Modeling the impact of early antiretroviral therapy for adults coinfected with HIV and hepatitis B or C in South Africa
Martin, Natasha K.a,b; Devine, Angelac; Eaton, Jeffrey W.d; Miners, Alecc; Hallett, Timothy B.d; Foster, Graham R.e; Dore, Gregory J.f; Easterbrook, Philippa J.g; Legood, Rosac; Vickerman, Petera,b
aSchool of Social and Community Medicine, University of Bristol, Bristol
bSocial and Mathematical Epidemiology Group, London School of Hygiene and Tropical Medicine
cDepartment of Health Services Research and Policy, London School of Hygiene and Tropical Medicine
dDepartment of Infectious Disease Epidemiology, Imperial College London
eBlizard Institute of Molecular Medicine, Queen Mary's University of London, London, UK
fKirby Institute, University of New South Wales, Sydney, Australia
gDepartment of HIV/AIDS, World Health Organization, Geneva, Switzerland.
Correspondence to Natasha K. Martin, School of Social and Community Medicine, Canynge Hall, 39 Whatley Road, Bristol, BS8 2PS, UK. Tel: +44 (0) 117 3314570; e-mail: firstname.lastname@example.org
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 Website (http://www.AIDSonline.com).
Objective: There has been discussion about whether individuals coinfected with HIV and hepatitis C virus (HCV) or hepatitis B virus (HBV) (∼30% of all people living with HIV) should be prioritized for early HIV antiretroviral therapy (ART). We assess the relative benefits of providing ART at CD4+ count below 500 cells/μl or immediate ART to HCV/HIV or HBV/HIV-coinfected adults compared with HIV-monoinfected adults. We evaluate individual outcomes (HIV/liver disease progression) and preventive benefits in a generalized HIV epidemic setting.
Methods: We modeled disease progression for HIV-monoinfected, HBV/HIV-coinfected, and HCV/HIV-coinfected adults for differing ART eligibility thresholds (CD4+ <350 cells/μl, CD4+ <500 cells/μl, immediate ART eligibility upon infection). We report disability-adjusted life-years averted per 100 person-years on ART (DALYaverted/100PYonART) as a measure of the health benefits generated from incremental changes in ART eligibility. Sensitivity analyses explored impact on sexual HIV and vertical HIV, HCV, and HBV transmission.
Results: For HBV/HIV-coinfected adults, a switch to ART initiation at CD4+ count below 500 cells/μl from CD4+ below 350 cells/μl generates 9% greater health benefits per year on ART (48 DALYaverted/100PYonART) than for HIV-monoinfected adults (44 DALYaverted/100PYonART). Additionally, ART at CD4+ below 500 cells/μl could prevent 25% and 32% of vertical transmissions of HIV and HBV, respectively. For HCV/HIV-coinfected adults, ART at CD4+ below 500 cells/μl generates 10% fewer health benefits (40 DALYaverted/100PYonART) than for HIV monoinfection, unless ART reduces progression to cirrhosis by more than 70% (33% in base-case).
Conclusions: The additional therapeutic benefits of ART for HBV-related liver disease results in ART generating more health benefits among HBV/HIV-coinfected adults than HIV-monoinfected individuals, whereas less health benefits are generated amongst HCV/HIV coinfection in a generalized HIV epidemic setting.
Worldwide, an estimated 30% of people living with HIV are coinfected with either hepatitis C virus (HCV) or hepatitis B virus (HBV), of which 4–5 million people living with HIV are coinfected with HCV and 2–4 million are coinfected with HBV [1,2]. Adults living with HIV who are coinfected with HBV (HBV/HIV-coinfected) or HCV (HCV/HIV-coinfected) experience 2 to 5-fold increased liver disease progression [3–5] and mortality . However, there remains very limited access to HBV or HCV testing, and treatment in low and middle-income countries (LMICs), so few are diagnosed and receive specific treatment for their hepatitis B or C infection . In light of accelerated liver disease progression for HIV and hepatitis-coinfected persons, as well as potential benefits of antiretroviral therapy (ART) on liver disease progression [4,8–12], there has been debate as to whether HBV/HIV-coinfected or HCV/HIV-coinfected individuals should be initiated on HIV ART at an earlier stage of HIV infection than HIV-monoinfected individuals . However, the degree of benefit of initiating ART earlier in coinfected adults compared with HIV-monoinfected adults has not been quantified, which is necessary to inform optimal use of resources and development of evidence-based clinical guidance.
In the past decade, policy changes and substantial increases in funding have resulted in widespread scale-up of ART for people living with HIV in LMICs. At the end of 2012, approximately 9.7 million people in LMICs were receiving ART – a 32-fold increase since 2002 . In 2010, the WHO guidelines recommended ART initiation for all HIV-infected adults with a CD4+ cell count below 350 cells/μl or with WHO clinical stage 3 or 4 disease . Recent data have shown that ART initiation above CD4+ cell counts of 350 cells/μl in people living with HIV can substantially reduce onward HIV transmission to sexual partners, as well as providing individual therapeutic benefits on morbidity and mortality [16,17]. The 2013 WHO guidelines now recommend ART initiation for people living with HIV (aged more than 5 years) with a CD4+ below 500 cells/μl . However, despite continued ART scale-up, the median CD4+ cell count at ART initiation in many low-income countries remains less than 100 cells/μl as of December 2012 .
In addition to the 2013 recommendation for ART initiation in all adults living with HIV with a CD4+ below 500 cells/μl , there was an additional recommendation in both the 2010 and 2013 guidelines for ART initiation in HBV/HIV-coinfected adults with a CD4+ cell count above 500 cells/μl who also have severe chronic liver disease. For HBV/HIV coinfection, the guidelines recommend the use of an ART regimen containing tenofovir (TDF), with either lamivudine (3TC) or emtricitabine (FTC). This was recommended because of the dual HIV and HBV benefits of the TDF/3TC or FTC regimens, and because the subgroup with severe chronic liver disease have a higher short-term risk of disease progression and mortality . To date, there has been no comparable recommendation for earlier ART initiation regardless of CD4+ cell count in HCV/HIV-coinfected adults despite increasing evidence that ART slows liver disease progression in HCV/HIV coinfection [4,8], even though ART does not contain active anti-HCV agents. Nevertheless, uncertainty remains as to the magnitude of the impact of ART on HCV disease progression. In the deliberations for the current 2013 guidelines based on systematic review data , it was considered that there was insufficient evidence for a benefit of ART at higher CD4+ cell counts to recommend ART initiation regardless of CD4+ cell count.
