Population-level effect of HSV-2 therapy on the incidence of HIV in sub-saharan Africa

Transcription

1 Additional details are published online only at sti.bmj.om/ontent/vol84/ issuesuppl II 1 London Shool of Hygiene and Tropial Mediine, London, UK; 2 Erasmus MC, University Medial Centre Rotterdam, Rotterdam, The Netherlands; 3 Pasteur Suite, Ealing Hospital, London, UK; 4 Institute of Tropial Mediine, Antwerp, Belgium Correspondene to: Dr R G White, Infetious Disease Epidemiology Unit, Department of Epidemiology and Population Health, London Shool of Hygiene and Tropial Mediine, Keppel Street, London WC1E 7HT, UK; rihard.white@lshtm. a.uk Aepted 30 April 2008 This paper is freely available online under the BMJ Journals unloked sheme, see sti.bmj.om/info/unloked.dtl Population-level effet of HSV-2 therapy on the inidene of HIV in sub-saharan Afria R G White, 1 E E Freeman, 1 K K Orroth, 1 R Bakker, 2 H A Weiss, 1 N O Farrell, 3 A Buvé, 4 R J Hayes, 1 J R Glynn 1 ABSTRACT Bakground: Herpes simplex virus type 2 (HSV-2) infetion inreases aquisition and transmission of HIV, but the results of trials measuring the impat of HSV-2 therapy on HIV genital shedding and HIV aquisition are mixed, and the potential impat of HSV-2 therapy on the inidene of HIV at the population level is unknown. Methods: The effets of episodi and suppressive HSV-2 therapy were simulated using the individual-level model STDSIM fitted to data from Cotonou, Benin (relatively low HIV prevalene) and Kisumu, Kenya (high HIV prevalene). Cliniian- and patient-initiated episodi therapy, started when symptomati, were assumed to redue uler duration. Suppressive therapy, given regardless of symptoms, was also assumed to redue uler frequeny and HSV-2 infetiousness. Results: Cliniian-initiated episodi therapy in the general population had almost no effet on the inidene of HIV. The impat of patient-initiated therapy was higher beause of earlier treatment initiation, but still low (,5%) unless symptom reognition and treatment-seeking behaviour were very high. Suppressive therapy given to female sex workers (FSW) in Kisumu had little effet on population HIV inidene. In Cotonou, suppressive therapy in FSW with high overage and long duration redued population HIV inidene by.20% in the long term. Impat was inreased in both ities by also treating a proportion of their lients. Long-term suppressive therapy with high overage in the general population ould redue HIV inidene by more than 30%. Conlusions: These results show that HSV-2 therapy ould potentially have a population-level impat on the inidene of HIV, espeially in more onentrated epidemis. However, a substantial impat requires high overage and long duration therapy, or very high symptom reognition and treatment-seeking behaviour. HIV and herpes simplex virus type 2 (HSV-2) have a synergisti relationship. HIV affets HSV-2 shedding, uler reurrene rate and uler duration. 1 5 HSV-2, in turn, has a strong impat on HIV transmission and aquisition and probably also affets the natural history of HIV infetion A meta-analysis of longitudinal studies found HSV-2 seropositivity to be assoiated with a risk ratio of HIV aquisition of 2.7 (95% onfidene interval (CI) 1.9 to 3.9) in men and 3.1(95% CI 1.7 to 5.6) in women, 15 and most ross-setional studies have found a orrelation between HSV-2 and HIV viral shedding and/or quantity. Antiviral therapy against HSV-2 ould thus have a population-level impat on the global HIV epidemi in areas with a high HSV-2 prevalene suh as sub-saharan Afria. Two types of therapy exist: episodi and suppressive. Episodi therapy is given to individuals with HSV-2 ulers who notie their symptoms. Cliniian-initiated therapy requires the patient to seek treatment for eah uler episode, while patient-initiated therapy requires patients to selfmediate, reduing treatment delays. Three trials have assessed the effet of liniian-initiated episodi HSV-2 therapy on HIV infetiousness and results are urrently available for two of these. A trial in women in Ghana and the Central Afrian Republi found no signifiant impat on genital HIV RNA or time to uler healing. 