Impact of Hepatitis C Treatment as Prevention for People Who Inject Drugs is sensitive to contact network structure

Potential impact of curative and preventive interventions toward hepatitis C elimination in people who inject drugs-A network modeling study

BACKGROUND

Injection-equipment-sharing networks play an important role in hepatitis C virus (HCV) transmission among people who inject drugs (PWID). Direct-acting antiviral (DAA) treatments for HCV infection and interventions to prevent HCV transmission are critical components of an overall hepatitis C elimination strategy, but how they contribute to the elimination outcomes in different PWID network settings are unclear.

METHODS

We developed an agent-based network model of HCV transmission through the sharing of injection equipment among PWID and parameterized and calibrated the model with rural PWID data in the United States. We modeled curative and preventive interventions at annual coverage levels of 12.5 %, 25 %, or 37.5 % (cumulative percentage of eligible individuals engaged), and two allocation approaches: random vs targeting PWID with more injection partners (hereafter 'degree-based'). We compared the impact of these intervention strategies on prevalence and incidence of HCV infections. We conducted sensitivity analysis on key parameters governing the effects of curative and preventive interventions and PWID network characteristics.

RESULTS

Combining curative and preventive interventions at 37.5 % annual coverage with degree-based allocation decreased prevalence and incidence of HCV infection by 67 % and 70 % over two years, respectively. Curative interventions decreased prevalence by six to 12 times more than preventive interventions, while curative and preventive interventions had comparable effectiveness on reducing incidence. Intervention impact increased with coverage almost linearly across all intervention strategies, and degree-based allocation was always more effective than random allocation, especially for preventive interventions. Results were sensitive to parameter values defining intervention effects and network mean degree.

CONCLUSION

DAA treatments are effective in reducing both prevalence and incidence of HCV infection in PWID, but preventive interventions play a significant role in reducing incidence when intervention coverage is low. Increasing coverage, including efforts in reaching individuals with the most injection partners, preventing reinfection, and improving compliance and retention in preventive services can substantially improve the outcomes. PWID network characteristics should be considered when designing hepatitis C elimination programs.

Sociodemographic Variables Can Guide Prioritized Testing Strategies for Epidemic Control in Resource-Limited Contexts

BACKGROUND

Targeted surveillance allows public health authorities to implement testing and isolation strategies when diagnostic resources are limited, and can be implemented via the consideration of social network topologies. However, it remains unclear how to implement such surveillance and control when network data are unavailable.

METHODS

We evaluated the ability of sociodemographic proxies of degree centrality to guide prioritized testing of infected individuals compared to known degree centrality. Proxies were estimated via readily available sociodemographic variables (age, gender, marital status, educational attainment, household size). We simulated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) epidemics via a susceptible-exposed-infected-recovered individual-based model on 2 contact networks from rural Madagascar to test applicability of these findings to low-resource contexts.

RESULTS

Targeted testing using sociodemographic proxies performed similarly to targeted testing using known degree centralities. At low testing capacity, using proxies reduced infection burden by 22%-33% while using 20% fewer tests, compared to random testing. By comparison, using known degree centrality reduced the infection burden by 31%-44% while using 26%-29% fewer tests.

CONCLUSIONS

We demonstrate that incorporating social network information into epidemic control strategies is an effective countermeasure to low testing capacity and can be implemented via sociodemographic proxies when social network data are unavailable.

Deep learning and social network analysis elucidate drivers of HIV transmission in a high-incidence cohort of people who inject drugs

Globally, people who inject drugs (PWID) experience some of the fastest-growing HIV epidemics. Network-based approaches represent a powerful tool for understanding and combating these epidemics; however, detailed social network studies are limited and pose analytical challenges. We collected longitudinal social (injection partners) and spatial (injection venues) network information from 2512 PWID in New Delhi, India. We leveraged network analysis and graph neural networks (GNNs) to uncover factors associated with HIV transmission and identify optimal intervention delivery points. Longitudinal HIV incidence was 21.3 per 100 person-years. Overlapping community detection using GNNs revealed seven communities, with HIV incidence concentrated within one community. The injection venue most strongly associated with incidence was found to overlap six of the seven communities, suggesting that an intervention deployed at this one location could reach the majority of the sample. These findings highlight the utility of network analysis and deep learning in HIV program design.