Our objective was to use a mathematical model to evaluate the relative benefits of ART at CD4+ below 350 cells/μl, ART at CD4+ below 500 cells/μl, and immediate ART regardless of CD4+ cell count among adults with HCV/HIV coinfection or HBV/HIV coinfection compared with HIV monoinfection in South Africa to further inform the development of the 2013 guidelines and recommendations on ART use in HBV and HCV coinfection. South Africa represents a generalized HIV epidemic setting, with an estimated 5.6 million people living with HIV in 2011 . An estimated 1–2% of adults living with HIV are coinfected with HCV in South Africa [20–24], though heterogeneity exists by region . HCV transmission routes are not well established, but include transmission in health settings through poor infection control practices, via blood transfusions and injections . The proportion of adults living with HIV who are HBV surface antigen positive (HBsAg+) in South Africa is higher than for HCV, with reported rates from 5 to 30% [23,27–30]. HBV in South Africa is mainly acquired at birth or during childhood through vertical transmission. We evaluate individual outcomes (HIV and liver disease progression and mortality) and additional preventive benefits on vertical and horizontal transmission.
Mathematical models and parameterization
A deterministic, compartmental model was developed to simulate disease progression of HIV-monoinfected, HBV/HIV-coinfected, and HCV/HIV-coinfected adults with different CD4+ count eligibility thresholds for ART initiation (schematics in Fig. 1). To compare the full potential health benefits of different ART eligibility between cohorts, we assumed all individuals are diagnosed and promptly initiated on to ART when eligible according to the different scenarios. The model simulated a cohort of coinfected adults followed until death. The model was parameterized to South Africa (background and HIV mortality, fertility, and estimated levels of serodiscordancy; Appendix Table A1, http://links.lww.com/QAD/A419). Individuals were assumed to become HIV-infected at age 25 years. Because of lack of data about age of HCV acquisition in generalized HIV epidemics driven by sexual transmission (where HCV transmission routes are unknown but include iatrogenic transmission ), it was assumed individuals are infected with HIV and HCV at a similar time-point. In contrast, individuals in this setting will likely have acquired HBV at birth or early childhood. However, childhood liver disease progression rates are very slow and few will have progressed to cirrhosis by the age of 25 years [31,32]. Accordingly, both coinfection models assumed adults are initially in the mildest stages of liver disease (chronic HBV, mild HCV) and have a high CD4+ cell count (CD4+ >500 cells/μl) at HIV acquisition. We assumed individuals would not have received treatment specifically for their hepatitis B (TDF/3TC or FTC) or hepatitis C (interferon/pegylated interferon and ribavirin) infection. For the sensitivity analysis, the model included HIV transmission to serodiscordant stable sexual partners, and vertical transmission of HIV, HBV, and HCV (see appendix, http://links.lww.com/QAD/A419).
HIV monoinfection model
For HIV monoinfection, individuals progress through HIV disease stages based on CD4+ categories (CD4+ >500cells/μl, 350–500 cells/μl, 200–350 cells/μl, <200cells/μl) to HIV-related death, and also experience background mortality from each stage. All individuals start in the highest CD4+ category; ART was assumed to slow HIV-progression rates by four-fold for each disease stage at base-case [33–37]. We assumed all persons remain on ART after initiation in order to compare the potential full benefits of ART.
Hepatitis C virus/HIV coinfection model
For HCV/HIV coinfection, we simulated coinfection progression among adults with chronic HCV (HCV RNA+), where liver disease stages (mild HCV, moderate HCV, compensated cirrhosis, decompensated cirrhosis, hepatocellular carcinoma) were stratified by HIV stage. Age-stratified HCV monoinfection progression to cirrhosis rates [38,39] were accelerated in HCV/HIV coinfection by 2.5-fold [95% confidence interval (95% CI) 1.8–3.4] . Death rates from decompensated cirrhosis were also elevated in coinfection (by 2.3-fold; 95% CI 1.5–3.4) . ART was assumed to reduce progression to cirrhosis by 33% (from a recent meta-analysis ) in the base-case. For the base-case we assumed HCV does not affect HIV progression  or response to ART [6,41–45], but we explored this further in the sensitivity analysis.
Hepatitis B virus/HIV coinfection model
The HBV/HIV coinfection model simulated progression through liver disease stages [chronic HBV (defined as HBsAg+), compensated cirrhosis, decompensated cirrhosis, hepatocellular carcinoma], each stratified by HIV stage. Given that the impact of HBV on rate of disease progression in immune-compromised individuals is driven by the level of HBV viremia, and TDF/3TC or FTC-containing regimens are highly effective in both HBV e antigen (HBeAg)-positive and negative individuals, we utilized a simplified model of HBV/HIV coinfection progression in which HBV viremia was the dominant influence on disease. We modeled a homogenous population where some transition rates (HBV to compensated cirrhosis, HBV to HCC, HBV to death) were calculated by averaging low/high HBV viremia transition rates weighted by the proportion of the population with low/high viremia. We recognize this ignores subtle differences in outcomes between HBeAg+ and HBeAg− active disease, but given the uncertainty surrounding coinfection transition rates in these two different disease states, we believe our model uses the most appropriate approach with available data.
We assumed coinfection accelerates progression to cirrhosis by 4.6-fold (95% CI 1.5–13.7) for those with high HBV viremia and CD4+ below 200cells/μl [5,46]; for CD4+ above 200 cells/μl progression was equal to monoinfection [46,47]. It was assumed that consistent with 2010 (and 2013) WHO guidelines, HBV/HIV-coinfected individuals would receive ART containing TDF and 3TC or FTC. These regimens were assumed to reduce all liver disease progression rates (by 63–90% depending on disease transition) except from hepatocellular carcinoma to death [9–11]. Although some data indicate TDF may reverse fibrosis [9,12], we conservatively assumed this does not occur. We reduced the impact of TDF on progression to cirrhosis amongst coinfected individuals with CD4+ below 200cells/μl [elevating progression in this group by 2.3-fold (95% CI 1.5–5.3) compared with those with CD4+ >200 cells/μl] . It was assumed individuals would be switched to second-line ART regimens containing anti-HBV treatment if necessary. For the base-case, we assumed no discontinuation of ART, because there is no evidence that discontinuation rates differ by coinfection status, and reported discontinuation rates are negligible [9,49]. We assumed HBV does not impact HIV progression or response to ART [47,50,51].