19 A trial in men in South Afria found a signifiant redution in uler healing times and a borderline signifiant redution in detetion and quantity of uler HIV-1 shedding at day Suppressive therapy is given to HSV-2 positive individuals regardless of whether they have symptoms. Six published trials have assessed the effet of HSV-2 suppressive therapy on genital HIV viral load in HIV/HSV-2-infeted individuals who had not reeived antiretroviral therapy Five found a signifiant redution in frequeny of genital shedding of HIV RNA and four found a signifiant redution in the quantity of genital shedding of HIV RNA. All four studies that used valaylovir or 800 mg twie daily aylovir found a signifiant impat on frequeny and quantity, whereas only one of those using aylovir 400 mg twie daily found an impat on frequeny and neither had an impat on quantity. The half-life of valaylovir is about double that of aylovir. 27 Two trials have assessed the effet of aylovir 400 mg twie daily on HIV aquisition No impat was observed in either trial, suggesting that this dose may be too low for HIV prevention. The results of a randomised trial of suppressive therapy with aylovir 400 mg twie daily on HIV transmission among disordant ouples are expeted in Mathematial modelling an be used to explore the potential population-level impat of data from individual-level trials. Very few modelling studies have examined the relationship between HSV-2 and HIV, and none have examined the impat of HSV-2 antiviral therapy on the inidene of HIV in sub-saharan Afria where suh interventions are likely to prove most useful. The aim of this paper is to explore the potential population-level impat of episodi and suppressive HSV-2 therapy on the inidene of HIV in one low and one high HIV prevalene ity in sub-saharan Afria. ii12

2 METHODS Data, model and baseline model senarios The mathematial model STDSIM has been fitted to data olleted from four ities in sub-saharan Afria as detailed in our previous publiations The model was fitted to data from the Study Group on Heterogeneity of HIV epidemis in Afrian Cities 35 for the year 1997 and to available data on trends over time. STDSIM is an individual-level stohasti model that simulates the natural history and interations between HIV, HSV-2, syphilis, gonorrhoea, hlamydia and hanroid. It has been desribed in detail elsewhere STDSIM is able to simulate realisti sexual networks and heterogeneity between individuals in sexual behaviour and in the natural history of infetion. It has also previously been used to explore the impat of treatment for sexually transmitted infetions and male irumision on HIV-1 prevention and the heterogeneous spread of HIV-1 in Afria, and the diverging HIV-1 and HIV-2 prevalene trends in West Afria. The model representation of the natural history of sexually transmitted infetions and, importantly, the interation between HIV and other sexually transmitted infetions were parameterised based on the literature where possible and poorly known parameter values have been subjet to sensitivity analysis Full details are shown in Setion S1 of the online supplement. In this paper we present results for the two ities in the study with the lowest (Cotonou, Benin) and highest (Kisumu, Kenya) HIV prevalene. Cotonou represents an HIV epidemi highly onentrated among female sex workers (FSW) and their lients, while Kisumu represents a more generalised HIV epidemi typial of Eastern and Southern Afria. 34 Simulated interventions For episodi and suppressive therapy we proposed a low, medium and high value for eah of the required model parameters. The medium values were ombined in one senario to give the most likely estimate of impat and the low and high parameter values were ombined in two further senarios to give a plausible range for the impat of the intervention. Episodi therapy Cliniian-initiated episodi therapy for HSV-2 was targeted at a proportion of symptomati individuals who were simulated to reognise their uler, seek treatment and benefit from redued uler duration. Full details are shown in Setion S2 of the online supplement. Patient-initiated episodi therapy for reurrent HSV-2 ulers was targeted at a proportion of symptomati individuals who were simulated to reognise their