Extent and implications of omitted ties on network measures in a longitudinal social network survey of people who use drugs

BACKGROUND

Previous research has revealed under-reporting of personal network members (i.e., alters) in studies involving people who use drugs (PWUD). This analysis (1) characterizes relationships that were more likely to be omitted but later recalled with prompting and (2) identifies network structural characteristics most impacted by these omissions among a sample of PWUD in rural Appalachian Kentucky, an epicenter of the opioid epidemic.

METHODS

Data were collected through longitudinal assessments as part of the Social Networks Among Appalachian People (SNAP) study (2008-2017). Study participants completed interviewer-administered questionnaires that collected social network data via free-listing at baseline and six-month intervals. At visit 5, after free-listing, interviewers prompted participants with the names of previously reported alters. We used modified Poisson regression with generalized estimating equations to identify individual- and relationship-level characteristics associated with an alter being reported only after prompting. We examined the impact of including vs. excluding relationships reported after prompting on local and global sociometric network measures (i.e., betweenness centrality, bridging, density, mean degree, transitivity, cliques, and 2-cores).

RESULTS

Relationships reported only after prompting were more likely to be immediate family (Adjusted Prevalence Ratio [APR]:1.29; 95% Confidence Interval [CI]: 1.03-1.63) and less likely to involve sex (APR:0.54; 95% CI: 0.43-0.67). Considerable differences were observed for participant positional rankings of betweenness centrality and bridging, and differences in network density and average degree pre- and post-prompting were statistically significant.

CONCLUSION

Longitudinal network studies that aim to assess transmission dynamics, information diffusion, or peer influence should consider the effects of omitted relationships.

Modelling the impact of HIV and HCV prevention and treatment interventions for people who inject drugs in Dar es Salaam, Tanzania

Introduction

People who inject drugs (PWID) in Dar es Salaam, Tanzania, have a high prevalence of HIV and hepatitis C virus (HCV). While needle and syringe programmes (NSP), opioid agonist therapy (OAT) and anti‐retroviral therapy (ART) are available in Tanzania, their coverage is sub‐optimal. We assess the impact of existing and scaled up harm reduction (HR) interventions on HIV and HCV transmission among PWID in Dar es Salaam.

Methods

An HIV and HCV transmission model among PWID in Tanzania was calibrated to data over 2006–2018 on HIV (∼30% and ∼67% prevalence in males and females in 2011) and HCV prevalence (∼16% in 2017), numbers on HR interventions (5254 ever on OAT in 2018, 766–1479 accessing NSP in 2017) and ART coverage (63.1% in 2015). We evaluated the impact of existing interventions in 2019 and impact by 2030 of scaling‐up the coverage of OAT (to 50% of PWID), NSP (75%, both combined termed “full HR”) and ART (81% with 90% virally suppressed) from 2019, reducing sexual HIV transmission by 50%, and/or HCV‐treating 10% of PWID infected with HCV annually.

Results

The model projects HIV and HCV prevalence of 19.0% (95% credibility interval: 16.4–21.2%) and 41.0% (24.4–49.0%) in 2019, respectively. For HIV, 24.6% (13.6–32.6%) and 70.3% (59.3–77.1%) of incident infections among male and female PWID are sexually transmitted, respectively. Due to their low coverage (22.8% for OAT, 16.3% for NSP in 2019), OAT and NSP averted 20.4% (12.9–24.7%) of HIV infections and 21.7% (17.0–25.2%) of HCV infections in 2019. Existing ART (68.5% coverage by 2019) averted 48.1% (29.7–64.3%) of HIV infections in 2019. Scaling up to full HR will reduce HIV and HCV incidence by 62.6% (52.5–74.0%) and 81.4% (56.7–81.4%), respectively, over 2019–2030; scaled up ART alongside full HR will decrease HIV incidence by 66.8% (55.6–77.5%), increasing to 81.5% (73.7–87.5%) when sexual risk is also reduced. HCV‐treatment alongside full HR will decrease HCV incidence by 92.4% (80.7–95.8%) by 2030.