Individual health impact evaluation
We evaluated the individual impact of two incremental ART eligibility changes for each cohort (first excluding transmission benefits): ART initiation at CD4+ below 500 cells/μl (intervention) compared with ART initiation at CD4+ below 350 cells/μl (baseline); and immediate ART regardless of CD4+ cell count (intervention) compared with ART initiation at CD4+ below 500 cells/μl (baseline). To consider the impact of uncertainty in underlying parameters, we performed a multivariate probabilistic uncertainty analysis, where 1000 parameter sets were randomly sampled from the parameter distributions in Appendix Tables A1 and A2 (http://links.lww.com/QAD/A419). All parameters were sampled except background mortality rates, the impact of ART on HIV disease progression, and the impact of ART on progression to cirrhosis for HCV/HIV-coinfected individuals (a conservative point estimate was used for base-case and increased in the univariate sensitivity analysis). For each of the 1000 parameter sets, the model was run for the baseline and intervention analysis, with a lifetime time horizon. The model tracked life expectancy, health benefits measured in averted disability-adjusted life-years (DALYs), and person-years on ART. Disability weights were taken from the 2010 Global Burden of Disease estimates  (Appendix Table A2, http://links.lww.com/QAD/A419). For coinfected individuals, disability weights were compounded multiplicatively: disability weight = 1 − [(1 − HIV disability weight) × (1 − hepatitis disability weight)]. For all outputs, the median and 95% interval values (taken from outputs generating the 2.5 and 97.5% percentiles) were generated from the distribution in outputs from the multivariate uncertainty sampling runs.
We report DALYs averted per 100 person-years on ART (DALYaverted/100PYonART) as a measure of the incremental health benefits of a given ART provision strategy; both DALYs and person-years of ART were discounted at 3% per year at baseline . We also report the ratio of DALYaverted/100PYonART for HCV/HIV or HBV/HIV coinfection over HIV monoinfection, where a ratio of more than 1 indicates that ART for HCV/HIV or HBV/HIV coinfection generates more health benefits (DALYs averted) per 100 person-years on ART than for those with HIV monoinfection. A linear regression analysis of covariance (ANCOVA) was performed to decompose uncertainty in DALYaverted/100PYonART, into that attributable to each individual parameter. The model was coded and solved in MATLAB (version R2010a).
Sensitivity analyses on individual impact
For each of the 1000 parameter sets, we re-ran the simulations varying a single parameter to explore the sensitivity to: discount rate (0 or 6% per year, compared with 3% at base-case), discontinuation from ART for coinfected individuals only (1 or 3% annually, compared with 0% at base-case), and a 20-year time horizon (lifetime at base-case). We also explored hypothesized associations found in the literature about which there is insufficient evidence for precise parameter estimates: impact of ART on reducing HCV liver disease progression by up to 80% (33% at base-case, but evidence of up to 67% ); impaired ART impact on HIV by 25% for coinfected individuals (three-fold reduction in progression compared with four-fold at base-case) [54–56], or if 5% start in a progressed liver disease stage (moderate HCV or HBV compensated cirrhosis). Finally, we explore the impact if all the HBV/HIV-coinfected individuals have high levels of viremia (33% at base-case).
Analyses of impact on transmission
We explored the impact of ART on sexual transmission of HIV to long-term serodiscordant partners. ART was assumed to reduce heterosexual transmission rates by 96%  at base-case (reduced to 70% in a sensitivity analysis). We assumed transmission rates are equal to monoinfection as no evidence indicates that HIV viral loads are elevated in HBV or HCV coinfection [57,58], or other data suggesting differential sexual transmission risk from coinfected persons.
We also evaluated the impact on mother-to-child vertical transmission of HIV, HCV, and HBV in a cohort of women (full details in appendix, http://links.lww.com/QAD/A419). We assumed all mothers receive initiatives for preventing mother-to-child transmission (PMTCT) which include maternal ART and nevirapine for infants through the end of breastfeeding . Infants born from HBV/HIV-coinfected mothers were assumed to receive HBV vaccination and hepatitis B immunoglobulin (HBIG). The 2010 South African guidelines did not recommend TDF-containing ART for PMTCT in pregnant women , but TDF-PMTCT became available in South Africa in April 2013 , so we performed two analyses exploring the impact if ART for pregnant women does or does not contain TDF. We assumed no vertical transmission of HBV from HBeAg-seroconverted mothers, 23% HBV transmission from mothers with high HBV viremia where PMTCT regimens do not include TDF , and a 0.31 relative risk of transmitting HBV (95% CI 0.15–0.63) with TDF-PMTCT or lifelong ART compared with PMTCT without TDF . HCV vertical transmission from coinfected mothers was assumed to be 1.9-fold (95% CI 1.4–2.7) , higher than in HCV monoinfection. The impact of ART on HCV vertical transmission is unknown; ART is known to increase HCV viral loads [64–66], but one study reported a reduction in transmission on ART , so we assumed no ART impact at base-case.
Life expectancy gains
In South Africa, HBV/HIV or HCV/HIV coinfection marginally reduces life expectancy at the age of 25 years compared with HIV monoinfection both in the absence of ART and under all ART eligibility scenarios (by <3 years per individual; Table 1). A change from ART initiation at CD4+ below 350 cells/μl to CD4+ below 500 cells/μl results in slightly more impact on life expectancy in HBV/HIV coinfection as compared with HIV monoinfection [0.4 years (8%) greater for HBV/HIV coinfection], whereas the impact is 0.7 years less (14% lower) in HCV/HIV coinfection compared with HIV monoinfection.