symptoms, self-treat and benefit from redued uler duration. In this senario we simulated liniian-initiated episodi therapy for primary ulers (full details in Setion S2 of the online supplement). For reurrent ulers we assumed the same proportion reognised their symptoms as for lini-based therapy, but assumed a higher proportion of those with reurrent ulers reeived treatment beause of the less frequent need for lini visits, and assumed a greater redution in uler duration beause of the earlier initiation of treatment. Full details are shown in Setion S3 of the online supplement. For both liniian- and patient-initiated episodi therapy, the target group was the general population and the simulated intervention was ongoing from 1 January Suppressive therapy Suppressive HSV-2 therapy was initiated in a proportion of HSV-2 infeted individuals who were simulated to start therapy and stop after a finite period of time. Suppressive therapy was assumed to inrease the interval between ulers, redue the duration of ulers and redue the infetiousness of HSV-2 (but had no diret effet on HIV). We proposed a low, medium and high value for eah of these parameters. Full details are shown in Setion S4 of the online supplement. Suppressive therapy was targeted at a proportion of the simulated population in a risk group. Individuals within this target group who were HSV-2 infeted on 1 January 2008 and individuals who beame HSV-2 infeted after this date were simulated to reeive suppressive therapy for 2 years, 10 years or until death. We assumed that, one individuals stopped suppressive treatment, they did not restart. The simulated target groups and overage were: (1) 25%, 50%, and 75% of FSW; (2) 50% of FSW and 10%, 30% and 50% of male lients; (3) 10%, 30% and 50% of the general population. Outome for episodi and suppressive therapy We alulated the perentage redution in the mean annual HSV-2 and HIV inidene among subjets aged years in the general population over 5 and 20 years in the intervention senarios ompared with the baseline senarios. Sensitivity analysis We assessed the robustness of our findings to key baseline and intervention parameter values known to affet impat. Full details are shown in Setion S5 of the online supplement. RESULTS Baseline senario A good fit of the model simulations to data for demography, sexual behaviour and epidemiology of the two populations was ahieved and has been presented elsewhere. The fit of model-simulated HSV-2 prevalene to data from the two ities is shown in fig 1. The prevalene of HSV-2 in younger age groups was fitted preferentially beause younger individuals aount for a higher proportion of new HIV and HSV-2 infetions. As a result, the prevalene of HSV-2 was underestimated in older women in Cotonou. In both sites the prevalene of HSV-2 in younger men was allowed to be overestimated in favour of fitting the prevalene of HSV-2 in younger women beause higher-risk men may not have been ompletely aptured in the original surveys. 46 The fit of HSV-2 prevalene reflets the general age trends in HSV-2 prevalene and the important differenes between the two sites, suh as the differene in HSV-2 prevalene in women aged years in Cotonou (9%) and Kisumu (39%). 47 Owing to the sensitivity of the simulated prevalene of HIV and the prevalene of shortduration sexually transmitted infetions to the hange in sexual behaviour parameters, these parameters ould not be further altered to improve the fit for HSV-2 without worsening the fit for HIV and the other sexually transmitted infetions. The model provided a reasonable fit to the age and sex patterns in HIV in both sites (fig 1). The model repliated the observed patterns, inluding the muh higher prevalene of HIV among young women ompared with young men. In Cotonou the prevalene of HIV was very similar among men and women, as observed. In Kisumu the prevalene of HIV peaked at a younger age in women than in men. The model fits the observed prevalene of HIV in the general population in 1997 ii13