Conclusions

Combination interventions, including sexual risk reduction and HCV treatment, are needed to eliminate HCV and HIV among PWID in Tanzania.

Injecting network structure determines the most efficient strategy to achieve Hepatitis C elimination in people who inject drugs

Transmission of Hepatitis C (HCV) continues via sharing of injection equipment between people who inject drugs (PWID). Network-based modelling studies have produced conflicting results about whether random treatment is preferable to targeting treatment at PWID with multiple partners. We hypothesise that differences in the modelled injecting network structure produce this heterogeneity. The study aimed to test how changing network structure affects HCV transmission and treatment effects. We created three dynamic injecting network structures connecting 689 PWID (UK-net, AUS-net and USA-net) based on published empirical data. We modelled HCV in the networks and at 5 years compared prevalence of HCV 1) with no treatment, 2) with randomly targeted treatment and 3) with treatment targeted at PWID with the most injecting partnerships (degree-based treatment). HCV prevalence at 5 years without treatment differed significantly between the three networks (UK-net (42.8%) vs. AUS-net (38.2%), p < 0.0001 and vs. USA-net (54.0%), p < 0.0001). In the treatment scenarios UK-net and AUS-net showed a benefit of degree-based treatment with a 5-year prevalence of 1.0% vs. 9.6% p < 0.0001 and 0.15% vs. 0.44%, p < 0.0001. USA-net showed no significant difference (29.3% vs. 29.2%, p = 0.0681). Degree-based treatment was optimised with low prevalence, moderate treatment coverage conditions whereas random treatment was optimised in low treatment coverage, high prevalence conditions. In conclusion, injecting network structure determines the transmission rate of HCV and the most efficient treatment strategy. In real-world injecting network structures, the benefit of targeting HCV treatment at individuals with multiple injecting partnerships may have been underestimated.

Psychosocial Factors and the Care Cascade for Hepatitis C Treatment Colocated at a Syringe Service Program

BACKGROUND

Direct-acting antiviral (DAA) therapy for hepatitis C virus (HCV) is highly effective. However, people who inject drugs face significant barriers to DAA access.

METHODS

We describe a program that colocates HCV management within a syringe service program in New York City. We performed a retrospective chart review of all patients with confirmed HCV viremia.

RESULTS

From 2015 to 2018, 102 patients with viremia completed intake. Fifty-eight patients started DAAs. Nine patients discontinued treatment or were lost to follow-up before completion; 1 is continuing DAA treatment. Of 48 patients who completed therapy, sustained virologic response (SVR) was achieved in 43 (89.6%). Age and established mental health treatment at intake were associated with SVR. Regular cocaine use was negatively associated with SVR in univariate analysis, but this association was not significant after adjustment for age. Of 30 patients completing DAA therapy with active illicit opioid use at intake, 14 (46.4%) engaged in opioid use disorder (OUD) treatment during therapy, and 9 remained in OUD treatment after completion of DAA treatment.

CONCLUSIONS

Loss to follow-up is a challenge for people who inject drugs, but among those who completed treatment, SVR was achieved at a high rate. Mental health treatment may facilitate HCV cure. Conversely, HCV therapy may facilitate engagement in OUD treatment and other services.

Cost-effectiveness of screening and treatment using direct-acting antivirals for chronic Hepatitis C virus in a primary care setting in Karachi, Pakistan

Despite the availability of effective direct-acting antiviral (DAA) treatments for Hepatitis C virus (HCV) infection, many people remain undiagnosed and untreated. We assessed the cost-effectiveness of a Médecins Sans Frontières (MSF) HCV screening and treatment programme within a primary health clinic in Karachi, Pakistan. A health state transition Markov model was developed to estimate the cost-effectiveness of the MSF programme. Programme cost and outcome data were analysed retrospectively. The incremental cost-effectiveness ratio (ICER) was calculated in terms of incremental cost (2016 US$) per disability-adjusted life year (DALY) averted from the provider's perspective over a lifetime horizon. The robustness of the model was evaluated using deterministic and probabilistic sensitivity analyses (PSA). The ICER for implementing testing and treatment compared to no programme was US$450/DALY averted, with 100% of PSA runs falling below the per capita Gross Domestic Product threshold for cost-effective interventions for Pakistan (US$1,422). The ICER increased to US$532/DALY averted assuming national HCV seroprevalence (5.5% versus 33% observed in the intervention). If the cost of liver disease care was included (adapted from resource use data from Cambodia which has similar GDP to Pakistan), the ICER dropped to US$148/DALY, while it became cost-saving if a recently negotiated reduced drug cost of $75/treatment course was assumed (versus $282 in base-case) in addition to cost of liver disease care. In conclusion, screening and DAA treatment for HCV infection are expected to be highly cost-effective in Pakistan, supporting the expansion of similar screening and treatment programmes across Pakistan.

