunodeﬁciencyvirus-infectedadults.JNatlCancerInst 2000;92:1823—30.
OliveiraCobucciRN,SaconatoH,LimaPH,RodriguesHM, PrudencioTL,JuniorJE, etal.Comparativeincidenceof cancerinHIV-AIDSpatientsandtransplantrecipients. Can-cerEpidemiol2012;36:e69—73.
StroupDF,Berlin JA, MortonSC, OlkinI,WilliamsonGD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis OfObservationalStudiesinEpidemiology(MOOSE)group. JAMA2000;283:2008—12.
Lo CK, Mertz D, LoebM. Newcastle-Ottawa Scale: com-paringreviewers’ toauthors’ assessments. BMC MedRes Methodol2014;14:45.
Olivero OA. Mechanisms of genotoxicity of nucleoside reverse transcriptase inhibitors. Environ Mol Mutagen 2007;48(3—4):215—23.
Franzetti M, AdorniF, Parravicini C, Vergani B, Antinori S, Milazzo L, et al. Trends and predictors of non AIDS-deﬁning cancersin men and women with HIV-infection. A single-institution retrospective studybefore and after theintroductionof HAART.JAcquir ImmuneDeﬁcSyndr 2013;62(4):414—20.
LegarthR,HellebergM,KronborgG,LarsenCS,PedersenG, PedersenC,etal.AnalcarcinomainHIV-infectedpatients intheperiod1995—2009:aDanishnationwidecohortstudy. ScandJInfectDis2013;45(6):453—9.
GrulichAE.Cancer:theeffects ofHIVand antiretroviral therapy,and implicationsfor earlyantiretroviraltherapy initiation.CurrOpinHIVAIDS2009;4:183—7.
Van der Snoek EM, van der Ende ME, den Hollander JC, Schutten M, Neumann HA, van Doornum GJ. Use of highlyactiveantiretroviraltherapyisassociatedwithlower prevalence ofanal intraepithelial neoplasticlesions and lowerprevalenceofhumanpapillomavirusinHIV-infected men who have sex with men. Sex Transm Dis 2012;39: 495—500.
Palefsky JM. Antiretroviral therapy and anal cancer: the good, the bad, and the unknown. Sex Transm Dis 2012;39:501—3.
BiggarRJ,JaffeES,GoedertJJ,ChaturvediA,PfeifferR, EngelsEA.Hodgkinlymphomaandimmunodeﬁciencyin per-sonswithHIV/AIDS.Blood2006;108:3786—91.
CliffordGM,PoleselJ,RickenbachM,Dal MasoL,Keiser O, Koﬂer A, et al. Cancer risk in the Swiss HIV Cohort Study:associationswithimmunodeﬁciency,smoking,and highly active antiretroviral therapy. J Natl Cancer Inst 2005;97:425—32.
Herida M, Mary-Krause M, Kaphan R, Cadranel J, Poizot-Martin I,Rabaud C, etal. Incidence of non-AIDS-deﬁning cancers before and during the highly active antiretroviraltherapy erainacohortofhuman immuno-deﬁciency virus-infected patients. J ClinOncol 2003;21: 3447—53.
SahasrabuddheVV,ShielsMS,McGlynnKA,EngelsEA.The risk of hepatocellularcarcinoma among individuals with acquiredimmunodeﬁciencysyndromeintheUnitedStates. Cancer2012;118:6226—33.
Pinzone MR, Fiorica F, Di Rosa M, Malaguarnera G, MalaguarneraL,CacopardoB,etal.Non-AIDS-deﬁning can-cersamongHIV-infectedpeople.EurRevMedPharmacolSci 2012;16:1377—88.
vanLeeuwenMT,VajdicCM,MiddletonMG,McDonaldAM, LawM,KaldorJM,etal.Continuingdeclinesinsomebut notallHIV-associatedcancersinAustraliaafterwidespread useofantiretroviraltherapy.AIDS2009;23:2183—90. Polesel J, Franceschi S, Suligoi B, Crocetti E, Falcini F,
GuzzinatiS, etal. Cancerincidence inpeoplewithAIDS inItaly.IntJCancer2010;127(6):1437—45.
SilbersteinJ,DownsT,LakinC,KaneCJ.HIVandprostate cancer:asystematicreviewoftheliterature.Prostate Can-cerProstaticDis2009;12:6—12.
PantanowitzL,Bohac G,Cooley TP,AboulaﬁaD,Dezube BJ. Human immunodeﬁciency virus-associated prostate cancer:clinicopathologicalﬁndingsandoutcomeina multi-institutionalstudy.BJUInt2008;101:1519—23.
in the era of highly active antiretroviraltherapy. Blood 2001;98:3406—12.