3 Figure 1 Observed (95% CI) and simulated prevalene of HIV and herpes simplex virus 2 (HSV-2) by age and sex in 1997 and over time in Cotonou and Kisumu (15 49 years). Note the differene in the y-axis sale used on HIV prevalene graphs. Gen pop, general population. ANC, antenatal lini; F, female; M, male. and the available data on the trends in HIV prevalene reasonably well. Interventions The effets of the interventions in terms of perentage redution in HSV-2 and HIV inidene over 5 and 20 years in the two ities are shown in figs 2 and 3. The point estimates result from simulations assuming medium senario options for all parameters, and the plausible bound from assuming all high or all low values. For simpliity, it is assumed that there are no other hanges in interventions and no antiretroviral therapy over the period. Episodi therapy With the medium senario options, liniian-initiated episodi therapy had almost no effet on the inidene of HSV-2 or HIV in the short or long term in either ity (right-hand side of graphs in figs 2 and 3). Patient-initiated therapy was only slightly more effetive. The upper plausible bound for episodi therapy, whih assumes that the highest value of eah parameter is orret, suggests that redutions in inidene are possible, espeially with patient-initiated therapy. The sensitivity analysis of the effet of hanging eah parameter individually is shown in fig S1 in the online supplement. The inrease in the proportion who reognise their symptoms had the largest effet on the impat on both HSV-2 and HIV inidene. Assuming 60% of individuals reognised their ulers (and medium options for the other parameters), over 20 years liniian-initiated therapy redued the inidene of HIV by 5.0% in Cotonou and 2.9% in Kisumu and patient-initiated therapy redued the inidene of HIV by 15.1% and 9.8%, respetively. Suppressive therapy Suppressive therapy in FSW in Kisumu had only a marginal effet on the inidene of HSV-2 and HIV, even with high overage long-term therapy (figs 2 and 3). Treating half the FSW and a smaller proportion of their lients was more effetive, leading to redutions in the inidene of HSV-2 of up to 30% and in the inidene of HIV of more than 10% in the long term. In Cotonou, even giving long-duration suppressive therapy to FSW alone resulted in redutions of.30% in HSV-2 and.20% in HIV over the long term, and these were further inreased by treating lients. Short-duration therapy (2 years) produed only small redutions in inidene in both ities. ii14

4 Figure 2 Impat of herpes simplex virus 2 (HSV-2) therapy on the inidene of HSV-2 over 5 and 20 years in adults aged years. For eah ombination of therapy type, target group, treatment duration and overage, three senarios are shown orresponding to the low (H), medium ( X ) and high (i) parameter sets. See Methods for full details. FSW, female sex workers; Clin init, liniian initiated; Pat init, patient initiated. Long-duration suppressive therapy in the general population an produe greater redutions in inidene over 20 years, with the effet rising steeply with the proportion of the population treated in both ities. With 30 50% of the population treated for 10 years, redutions in the inidene of HIV of 20 30% over 20 years were predited (fig 3). The sensitivity analysis of the effet of hanging eah intervention parameter value individually (see fig S2 in the online supplement) shows that the effet on HSV-2 infetivity has the largest effet on HSV-2 inidene, and that effets on uler duration are more important for HIV inidene. However, the results were fairly insensitive to altering any one parameter alone. Sensitivity analysis of baseline parameter values Figure S3 in the online supplement shows the sensitivity of the results for episodi and suppressive therapy to the assumptions made about o-fator effets in different stages of HSV-2 infetion. As expeted, these hanges had no impat on the inidene of HSV-2 inidene. Assuming a between-uler o-fator effet, while keeping the population attributable fration of HSV-2 on HIV transmission onstant by reduing the o-fator effets during uleration, redued the impat of both episodi and suppressive therapy on the inidene of HIV in both ities. This is beause therapy ats on the uler stages, so if they ontribute less to the total ofator effet, the potential benefit of therapy would be redued. Changing HSV-2 uler o-fators