A review of network simulation models of hepatitis C virus and HIV among people who inject drugs

Network modelling is a valuable tool for simulating hepatitis C virus (HCV) and HIV transmission among people who inject drugs (PWID) and assessing the potential impact of treatment and harm-reduction interventions. In this paper, we review literature on network simulation models, highlighting key structural considerations and questions that network models are well suited to address. We describe five approaches (Erdös-Rényi, Stochastic Block, Watts-Strogatz, Barabási-Albert, and Exponential Random Graph Model) used to model partnership formation with emphasis on the strengths of each approach in simulating different features of real-world PWID networks. We also review two important structural considerations when designing or interpreting results from a network simulation study: (1) dynamic vs. static network and (2) injection only vs. both injection and sexual networks. Dynamic network simulations allow partnerships to evolve and disintegrate over time, capturing corresponding shifts in individual and population-level risk behaviour; however, their high level of complexity and reliance on difficult-to-observe data has driven others to develop static network models. Incorporating both sexual and injection partnerships increases model complexity and data demands, but more accurately represents HIV transmission between PWID and their sexual partners who may not also use drugs. Network models add the greatest value when used to investigate how leveraging network structure can maximize the effectiveness of health interventions and optimize investments. For example, network models have shown that features of a given network and epidemic influence whether the greatest community benefit would be achieved by allocating hepatitis C or HIV treatment randomly, versus to those with the most partners. They have also demonstrated the potential for syringe services and "buddy sharing" programs to reduce disease transmission.

Estimation and correction of bias in network simulations based on respondent-driven sampling data

Respondent-driven sampling (RDS) is widely used for collecting data on hard-to-reach populations, including information about the structure of the networks connecting the individuals. Characterizing network features can be important for designing and evaluating health programs, particularly those that involve infectious disease transmission. While the validity of population proportions estimated from RDS-based datasets has been well studied, little is known about potential biases in inference about network structure from RDS. We developed a mathematical and statistical platform to simulate network structures with exponential random graph models, and to mimic the data generation mechanisms produced by RDS. We used this framework to characterize biases in three important network statistics – density/mean degree, homophily, and transitivity. Generalized linear models were used to predict the network statistics of the original network from the network statistics of the sample network and observable sample design features. We found that RDS may introduce significant biases in the estimation of density/mean degree and transitivity, and may exaggerate homophily when preferential recruitment occurs. Adjustments to network-generating statistics derived from the prediction models could substantially improve validity of simulated networks in terms of density, and could reduce bias in replicating mean degree, homophily, and transitivity from the original network.

A Maximum Entropy Method for the Prediction of Size Distributions

We propose a method to derive the stationary size distributions of a system, and the degree distributions of networks, using maximisation of the Gibbs-Shannon entropy. We apply this to a preferential attachment-type algorithm for systems of constant size, which contains exit of balls and urns (or nodes and edges for the network case). Knowing mean size (degree) and turnover rate, the power law exponent and exponential cutoff can be derived. Our results are confirmed by simulations and by computation of exact probabilities. We also apply this entropy method to reproduce existing results like the Maxwell-Boltzmann distribution for the velocity of gas particles, the Barabasi-Albert model and multiplicative noise systems.