Impact of antiretroviral therapy on individual benefits
The individual impact of incremental eligibility changes (ART at CD4+ <500 cells/μl versus CD4+ <350 cells/μl, or immediate ART regardless of CD4+ cell count compared with CD4+ <500 cells/μl) is summarized in Table 2 and Fig. 2. For both coinfection cohorts, most deaths (>46%) are due to HIV under all ART eligibility thresholds; expanding ART eligibility progressively averts HIV deaths (by a relative 21% or 15% with ART at CD4+ <500 cells/μl for HCV/HIV or HBV/HIV coinfection, respectively, and a further 11% or 7% with immediate ART, respectively). Because the direct effect of ART on HBV progression is strong, ART at CD4+ below 500 cells/μl or immediate ART for HBV/HIV-coinfected persons decreases HBV-related mortality (by a relative 28% with ART at CD4+ <500 cells/μl and a further 17% with immediate ART). However, ART for HCV/HIV-coinfected individuals increases the share of liver-related deaths (by a relative 34% or 10% with ART at CD4+ <500 cells/μl or immediate ART, respectively) as individuals survive longer and succumb to liver-related mortality because ART impact on liver progression is weaker.
Initiating ART at CD4+ below 500 cells/μl instead of CD4+ below 350 cells/μl averts more lifetime DALYs for HBV/HIV-coinfected individuals than with HIV-monoinfected individuals (5.1 and 4.8 DALYs averted per HBV/HIV-coinfected and HIV-monoinfected adult, respectively; Fig. 2a). Treatment of HBV/HIV coinfection provides 9% greater health benefit in DALYaverted/100PYonART than for HIV monoinfection (48 and 44 DALYaverted/100PYonART for HBV/HIV coinfection and HIV monoinfection, respectively; Fig. 2b). By contrast, initiating ART at CD4+ below 500 cells/μl instead of CD4+ below 350 cells/μl for HCV/HIV-coinfected individuals averts fewer DALYs overall (3.9 per HCV/HIV-coinfected adult) and provides less health benefit per year on ART (40 DALYaverted/100PYonART) than for HIV monoinfection. For each sampled projection, Fig. 2c shows the ratio of the DALYaverted/100PYonART of coinfection over monoinfection; in 94% of the simulations, ART at CD4+ below 500 cells/μl for HBV/HIV coinfection provided more DALYaverted/100PYonART than for HIV monoinfection, whereas ART at CD4+ below 500 cells/μl for HCV/HIV coinfection never provided more DALYaverted/100PYonART. Similar relative efficiencies hold for immediate ART eligibility. Increasing eligibility to immediate ART from CD4+ below 500 cells/μl provides slightly more DALYaverted/100PYonART (by 13–14%) than a switch to ART at CD4+ below 500 cells/μl from CD4+ below 350 cells/μl for all cohorts.
In the ANCOVA analysis for HCV/HIV coinfection, uncertainty in progression rates to cirrhosis and uncertainty in disability weights for ART and compensated cirrhosis contributed most to the variability in DALYaverted/100PYonART, explaining 42% and 48% of the variability, respectively. For HBV/HIV coinfection, variability was primarily due to uncertainty in disability weights for ART and chronic HBV (contributing 72%), the rate of chronic HBV to compensated cirrhosis with CD4+ above 200 cells/μl (16%).
Univariate sensitivity analyses on the health benefit per 100 person-years on ART (DALYaverted/100PYonART) with ART at CD4+ below 500 cells/μl compared with CD4+ below 350 cells/μl indicate that 0% discounting (compared with 3% at base-case) changed the absolute benefits, but not the relative benefits of treating coinfected or monoinfected individuals (Fig. 3a). No relative difference was seen for HBV/HIV coinfection when using 6% discounting, or using a 20-year time horizon. However, ART at CD4+ below 500 cells/μl became slightly more beneficial for HCV/HIV-coinfected individuals than HIV-monoinfected individuals using a 6% discount rate, and was equal using a 20-year time horizon, indicating some benefits for HCV/HIV coinfection were accrued early on. No relative differences were seen, assuming 1% annual discontinuation from ART among coinfected individuals (0% at base-case). However, at a 3% annual discontinuation rate among HBV/HIV-coinfected individuals, ART became slightly less beneficial than for HIV monoinfection (where no discontinuation was assumed). If ART impact on HIV is impaired in coinfection (three-fold increase of HIV survival, base-case four-fold), then ART became less beneficial for both coinfected cohorts, and roughly equal between HBV/HIV coinfection and HIV monoinfection. For HCV/HIV-coinfected individuals, ART at CD4+ below 500 cells/μl (compared with CD4+ <350cells/μl) only becomes more beneficial than for HIV monoinfection if ART reduces the progression to cirrhosis by more than 70% (Fig. 3b). Negligible (<1%) impact change was seen if 5% of the cohort started ART at a more advanced stage of liver disease; if all HBV/HIV-coinfected individuals have high HBV viremia, benefit was marginally increased (by <4%) compared with baseline (not shown).
Impact of antiretroviral therapy on HIV sexual transmission
When including the DALYs averted from prevention of HIV transmission to long-term serodiscordant heterosexual partners, the benefit of a switch to ART at CD4+ below 500 cells/μl from CD4+ below 350 cells/μl is increased by 19–24%, with ART for HBV/HIV-coinfected adults still resulting greater benefit than for HIV monoinfection (Fig. 3a). Relative comparisons hold when further increasing the eligibility from ART at CD4+ below 500 cells/μl to immediate ART. Reducing the impact of ART on HIV sexual transmission (to 70%) lowered impact for all cohorts, but did not change relative impact (not shown).