had larger impats on the effets of treatment on the inidene of HIV, affeting both episodi and, partiularly, suppressive therapy. However, ofator effets of this magnitude do not fit well with the observed relative risks for the assoiation of HSV-2 and HIV. DISCUSSION This modelling study shows that effetive treatment of HSV-2 ould theoretially redue the inidene of HIV suffiiently for a substantial publi health impat. However, the impat depends on high overage and long duration of therapy, or very high symptom reognition and treatment-seeking behaviour. Episodi therapy would be easier to introdue and sustain than suppressive therapy as patients would already have presented for treatment and should be more motivated to ontinue. However, liniian-initiated therapy was shown to have very little population-level impat on the inidene of HIV. The impat of patient-initiated therapy was higher, but still low unless symptom reognition and treatment seeking was very high. Although studies have shown that individuals an be taught to reognise ulers, 48 this would be hallenging to ahieve on a large sale. The addition of aylovir to syndromi management guidelines for ountries with a high HSV-2 prevalene 49 is only likely to have a population-level impat on HSV-2 and HIV if symptom reognition, treatment-seeking behaviour and orret syndromi management by providers an be substantially improved. ii15

5 Figure 3 Impat of herpes simplex virus 2 (HSV-2) therapy on the inidene of HIV over 5 and 20 years in adults aged years. For eah ombination of therapy type, target group, treatment duration and overage, three senarios are shown orresponding to the low (H), medium ( X ) and high (i) parameter sets. See Methods for full details. FSW, female sex workers; Clin init, liniian initiated; Pat init, patient initiated. FSW might be easiest to target for suppressive therapy but, in the generalised HIV epidemi in Kisumu, suppressive therapy in FSW alone had almost no effet. There was an effet in the onentrated HIV epidemi in Cotonou, possibly beause the proportion of HIV infetions due to FSW and the population attributable fration of HSV-2 on the inidene of HIV among FSW is higher in onentrated HIV epidemis. Additional treatment of lients inreased the impat in both ities, suggesting that targeting ore groups with high-risk behaviour, suh as lients of FSW, an still be an effetive strategy even in some generalised HIV epidemis, but only with high overage and long duration. It is important to note that the impat of 10 years of suppressive therapy was almost as large as lifelong treatment. This suggests that HSV-2 treatment during the first few years following infetion may avert most of the HIV infetions preventable by HSV-2 therapy. The general population senarios show what ould theoretially be ahieved and underline the importane of HSV-2 in HIV transmission, but long-term treatment of large proportions of the population is unlikely to be feasible. The impat of HSV-2 therapy on the inidene and prevalene of HIV in the model is diretly dependent on the assumed strength of the relationship between HSV-2 and HIV. The effets of treatment were dependent on assumptions about the level of the o-fator effet and the relative importane of o-fators between ulers. Suppressive therapy might also redue any between-uler o-fator effet, although this was not modelled, so the redutions in the effets of therapy in these alternative senarios may have been overestimated. Speifi assumptions about the ation and mehanism of the interventions on HSV-2 were informed by the literature wherever possible but, as many of these interventions are not yet well quantified or are hypothetial, a plausible range of values was explored for unertain parameters. Even with these ranges, it is possible that our assumptions may have been overoptimisti in terms of some parameters suh as the high parameter value for symptom reognition for episodi therapy. We assumed the same symptom reognition rate for both sexes, although it tends to be higher among men than women. 