Prevalence of HIV/Hepatitis C Virus Co-Infection and Injection Risk Correlations in People Who Inject Drugs in Colombia: A Cross-Sectional Study Using Respondent Driven Sampling

Background: In Colombia, an exponential growth of people who inject drugs (PWID) has been reported over the past decades. The main burden of disease in PWID is attributed to human immunodeficiency virus (HIV) and hepatitis C virus (HCV) and their co-infection is associated with poor prognosis, high morbidity and treatment related implications. Nevertheless, the prevalence of HIV/HCV co-infection in Colombian PWIDs is unknown. Objective: To determine the prevalence, demographic characteristics, and injecting behaviors of HIV/HCV co-infected PWID in Colombia. Methods: This was a cross-sectional study of 1,123 PWID recruited by respondent driven sampling in five Colombian cities between January and June of 2014. Each participant completed a quantitative survey, and blood samples for HIV and HCV antibody testing were obtained. A multinomial logistic regression was used for statistical analysis. Results: Average participant age was 26.3 ± 6.5 years, and the majority was male (86%). HIV or HCV mono-infection prevalence was 27.6%, while co-infection was 3.3%. Compared with PWID with mono-infection, co-infected PWID exhibited higher odds of: injecting ≥4 times daily (OR: 3.5; CI: 1.7-7.2; p < .001), cleaning needles and syringes with water (OR 3.2; CI: 1.6-6.3; p < .001), passing drug mix between syringes (OR: 2.7; CI: 1.3-5.3; p = .04), injecting on illegal indoor shooting galleries (OR: 2.4; CI: 1.0-5.3, p = .02), and getting injected by someone who charges for injecting (OR 2.3; CI: 1.0-5.2; p = .04). Conclusion: Prevalence of HIV/HCV co-infection among PWID in Colombia is lower than that reported in other countries. However, addressing the identified demographic characteristics and injection risk behaviors of co-infected PWID is essential for the implementation of broadly available harm reduction interventions as well as routine HIV/HCV testing and treatment strategies aiming to control the spread of both viruses and their associated morbidity and mortality.

Prioritising Hepatitis C treatment in people with multiple injecting partners maximises prevention: A real-world network study

OBJECTIVE

To describe an injecting network of PWID living in an isolated community on the Isle of Wight (UK) and the results of a agent-based simulation, testing the effect of Hepatitis C (HCV) treatment on transmission.

METHOD

People who inject drugs (PWID) were identified via respondent driven sampling and recruited to a network and bio-behavioural survey. The injecting network they described formed the baseline population and potential transmission pathways in an agent-based simulation of HCV transmission and the effects of treatment over 12 months.

RESULTS

On average each PWID had 2.6 injecting partners (range 0-14) and 137 were connected into a single component. HCV in the network was associated with a higher proportion of positive injecting partners (p = 0.003) and increasing age (p = 0.011). The treatment of well-connected PWID led to significantly fewer new infections of HCV than treating at random (10 vs. 7, p<0.001). In all scenarios less than one individual was re-infected.

CONCLUSION

In our model the preferential treatment of well-connected PWID maximised treatment as prevention. In the real-world setting, targeting treatment to actively injecting PWID, with multiple injecting partners may therefore represent the most efficient elimination strategy for HCV.

Mathematical modeling of hepatitis c virus (HCV) prevention among people who inject drugs: A review of the literature and insights for elimination strategies

In 2016, the World Health Organization issued global elimination targets for hepatitis C virus (HCV), including an 80% reduction in HCV incidence by 2030. The vast majority of new HCV infections occur among people who inject drugs (PWID), and as such elimination strategies require particular focus on this population. As governments urgently require guidance on how to achieve elimination among PWID, mathematical modeling can provide critical information on the level and targeting of intervention are required. In this paper we review the epidemic modeling literature on HCV transmission and prevention among PWID, highlight main differences in mathematical formulation, and discuss key insights provided by these models in terms of achieving WHO elimination targets among PWID. Overall, the vast majority of modeling studies utilized a deterministic compartmental susceptible-infected-susceptible structure, with select studies utilizing individual-based network transmission models. In general, these studies found that harm reduction alone is unlikely to achieve elimination targets among PWID. However, modeling indicates elimination is achievable in a wide variety of epidemic settings with harm reduction scale-up combined with modest levels of HCV treatment for PWID. Unfortunately, current levels of testing and treatment are generally insufficient to achieve elimination in most settings, and require further scale-up. Additionally, network-based treatment strategies as well as prison-based treatment and harm reduction provision could provide important additional population benefits. Overall, epidemic modeling has and continues to play a critical role in informing HCV elimination strategies worldwide.