Impact of antiretroviral therapy on vertical transmission
Antiretroviral therapy at CD4+ below 500 cells/μl or immediate ART for a cohort of women living with HIV is projected to increase the total live births due to increased life expectancy, but also reduces the proportion of infants infected with HIV (2.0%, 1.5%, and 1.3% with CD4+ <350 cells/μl, CD4+ <500 cells/μl, or immediate ART, respectively, in all cohorts; Fig. 3c). Additionally, if ART used for PMTCT does not contain TDF (2010 South African guidelines ), then expanding ART eligibility for HBV/HIV-coinfected mothers reduces the relative proportion of HBV vertical transmissions by 32% with early ART (from 4.4% with ART at CD4+ <350 cells/μl to 3.0% with ART at CD4+ <500 cells/μl), and a further 20% (to 2.4%) with immediate ART (Fig. 3c). However, if ART used for PMTCT contains TDF (available in South Africa from April 2013), no additional impact is achieved with lifelong ART. ART was not assumed to reduce HCV transmission from coinfected mothers, so no impact was seen on HCV vertical transmissions (Fig. 3c, 11.8% infants HCV-infected for all scenarios).
Although other modeling studies have explored the impact of specific HCV antiviral treatment among HCV/HIV coinfection [68–70] and HBV treatment among HBV monoinfection [71–73], to our knowledge, this is the first analysis to compare ART impact on liver disease progression among HBV/HIV or HCV/HIV-coinfected individuals. Our comparative modeling analysis shows that in a generalized HIV epidemic setting such as South Africa, a switch to ART at CD4+ below 500 cells/μl from CD4+ below 350 cells/μl has greater health benefits per year on ART (by 9%) for HBV/HIV-coinfected adults than for HIV-monoinfected individuals, in addition to preventing vertical HBV transmission (when ART for PMTCT does not already contain TDF). These findings support the earlier use of ART in HBV/HIV-coinfected individuals as recommended in the 2010 and 2013 guidelines, although several parameter uncertainties remain. By contrast, early ART for HCV/HIV-coinfected persons has fewer health benefits compared with early ART for HIV monoinfection, unless ART reduces liver disease progression by greater than 70%.
These projections are based on a theoretical model, and subject to a number of limitations. First, there is substantial uncertainty about a number of parameters. Liver disease progression rates were unavailable for coinfected African populations, and may differ from the estimates used from primarily Asian populations. Among HBV/HIV-coinfected individuals, there is uncertainty in outcomes and transition rates for HBeAg-negative and HBeAg-positive disease. We used wide sampling bounds to account for this uncertainty, and our analysis indicated projections were sensitive to cirrhosis progression rates; future research should target this area of uncertainty as a priority. For HCV/HIV coinfection, evidence on ART impact is variable, with studies reporting point estimates of 33–67% reduction in progression to cirrhosis [4,8]. Our results indicate an impact of above 70% could make treatment of HCV/HIV coinfection more beneficial than HIV monoinfection. For HBV/HIV coinfection, the impact of combination TDF/3TC or FTC therapy was taken from small-scale studies, or estimated from monoinfection studies. A small, nonrandomized study has also shown improvement in fibrosis among HBV/HIV-coinfected individuals  which we did not incorporate into our model, including this benefit would increase treatment benefits. It is important to note we did not examine potential harms of drug toxicities due to more prolonged ART exposure, or drug resistance if adherence is sub-optimal, but do not believe this would vary by coinfection status. Additionally, due to lack of data surrounding age of HCV acquisition in generalized HIV epidemics driven by sexual transmission, we assume individuals become infected with HIV and HCV around the same time. As with HBV, progression of liver disease in HCV-infected children occurs very slowly , and few will have progressed to cirrhosis by the age of 25 years. Therefore, if HCV acquisition occurs in childhood prior to HIV acquisition, we do not believe it would significantly change our findings and conclusion. If, however, HCV acquisition occurs much later than HIV acquisition, then individuals may have already progressed to advanced immunodeficiency prior to HCV infection, and would not be initiating ART as a coinfected individual at higher CD4+ cell counts.
Further, evidence is mixed regarding the impact of HCV infection on HIV progression [6,54,75] and CD4+ response to ART, with some studies reporting delayed CD4+ recovery [54–56], whereas others showing little to no impairment [41–45] or suggesting delayed responses could be due to injecting drug use cofactors [76,77]. Our results indicate that impaired recovery would further reduce impact among HCV/HIV-coinfected individuals. Additionally, we assume that ART does not affect HCV vertical transmission, based on a lack of strong data [45,64–66]. However, one study suggests ART may reduce HCV vertical transmission , which would mean that ART could provide more benefit than we have projected.
Second, we modeled prompt initiation of ART as soon as CD4+ eligibility criteria are met, in order to compare the optimal health impact of ART, and because no data suggest linkage to ART is related to coinfection status. Therefore, projections do not include ‘real world’ programmatic issues such as HIV testing and linkage to care, or sub-optimal ART adherence, which would likely make provision of ART less beneficial for all the scenarios examined. Further data are needed to inform more detailed models of ART provision in South Africa.
Third, we do not include costs in this analysis, as uncertainties surround the costs associated with case-finding, linkage-to-care, and treatment of coinfection. Consequently, we were not able to determine cost-effectiveness. Importantly, since there remains limited access to HBV or HCV testing among people living with HIV, changes in the recommended timing of ART among this population could have limited impact without a corresponding scale-up of screening. However, if a program were to prioritize treatment of individuals with coinfection (or prioritize TDF treatments for HBV/HIV-coinfected adults if supply is limited), screening costs could be substantial in generalized HIV epidemics where prevalence is low. However, in concentrated epidemics (such as in epidemics driven by injecting drug use where HCV seroprevalence among adults living with HIV can reach 95% ), screening costs could be much lower.
Finally, our analysis was based on a generalized HIV epidemic setting (South Africa). We neglect any potential additional benefits of ART on preventing transmission of HIV, HCV, or HBV among populations at higher risk such as people who inject drugs or MSM. In particular, in settings driven by injection drug use, being coinfected with HCV is likely to be a marker for individuals with behaviors that make them at high risk for HIV transmission, because the vast majority (>85%) of HIV-infected people who inject drugs are also coinfected with HCV . In these concentrated epidemic settings, ART for HCV/HIV-coinfected individuals could have a substantial impact on the prevention of injection-related HIV transmission [80,81] and so is likely to be more beneficial than ART for lower-risk HIV-monoinfected individuals due to the additional population prevention benefits.