50 We may therefore have underestimated the impat of episodi therapy in men and overestimated it in women. However, these effets will tend to anel out in the results we presented on both sexes for episodi therapy. We have presented the hypothetial redutions in HIV inidene and prevalene that ould be ahieved through HSV- 2 interventions in ertain senarios. The real-world impat of saled-up HSV-2 interventions would be tempered by logistial delivery onstraints, suh as the feasibility of funding and arrying out large-sale interventions in the general publi, or of loating lients of FSW and persuading them of the utility of daily suppressive therapy for HSV-2, and by interations with other HIV prevention and treatment programmes. Both episodi and suppressive therapies have limited potential as realisti HIV interventions. To ahieve the high overage and long duration effets on HSV-2 that are required to have an important impat on the inidene of HIV, an effetive vaine against HSV-2 is needed. ii16

6 Key messages Herpes simplex virus type-2 (HSV-2) infetion inreases aquisition and transmission of HIV. Results from trials of HSV- 2 therapy on HIV aquisition have been disappointing, perhaps beause of insuffiient herpes suppression. The potential impat of HSV-2 therapy on the inidene of HIV at the population level in sub-saharan Afria is unknown. Simulated liniian-initiated episodi therapy in the general population had almost no effet on the inidene of HIV. The impat of patient-initiated therapy was higher beause of earlier treatment initiation, but still low unless symptom reognition and treatment-seeking behaviour were very high. Simulated suppressive therapy in high-risk groups with high overage and long duration in settings with onentrated epidemis redued population HIV inidene in the long term. Long-term suppressive therapy with high overage in the general population ould substantially redue the inidene of HIV. HSV-2 therapy ould potentially have a population-level impat on HIV inidene, espeially in more onentrated HIV epidemis. However, substantial impat requires high overage and long duration therapy, or very high symptom reognition and treatment-seeking behaviour. To ahieve the high overage and long duration effets on HSV-2 that are required to have an important impat on the inidene of HIV, an effetive vaine against HSV-2 is needed. Aknowledgements: The authors thank the partiipants and the Study Group on the Heterogeneity of HIV epidemis in Afrian Cities for ooperation and assistane in arrying out this study, and Craig Cohen and Nuala MGrath for aess to unpublished data. Funding: RGW is funded by the MRC (UK) and the Wellome Trust, and EEF is funded by the Wellome Trust (grant /Z/02/Z) and the Marshall Sholarship Commission for their finanial support. The funders had no involvement in the design, olletion, analysis or interpretation of the data, in writing the report or in the deision to submit. Competing interests: None. REFERENCES 1. Mbopi-Keou FX, Gresenguet G, Mayaud P, et al. Genital herpes simplex virus type 2 shedding is inreased in HIV-infeted women in Afria. AIDS 1999;13: Mbopi-Keou FX, Gresenguet G, Mayaud P, et al. Interations between herpes simplex virus type 2 and human immunodefiieny virus type 1 infetion in Afrian women: opportunities for intervention. J Infet Dis 2000;182: Shaker T, Zeh J, Hu HL, et al. Frequeny of symptomati and asymptomati herpes simplex virus type 2 reativations among human immunodefiieny virusinfeted men. J Infet Dis 1998;178: Corey L, Wald A, Celum CL, et al. The effets of herpes simplex virus-2 on HIV-1 aquisition and transmission: a review of two overlapping epidemis. J Aquir Immune Defi Syndr 2004;35: MFarland W, Gwanzura L, Bassett MT, et al. Prevalene and inidene of herpes simplex virus type 2 infetion among male Zimbabwean fatory workers. J Infet Dis 1999;180: Shaker T, Zeh J, Hu H, et al. Changes in plasma human immunodefiieny virus type 1 RNA assoiated with herpes simplex virus reativation and suppression. J Infet Dis 2002;186: Serwadda D, Gray RH, Sewankambo NK, et al. Human immunodefiieny virus aquisition assoiated with genital uler disease and herpes simplex virus type 2 infetion: a nested ase-ontrol study in Rakai, Uganda. JInfetDis2003;188: Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per oital at in