Hepatitis C

Hepatitis C is a global health problem, and an estimated 71·1 million individuals are chronically infected with hepatitis C virus (HCV). The global incidence of HCV was 23·7 cases per 100 000 population (95% uncertainty interval 21·3-28·7) in 2015, with an estimated 1·75 million new HCV infections diagnosed in 2015. Globally, the most common infections are with HCV genotypes 1 (44% of cases), 3 (25% of cases), and 4 (15% of cases). HCV transmission is most commonly associated with direct percutaneous exposure to blood, via blood transfusions, health-care-related injections, and injecting drug use. Key high-risk populations include people who inject drugs, men who have sex with men, and prisoners. Approximately 10-20% of individuals who are chronically infected with HCV develop complications, such as cirrhosis, liver failure, and hepatocellular carcinoma over a period of 20-30 years. Direct-acting antiviral therapy is now curative, but it is estimated that only 20% of individuals with hepatitis C know their diagnosis, and only 15% of those with known hepatitis C have been treated. Increased diagnosis and linkage to care through universal access to affordable point-of-care diagnostics and pangenotypic direct-acting antiviral therapy is essential to achieve the WHO 2030 elimination targets.

Research gaps in viral hepatitis

Introduction

The World Health Organization has aimed for global elimination of both hepatitis B virus (HBV) and hepatitis C virus (HCV) by 2030. Treatments available to cure HCV and control HBV, as well as vaccination to prevent HBV infection, have certainly allowed for such bold goals, yet the final steps to usher in elimination require further evidence.

Discussion

We broadly discuss the needs for three major public health approaches. First, an effective vaccine exists for HBV and mass‐vaccination campaigns have resulted in decreases in hepatitis B surface antigen seroprevalence and overall rates of liver‐related morality. Still, HBV vaccination coverage is poor in certain regions of the world, while the reasons for such low coverage require further study. A prophylactic vaccine is probably needed to eliminate HCV, but is not being readily developed. Second, identifying HBV/HCV infected individuals remains a priority to increase awareness of disease status, particularly for key populations. Research evaluating large‐scale implementation of novel, rapid and mobile point‐of‐care tests would be helpful to determine whether increased awareness is achievable in these settings. Third, antiviral therapy allows for strong HBV suppression and HCV cure, while its access depends on financial factors among many others. Although there is strong evidence to treat key populations and specific groups with progressed disease, as stated in current guidelines, the advantages of extending treatment eligibility to decrease onward spread of HBV/HCV infection and prevent further burden of disease are lacking “real world” evidence. Novel anti‐HBV treatments are being developed to target intrahepatocellular HBV replication, but are still in the early phases of clinical development. Each of the strategies mentioned above has specific implications for HIV infection.

Conclusions

There are certainly effective tools to combat the spread of viral hepatitis and treat infected individuals – yet how they are able to reach key populations, and the infrastructure required to do so, continue to represent the largest research gap when evaluating the progress towards elimination. Continuously adapted and informed research is required to establish the priorities in achieving elimination goals.

Prevalence of hepatitis C infection among the general population and high-risk groups in the EU/EEA: a systematic review update

Background

Although significant improvement in efficacy measured by a sustained virological response, the high acquisition costs of direct-acting antivirals limit the access for patients and influence the costs of healthcare resource utilisation in hepatitis C. It is important to have the latest estimates of prevalence, especially in high-risk groups, for cost of illness, cost-effectiveness and budget impact studies.

Methods

Original studies on the estimates of the prevalence among general and high-risk groups in the European Union/European Economic Area (EU/EEA) were retrieved from Medline and Embase for the period from 2015 to 2018. All included studies were evaluated for risk of selection bias and summarised together in a narrative form. Results from previous reviews and updated searches were compared per country among different populations, respectively.