Patel P, Hanson DL, Sullivan PS, Novak RM, Moorman AC, Tong TC, etal. Incidence oftypesof canceramong HIV-infected persons compared withthegeneral popula-tion in the United States, 1992—2003. Ann Intern Med 2008;148:728—36.
GinguesS,GillMJ.Theimpactofhighlyactive antiretro-viral therapy on the incidence and outcomes of AIDS-deﬁningcancersinSouthernAlberta.HIVMed2006;7(6): 369—77.
Dorrucci M, Suligoi B, Serraino D, Tirelli U, Rezza G. Incidence ofinvasive cervicalcancerin acohortof HIV-seropositive womenbefore and aftertheintroductionof highlyactiveantiretroviraltherapy.JAcquirImmuneDeﬁc Syndr2001;26:377—80.
Frisch M, Biggar RJ, GoedertJJ. Human papillomavirus-associated cancers in patients with human immunode-ﬁciency virus infection and acquired immunodeﬁciency syndrome.JNatlCancerInst2000;92:1500—10.
Bower M,PowlesT, NelsonM,ShahP,CoxS,MandeliaS, et al.HIV-relatedlungcancerintheera ofhighlyactive antiretroviraltherapy.AIDS2003;17:371—5.
Bower M, Powles T, Newsom-Davis T, ThirlwellC, Steb-bingJ,MandaliaS,etal.HIV-associatedanalcancer:has highlyactiveantiretroviraltherapyreducedtheincidence or improved the outcome.J Acquir ImmuneDeﬁc Syndr 2004;37:1563—5.
HessolNA,SeabergEC,Preston-MartinS,MassadLS,Sacks HS,SilverS,etal.Cancerriskamongparticipantsinthe women’s interagency HIV study. J Acquir Immune Deﬁc Syndr2004;36:978—85.
D’SouzaG,WileyDJ,LiX,ChmielJS,MargolickJB,Cranston RD,etal.Incidenceandepidemiologyofanalcancerinthe multicenterAIDScohortstudy.JAcquirImmuneDeﬁcSyndr 2008;48:491—9.
FranceschiS,MasoLD,RickenbachM,PoleselJ,HirschelB, CavassiniM,etal.KaposisarcomaincidenceintheSwissHIV CohortStudybeforeandafterhighlyactiveantiretroviral therapy.BrJCancer2008;99:800—4.
PoleselJ,CliffordGM,RickenbachM,DalMasoL,Battegay M,BouchardyC,etal.Non-Hodgkinlymphomaincidencein theSwissHIVCohortStudybeforeandafterhighlyactive antiretroviraltherapy.AIDS2008;22:301—6.
Engels EA, Biggar RJ, Hall HI, Cross H, Crutchﬁeld A, FinchJL,etal.Cancerriskinpeopleinfectedwithhuman immunodeﬁciencyvirusintheUnitedStates.IntJCancer 2008;123:187—94.
CliffordGM,RickenbachM,LiseM,DalMasoL,BattegayM, BohliusJ,etal.HodgkinlymphomaintheSwissHIVCohort Study.Blood2009;113:5737—42.
Buchacz K, Baker RK, PalellaJr FJ, Chmiel JS, Lichten-stein KA, Novak RM, et al. AIDS-deﬁning opportunistic illnessesinUSpatients,1994—2007:a cohortstudy.AIDS 2010;24:1549—59.
Besson C, Goubar A, Gabarre J, Rozenbaum W, Pialoux G, ChateletFP,etal. ChangesinAIDS-relatedlymphoma sincetheeraofhighlyactiveantiretroviraltherapy.Blood 2001;98:2339—44.
Crum-Cianﬂone NF. HIV and the Gastrointestinal Tract. InfectDisClinPract(Baltim,MD)2010;18:283—5. Seaberg EC,WileyD,Martinez-Maza O,ChmielJS,
Kings-leyL, TangY, etal.Cancerincidence inthemulticenter AIDSCohortStudybeforeandduringtheHAARTera:1984 to2007.Cancer2010;116:5507—16.
Powles T, Robinson D, Stebbing J, Shamash J, Nel-son M, Gazzard B, et al. Highly active antiretroviral
therapy and the incidence of non-AIDS-deﬁning cancers in people with HIV infection. J Clin Oncol 2009;27: 884—90.
Franceschi S,LiseM, CliffordGM,RickenbachM,Levi F, MaspoliM,etal.Changingpatternsofcancerincidencein
theearly- andlate-HAARTperiods:theSwissHIV Cohort Study.BrJCancer2010;103:416—22.
DalMasoL,PoleselJ,SerrainoD,LiseM,PiselliP,FalciniF, etal.PatternofcancerriskinpersonswithAIDSinItalyin theHAARTera.BrJCancer2009;100:840—7.