monogamous, heterosexual, HIV-1-disordant ouples in Rakai, Uganda. Lanet 2001;357: Shaker T, Rynarz AJ, Goddard J, et al. Frequent reovery of HIV-1 from genital herpes simplex virus lesions in HIV-1-infeted men. JAMA 1998;280: Mbopi-Keou FX, Legoff J, Gresenguet G, et al. Genital shedding of herpes simplex virus-2 DNA and HIV-1 RNA and proviral DNA in HIV-1- and herpes simplex virus-2- oinfeted Afrian women. J Aquir Immune Defi Syndr 2003;33: MClelland RS, Wang CC, Overbaugh J, et al. Assoiation between ervial shedding of herpes simplex virus and HIV-1. AIDS 2002;16: Kilmarx PH, Limpakarnjanarat K, Mastro TD, et al. HIV-1 seroonversion in a prospetive study of female sex workers in northern Thailand: ontinued high inidene among brothel-based women. AIDS 1998;12: Celum C, Levine R, Weaver M, et al. Genital herpes and human immunodefiieny virus: double trouble. Bull World Health Organ 2004;82: Wald A, Link K. Risk of human immunodefiieny virus infetion in herpes simplex virus type 2-seropositive persons: a meta-analysis. J Infet Dis 2002;185: Freeman EE, Weiss HA, Glynn JR, et al. Herpes simplex virus 2 infetion inreases HIV aquisition in men and women: systemati review and meta-analysis of longitudinal studies. AIDS 2006;20: Cowan FF, Pasoe SJ, Barlow KL, et al. Assoiation of genital shedding of herpes simplex virus type 2 and HIV-1 among sex workers in rural Zimbabwe. AIDS 2006;20: LeGoff J, Weiss HA, Gresenguet G, et al. Cerviovaginal HIV-1 and herpes simplex virus type 2 shedding during genital uler disease episodes. AIDS 2007;21: Celum CL, Robinson NJ, Cohen MS. Potential effet of HIV type 1 antiretroviral and herpes simplex virus type 2 antiviral therapy on transmission and aquisition of HIV type 1 infetion. J Infet Dis 2005;191(Suppl 1):S Mayaud P. Impat of HSV-2 episodi therapy on HIV-1 and HSV-2 genital shedding, and uler healing among women in Ghana and Central Afrian Republi: randomised ontrolled trial (ANRS 1212). International AIDS Conferene; Toronto, Paz Bailey G, Sternberg M, Puren A, et al. Impat of episodi aylovir therapy on genital uler duration and HIV shedding from herpeti ulers among men in South Afria. ISSTDR; Seattle, Zukerman RA, Luhetti A, Whittington WL, et al. Herpes simplex virus (HSV) suppression with valaylovir redues retal and blood plasma HIV-1 levels in HIV-1/ HSV-2-seropositive men: a randomized, double-blind, plaebo-ontrolled rossover trial. J Infet Dis 2007;196: Delany S, Mayaud P, Clayton T, et al. Impat of HSV-2 suppressive therapy on genital and plasma HIV-1 RNA in HIV-1 and HSV-2-seropositive women not taking ART: a randomized, plaebo-ontrolled trial in Johannesburg, South Afria. CROI, Dunne E, Whitehead S, Whitehead S, et al. The effet of suppressive aylovir therapy on HIV erviovaginal shedding in HIV- and HSV-2-infeted women, Chiang Rai, Thailand. CROI, Nagot N, Ouedraogo A, Foulongne V, et al. Redution of HIV-1 RNA levels with therapy to suppress herpes simplex virus. N Engl J Med 2007;356: Tanton C, Watson-Jones D, Rusizoka M, et al. A randomized ontrolled trial in Tanzania to assess the impat of HSV-2 suppressive therapy on genital HIV viral load among HSV-2 and HIV-1 seropositive women. ISSTDR; Seattle, Baeten JM, Strik LB, Luhetti A, et al. Herpes simplex virus suppressive treatment dereases plasma HIV-1 viral load in HSV-2/HIV-1 o-infeted women: a randomized, plaebo-ontrolled, ross-over trial. ISSTDR; Seattle, Perry CM, Faulds D. Valailovir. A review of its antiviral ativity, pharmaokineti properties and therapeuti effiay in herpesvirus infetions. Drugs 1996;52: Watson-Jones DL, Rusizoka M, Weiss H, et al. A randomised ontrolled trial to determine the impat of HSV -2 suppressive therapy on HIV inidene in HSV-2 seropositive women in Tanzania. ISSTDR; Seattle, Celum C, Wald J, Hughes J, et al. Effet of ailovir on HIV-1 aquisition in herpes simplex virus 2 seropositive women and men who have sex with men: a randomised, double-blind, plaebo-ontrolled trial. Lanet 2008;371: Blower S, Ma L. Calulating the ontribution of herpes simplex virus type 2 epidemis to inreasing HIV inidene: treatment impliations. Clin Infet Dis 2004;39(Suppl 5):S Korenromp EL, Bakker R, De Vlas SJ, et al. Can behavior