Finally, we note that our analysis is limited to sexual transmission between long-term serodiscordant partners only, and does not consider transmission to casual partners, or partner change. In our analysis, including sexual transmission to serodiscordant partners did not alter the relative benefits of ART, and we would not expect adding casual partners to impact our qualitative results, unless there are differences in sexual behaviour by coinfection status. We do not believe sexual behavior would differ among coinfected or monoinfected individuals in South Africa, but acknowledge it may differ in concentrated epidemics where coinfection is driven by MSM or people who inject drugs.
Our findings underscore the individual and prevention benefits of earlier ART initiation (CD4+ <500 cells/μl) for all adults living with HIV, and especially HBV/HIV-coinfected adults. In generalized HIV epidemics, ART for HCV/HIV-coinfected adults is unlikely to be more beneficial than for HIV monoinfection, but further studies quantifying the impact of ART on HCV-related liver disease progression or HCV vertical transmission would reduce uncertainty surrounding this estimate. The 2013 WHO guidelines now recommend ART initiation for all HIV-infected persons at CD4+ cell counts below 500 cells/μl . Further modeling work should evaluate the cost-effectiveness of HBV screening and immediate ART regardless of CD4+ cell count or stage of liver disease among HBV/HIV-infected adults. For HCV, the advent of a new generation of well tolerated oral regimens that are highly effective against multiple HCV genotypes  will also necessitate further modeling of the combined impact of ART and HCV-specific treatment in HIV/HCV-coinfected individuals for both individual and population benefits .
Conflicts of interest
Funding: This work was funded by the Bill and Melinda Gates Foundation and the WHO through a grant to the HIV Modelling Consortium. N.K.M.: This work is produced under the terms of the postdoctoral research training fellowship issued by The National Institute for Health Research (NIHR). The views expressed in this publication are those of the author and not necessarily those of the UK National Health Service, NIHR or the Department of Health. P.V.: Medical Research Council New Investigator Award G0801627. T.B.H. thanks the Bill & Melinda Gates Foundation, UNAIDS, The Wellcome Trust and the UK MRC for funding support.
N.M. has received an honorarium for speaking at a conference sponsored by Janssen. G.F. has received funding from Roche, Novartis, Janssen, Gilead, BMS, BI, Idenix, and Merck for consultancy and lectures. G.D. is a consultant/advisor and has received research grants from Roche, Merck, Janssen, Gilead, Bristol Myers Squibb. T.B.H. directs the operations of the HIV Modelling Consortium and was, in part, responsible for the funding of this work. P.V., A.D., J.W.E., P.J.E., A.M., and R.L. have no competing interests.
1. Alter MJ. Epidemiology of viral hepatitis and HIV co-infection
. J Hepatol
2. Lacombe K, Rockstroh J. HIV and viral hepatitis coinfections: advances and challenges
3. Graham CS, Baden LR, Yu E, Mrus JM, Carnie J, Heeren T, et al. Influence of human immunodeficiency virus infection on the course of hepatitis C virus infection: a meta-analysis
. Clin Infect Dis
4. Thein H-H, Yi Q, Dore G, Krahn M. Natural history of hepatitis C virus infection in HIV-infected individuals and the impact of HIV in the era of highly active antiretroviral therapy: a meta-analysis
5. Colin JF, Cazals-Hatem D, Loriot MA, Martinot-Peignoux M, Pham BN, Auperin A, et al. Influence of human immunodeficiency virus infection on chronic hepatitis B in homosexual men
6. Chen T-Y, Ding EL, Seage GR, Kim AY. Meta-analysis: increased mortality associated with hepatitis C in HIV-infected persons is unrelated to HIV disease progression
. Clin Infect Dis
8. Limketkai BN, Mehta SH, Sutcliffe CG, Higgins YM, Torbenson MS, Brinkley SC, et al. Relationship of liver disease stage and antiviral therapy with liver-related events and death in adults coinfected with HIV/HCV
. J Am Med Assoc
9. Marcellin P, Gane E, Buti M, Afdhal N, Sievert W, Jacobson IM, et al. Regression of cirrhosis during treatment with tenofovir disoproxil fumarate for chronic hepatitis B: a 5-year open-label follow-up study
10. Hosaka T, Suzuki F, Kobayashi M, Seko Y, Kawamura Y, Sezaki H, et al. Long-term entecavir treatment reduces hepatocellular carcinoma incidence in patients with hepatitis B virus infection
11. Liaw Y, Raptopoulou-Gigi M, Cheinquer H, Sarin SK, Tanwandee T, Leung N, et al. Efficacy and safety of entecavir versus adefovir in chronic hepatitis B patients with hepatic decompensation: a randomized, open-label study
12. Matthews G, Cooper DA, Dore GJ. Improvements in parameters of end-stage liver disease in patients with HIV/HBV-related cirrhosis treated with tenofovir
. Antivir Ther
16. Cohen M, Chen Y, McCauley M, Gamble T, Hosseinipour M, Kumarasamy N, et al. Prevention of HIV-1 infection with early antiretorival therapy
. N Engl J Med
17. Sterne J, May M, Costagliola D, de Wolf F, Phillips A, Harris R, et al. Timing of initiation of antiretroviral therapy in AIDS-free HIV-1-infected patients: a collaborative analysis of 18 HIV cohort studies
20. Gededzha M, Mphahlele M, Lukhwareni A, Selabe S. Should routine serological screening for HCV be mandatory in HIV/AIDS patients enrolling for HAART in South Africa
. South Afr Med J
21. Amin J, Kaye M, Skidmore S, Pillay D, Cooper DA, Dore G. HIV and hepatitis C co-infection within CAESAR study
. HIV Med
22. Cherry CL, Affandi JS, Brew BJ, Creighton J, Djauzi S, Hooker DJ, et al. Hepatitis C seropositivity is not a risk factor for sensory neuropathy among patients with HIV
23. Lodenyo H, Schoub B, Ally R, Kairu S, Segal I. Hepatitis B and C virus infections and liver function in AIDS patients at Chris Hani Baragwanath Hospital, Johannesburg
. E Afr Med J
24. Barth RE, Huijgen Q, Tempelman HA, Mudrikova T, Wensing AMJ, Hoepelman AIM. Presence of occult HBV, but near absence of active HBV and HCV infections in people infected with HIV in rural South Africa
. J Med Virol
25. Parboosing R, Paruk I, Lalloo U. Hepatitis C virus seropositivity in a South African cohort of HIV co-infected, ARV naïve patients is associated with renal insufficiency and increased mortality
. J Med Virol
26. Madhava V, Burgess C, Drucker E. Epidemiology of chronic hepatitis C virus infection in sub-Saharan Africa
. Lancet Infect Dis
27. Lukhwareni A, Burnett RJ, Selabe SG, Mzileni MO, Mphahlele MJ. Increased detection of HBV DNA in HBsAg-positive and HBsAg-negative South African HIV/AIDS patients enrolling for highly active antiretroviral therapy at a Tertiary Hospital
. J Med Virol
28. Burnett RJ, Ngobeni JM, François G, Hoosen AA, Leroux-Roels G, Meheus A, et al. Increased exposure to hepatitis B virus infection in HIV-positive South African antenatal women
. Int J STD AIDS
29. Hoffman C, Charalambous S, Thio CL, Martin D, Pemba L, Fielding K, et al. Hepatotoxicity in an African antiretroviral therapy cohort: the effect of tuberculosis and hepatitis B
30. Firnhaber C, Reyneke A, Schulze D, Malope B, Maskew M, MacPhail P, et al. The prevalence of hepatitis B co-infection in a South African urban government HIV clinic
. S Afr Med J
31. Chang M-H. Chronic hepatitis virus infection in children
. J Gastroenterol Hepatol
32. Hoffmann CJ, Thio CL. Clinical implications of HIV and hepatitis B co-infection in Asia and Africa
. Lancet Infect Dis
33. Johnson LF, Mossong J, Dorrington R, Schomaker M, Hoffman C, Keiser O, et al. Life expectancies of HIV-positive adults receiving antiretroviral treatment in South Africa
. ASSA Convention
34. Johansson K, Robberstad B, Norheim O. Further benefits by early start of HIV treatment in low income countries: survival estimates of early versus deferred antiretroviral therapy
. AIDS Res Ther
35. Mills EJ, Bakanda C, Birungi J, Mwesigwa R, Chan K, Ford N, et al. Mortality by baseline CD4 cell count among HIV patients initiating antiretroviral therapy: evidence from a large cohort in Uganda
36. Brinkhof M, Boulle A, Weigel R, Messou E, Mathers C, Orell C, et al. Mortality of HIV-infected patients starting antiretroviral therapy in sub-Saharan Africa: comparison with HIV-unrelated mortality
. PLoS Med
37. Mahy M, Lewden C, Brinkhof MWG, Dabis F, Tassie J-M, Souteyrand Y, et al. Derivation of parameters used in spectrum for eligibility for antiretroviral therapy and survival on antiretroviral therapy
. Sex Transm Infect
38. Salomon JA, Weinstein MC, Hammitt JK, Goldie SJ. Cost-effectiveness of treatment for chronic hepatitis c infection in an evolving patient population
. J Am Med Assoc
39. Thein H-H, Yi Q, Dore GJ, Krahn MD. Estimation of stage-specific fibrosis progression rates in chronic hepatitis C virus infection: a meta-analysis and meta-regression
40. Pineda JA, Romero-Gómez M, Díaz-García F, Girón-González JA, Montero JL, Torre-Cisneros J, et al. HIV coinfection shortens the survival of patients with hepatitis C virus-related decompensated cirrhosis
41. Klein M, Lalonde R, Suissa S. The impact of hepatitis C virus coinfection on HIV progression before and after highly active antiretoriviral therapy
. J Acquir Immune Defic Syndr
42. Melvin D, Lee J, Belsey E, Arnold J, Murphy R. The impact of coinfection with hepatitis C virus and HIV on the tolerability of antiretroviral therapy
43. Sulkowski M, Moore RD, Mehta SH, Chaisson RE, Thomas DL. Hepatitis C and progression of HIV disease
. J Am Med Assoc
44. Law W, Dauncombe C, Mahanontharit A. Impact of viral hepatitis co-infection on response to antiretroviral therapy and HIV disease progression in the HIV-NAT cohort
45. Chung RT, Evans S, Yang Y, Theodore D, Valdez H, Clark R, et al. Immune recovery is associated with persistent rise in hepatitis C virus RNA, infrequent liver test flares, and is not impaired by hepatitis C virus in co-infected subjects
46. Di Martino V, Thevenot T, Colin JF, Boyer N, Martinot M, Degos F, et al. Influence of HIV infection on the response to interferon therapy and the long-term outcome of chronic hepatitis B
47. Gilson R, Hawkins A, Beecham M, Ross E, Waite J, Briggs M, et al. Interactions between HIV and hepatitis B virus in homosexual men: effects on the natural history of infection
48. Matthews GV, Seaberg EC, Avihingsanon A, Bowden S, Dore GJ, Lewin SR, et al. Patterns and causes of suboptimal response to tenofovir-based therapy in individuals coinfected with HIV and hepatitis B virus
. Clin Infect Dis
49. de Vries-Sluijs TEMS, Reijnders JGP, Hansen BE, Zaaijer HL, Prins JM, Pas SD, et al. Long-term therapy with tenofovir is effective for patients co-infected with human immunodeficiency virus and hepatitis B virus
50. Nikolopoulo GK, Paraskevis D, Hatzitheodorou E, Moschidis Z, Sypsa V, Zavitsanos X, et al. Impact of hepatitis B virus infection on the progression of AIDS and mortality in HIV-infected individuals: a cohort study and meta-analysis
. Clin Infect Dis
51. Konopnicki D, Mocroft A, de Wit S, Antunes F, Ledergerber B, Katlama C, et al. Hepatitis B and HIV: prevalence, AIDS progression, response to highly active antiretroviral therapy and increased mortality in the EuroSIDA cohort
52. Salomon JA, Vos T, Hogan DR, Gagnon M, Naghavi M, Mokdad A, et al. Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010