hange explain inreases in the proportion of genital ulers attributable to herpes in sub-saharan Afria? A simulation modeling study. Sex Transm Dis 2002;29: Korenromp EL, Bakker R, Gray R, et al. The effet of HIV, behavioural hange, and STD syndromi management on STD epidemiology in sub-saharan Afria: simulations of Uganda. Sex Transm Infet 2002;78(Suppl 1):i Freeman E, Orroth K, White RG, et al. The proportion of new HIV infetions attributable to HSV-2 inreases over time: simulations of the hanging role of sexually transmitted infetions in sub-saharan Afrian HIV epidemis. Sex Trans Infet 2007;83(Suppl 1):i Orroth KK, Freeman E, Bakker R, et al. Understanding differenes aross the ontrasting epidemis in East and West Afria: results from a simulation model of the Four Cities Study. Sex Transm Infet 2007;83(Suppl 1):i Buve A, Carael M, Hayes RJ, et al. Multientre study on fators determining differenes in rate of spread of HIV in sub-saharan Afria: methods and prevalene of HIV infetion. AIDS 2001;15(Suppl 4):S Orroth K. Investigations of the proportion of HIV infetions attributable to sexually transmitted diseases in Sub-Saharan Afria based on data from the Mwanza and Rakai trials. London: London Shool of Hygiene and Tropial Mediine, Korenromp E, van Vliet C, Bakker R, et al. HIV spread and partnership redution for different patterns of sexual behaviour: a study with the mirosimulation model STDSIM. Math Popul Studies 2000;8: van Vliet C, Holmes KK, Singer B, et al. The effetiveness of HIV prevention strategies under alternative senarios: evaluation with the STDSIM model. In: Bank W, ed. Confronting AIDS: Publi priorities in a global epidemi A World Bank Poliy Researh Report. Oxford: Oxford University Press, 1997: ii17

7 39. Korenromp EL, Van Vliet C, Grosskurth H, et al. Model-based evaluation of singleround mass treatment of sexually transmitted diseases for HIV ontrol in a rural Afrian population. AIDS 2000;14: Shmidt WP, Shim VLM, Aaby P, et al. Behaviour hange and ompetitive exlusion an explain the diverging HIV-1 and HIV-2 prevalene trends in Guinea- Bissau. Epidemiol Infet 2008;136: White RG, Orroth KK, Korenromp EL, et al. Can population differenes explain the ontrasting results of the Mwanza, Rakai, and Masaka HIV/sexually transmitted disease intervention trials?: a modeling study. J Aquir Immune Defi Syndr 2004;37: Korenromp EL, White RG, Orroth KK, et al. Determinants of the impat of sexually transmitted infetion treatment on prevention of HIV infetion: a synthesis of evidene from the Mwanza, Rakai, and Masaka intervention trials. J Infet Dis 2005;191(Suppl 1):S Orroth KK, White RG, Korenromp EL, et al. Empirial observations underestimate the proportion of human immunodefiieny virus infetions attributable to sexually transmitted diseases in the Mwanza and Rakai sexually transmitted disease treatment trials: simulation results. Sex Transm Dis 2006;33: White RG, Orroth KK, Glynn JR, et al. Treating urable sexually transmitted infetions to prevent HIV in Afria: still an effetive ontrol strategy? J Aquir Immune Defi Syndr 2008;47: White RG, Glynn J, Orroth KK, et al. Male irumision for HIV prevention in subsaharan Afria: who, what and when? AIDS 2008;22(in press). 46. Buve A, Lagarde E, Carael M, et al. Interpreting sexual behaviour data: validity issues in the multientre study on fators determining the differential spread of HIV in four Afrian ities. AIDS 2001;15(Suppl 4):S Weiss H, Buvé A,RobinsonN,etal. Theepidemiology of HSV-2 infetion and its assoiation with HIV infetion in four urban Afrian populations. AIDS 2001;15(Suppl 4):S Langenberg A, Benedetti J, Jenkins J, et al. Development of linially reognizable genital lesions among women previously identified as having asymptomati herpes simplex virus type 2 infetion. Ann Intern Med 1989;110: World Health Organization. Guidelines for the management of sexually transmitted infetions. Geneva: World Health Organization, Obasi A, Mosha F, Quigley M, et al. Antibody to HSV-2 as a marker of sexual risk behaviour in rural Tanzania. J Infet Dis 1999;179: Sex Transm Infet: first published as /sti on 17 September Downloaded from on 26 Marh 2019 by guest. Proteted by opyright. ii18