54. Greub G, Ledergerber B, Battegay M, Grob P, Perrin L, Furrer H, et al. Clinical progression, survival, and immune recovery during antiretroviral therapy in patients with HIV-1 and hepatitis C virus coinfection: the Swiss HIV Cohort Study
55. Braitstein P, Zala C, Yip B, Brinkhof MWG, Moore D, Hogg RS, et al. Immunologic response to antiretroviral therapy in hepatitis C virus-coinfected adults in a population-based HIV/AIDS treatment program
. J Infect Dis
56. De Luca A, Bugarini R, Lepri A, Puoti M, Girardi E, Antinori A. Coinfection with hepatitis viruses and outcome of initial antiretroviral regimens in previously naive HIV-infected subjects
. Arch Intern Med
57. Sheng W-H, Chen M-Y, Hsieh S-M, Hsiao C-F, Wang J-T, Hung C-C, et al. Impact of chronic hepatitis B virus (HBV) infection on outcomes of patients infected with HIV in an area where HBV infection is hyperendemic
. Clin Infect Dis
58. Hershow RC, O’Driscoll PT, Handelsman E, Pitt J, Hillyer G, Serchuck L, et al. Hepatitis C virus coinfection and HIV load, CD4+ cell percentage, and clinical progression to AIDS or death among HIV-infected women: Women and Infants Transmission Study
. Clin Infect Dis
59. National Department of Health South Africa and South African National AIDS Council. PMTCT (Prevention of Mother-to-Child Transmission) clinical guidelines. Republic of South Africa Department of Health; 2010. http://www.fidssa.co.za/images/PMTCT_Guidelines.pdf
61. Lee C, Gong Y, Brok J, Boxall EH, Gluud C. Effect of hepatitis B immunisation in newborn infants of mothers positive for hepatitis B surface antigen: systematic review and meta-analysis
. Br Med J
62. Shi Z, Yang Y, Ma L, Li X, Schreiber A. Lamivudine in late pregnancy to interrupt in utero transmission of hepatitis B virus: a systematic review and meta-analysis
. Obstet Gynecol
63. Polis C, Shah S, Johnson K, Gupta A. Impact of maternal HIV coinfection on the vertical transmission of hepatitis C virus: a meta-analysis
. Clin Infect Dis
64. Bower WA, Culver DH, Castor D, Wu Y, James VN, Zheng H, et al. Changes in hepatitis C virus (HCV) viral load and interferon-[alpha] levels in HIV/HCV-coinfected patients treated with highly active antiretroviral therapy
65. Liu X, He N, Fu Z, Duan S, Gao M, Zhang Z. Plasma hepatitis C virus viral load among hepatitis C virus mono-infected and HCV/HIV co-infected individuals in Yunnan Province, China
. Hepat Mon
66. Ragni M, Bontempo F. Increase in hepatitis C virus load in hemophiliacs during treatment with highly active antiretroviral therapy
. J Infect Dis
67. European Paediatric Hepatitis C Virus NetworkA significant sex – but not elective cesarean section – effect on mother-to-child transmission of hepatitis C virus infection
. J Infect Dis
68. Campos N, Salomon J, Servoss J, Nunes D, Samet J, Freedberg K, et al. Cost-effectiveness of treatment for hepatitis C in an urban cohort co-infected with HIV
. Am J Med
69. Hornberger J, Torriani FJ, Dieterich DT, Bräu N, Sulkowski MS, Torres MR, et al. Cost-effectiveness of peginterferon alfa-2a (40 kDa) plus ribavirin in patients with HIV and hepatitis C virus co-infection
. J Clin Virol
70. Kuehne F, Bethe U, Freedberg K, Goldie SJ. Treatment for hepatitis C virus in human immunodeficiency virus–infected patients: clinical benefits and cost-effectiveness
. Arch Intern Med
71. Jones J, Shepherd J, Baxter L, Gospodarevskaya E, Hartwell D, Harris P, et al. Adefovir dipivoxil and pegylated interferon alpha for the treatment of chronic hepatitis B: an updated systematic review and economic evaluation
. Health Technol Assess
72. Lacey LF, Gane E. The cost-effectiveness of long-term antiviral therapy in the management of HBeAg-positive and HBeAg-negative chronic hepatitis B in Singapore
. J Viral Hepat
73. Dakin H, Bentley A, Dusheiko G. Cost-utility analysis of tenofovir disoproxil fumarate in the treatment of chronic hepatitis B
. Value In Health
74. Le Campion A, Larouche A, Fauteux-Daniel S, Soudeyns H. Pathogenesis of hepatitis C during pregnancy and childhood
75. van der Helm J, Geskus R, Sabin C, Meyer L, del Amo J, Chêne G, et al. Effect of HCV infection on cause-specific mortality following HIV seroconversion before and after 1997
76. Vlahov D, Graham N, Hoover D, Flynn C, Bartlett JG, Margolick JB, et al. Prognostic indicators for aids and infectious disease death in hiv-infected injection drug users: plasma viral load and CD4+ cell count
. J Am Med Assoc
77. Celentano DD, Vlahov D, Cohn S, Shadle VM, Obasanjo O, Moore RD. Self-reported antiretroviral therapy in injection drug users
. J Am Med Assoc
78. Nelson PK, Mathers BM, Cowle B, Hagan H, Des Jarlais DC, Horyniak D, et al. Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews
79. Vickerman P, Hickman M, May M, Kretzschmar M, Wiessing L. Can hepatitis C virus prevalence be used as a measure of injection-related human immunodeficiency virus risk in populations of injecting drug users? An ecological analysis
80. Degenhardt L, Mathers B, Vickerman P, Rhodes T, Latkin C, Hickman M. Prevention of HIV infection for people who inject drugs: why individual, structural, and combination approaches are needed
81. Strathdee SA, Hallett TB, Bobrova N, Rhodes T, Booth R, Abdool R, et al. HIV and risk environment for injecting drug users: the past, present, and future
82. Dore GJ. The changing therapeutic landscape for hepatitis C
. Med J Aust
83. Martin NK, Vickerman P, Grebely J, Hellard M, Hutchinson SJ, Lima VD, et al. HCV treatment for prevention among people who inject drugs: modeling treatment scale-up in the age of direct-acting antivirals
antiretroviral therapy; coinfection; hepatitis B virus; hepatitis C virus; HIV
Supplemental Digital Content
© 2014 Lippincott Williams & Wilkins, Inc.
Highlight selected keywords in the article text.