Results

Among the 3871 studies identified, 46 studies were included: 20 studies were used for the estimate of the general population; 3 for men who have sex with men (MSM); 6 for prisoners; and 17 for people who inject drugs (PWID). Compared with the results reported in previous systematic reviews, the updated estimates were lower than previously in most available countries. Anti-HCV general population prevalence estimates ranged from 0.54 to 1.50% by country. The highest prevalence of anti-HCV was found among PWID (range of 7.90–82.00%), followed by prisoners (7.00–41.00%), HIV-positive MSM (1.80–7.10%), HIV-negative MSM (0.20–1.80%), pregnant women (0.10–1.32%) and first-time blood donors (0.03–0.09%).

Conclusions

Our study highlights the heterogeneity in anti-HCV prevalence across different population groups in EU/EEA. The prevalence also varies widely between European countries. There are many countries that are not represented in our results, highlighting the need for the development of robust epidemiological studies.

Electronic supplementary material

The online version of this article (10.1186/s12879-019-4284-9) contains supplementary material, which is available to authorized users.

Hepatitis C transmission in young people who inject drugs: Insights using a dynamic model informed by state public health surveillance

Increasing injection of heroin and prescription opioids have led to increases in the incidence of hepatitis C virus (HCV) infections in US young adults since the early 2000s. How best to interrupt transmission and decrease HCV prevalence in young people who inject drugs (PWID) is uncertain. We developed an age-stratified ordinary differential equation HCV transmission model of PWID aged 15-64, which we fit to Michigan HCV surveillance data among young PWID aged 15-29. We used Latin hypercube sampling to fit to data under 10,000 plausible model parameterizations. We used the best-fitting 10% of simulations to predict the potential impact of primary (reducing injection initiation), secondary (increasing cessation, reducing injection partners, or reducing injection drug use relapse), and tertiary (HCV treatment) interventions (over the period 2017-2030) on acute and chronic HCV cases by the year 2030. Treating 3 per 100 current and former PWID per year could reduce chronic HCV by 27.3% (range: 18.7-30.3%) and acute HCV by 23.6% (range: 6.7-29.5%) by 2030 among PWID aged 15-29 if 90% are cured (i.e. achieved sustained virologic response [SVR] to treatment). Reducing the number of syringe sharing partners per year by 10% was predicted to reduce chronic HCV by 15.7% (range: 9.4-23.8%) and acute cases by 21.4% (range: 14.2-32.3%) among PWID aged 15-29 by 2030. In simulations of combinations of interventions, reducing injection initiation, syringe sharing, and relapse rates each by 10% while increasing cessation rates by 10% predicted a 27.7% (range: 18.0-39.7%) reduction in chronic HCV and a 38.4% (range: 28.3-53.3%) reduction in acute HCV. Our results highlight the need for HCV treatment among both current and former PWID and the scale up of both primary and secondary interventions to concurrently reduce HCV prevalence and incidence in Michigan.

Phylogenies from dynamic networks

The relationship between the underlying contact network over which a pathogen spreads and the pathogen phylogenetic trees that are obtained presents an opportunity to use sequence data to learn about contact networks that are difficult to study empirically. However, this relationship is not explicitly known and is usually studied in simulations, often with the simplifying assumption that the contact network is static in time, though human contact networks are dynamic. We simulate pathogen phylogenetic trees on dynamic Erdős-Renyi random networks and on two dynamic networks with skewed degree distribution, of which one is additionally clustered. We use tree shape features to explore how adding dynamics changes the relationships between the overall network structure and phylogenies. Our tree features include the number of small substructures (cherries, pitchforks) in the trees, measures of tree imbalance (Sackin index, Colless index), features derived from network science (diameter, closeness), as well as features using the internal branch lengths from the tip to the root. Using principal component analysis we find that the network dynamics influence the shapes of phylogenies, as does the network type. We also compare dynamic and time-integrated static networks. We find, in particular, that static network models like the widely used Barabasi-Albert model can be poor approximations for dynamic networks. We explore the effects of mis-specifying the network on the performance of classifiers trained identify the transmission rate (using supervised learning methods). We find that both mis-specification of the underlying network and its parameters (mean degree, turnover rate) have a strong adverse effect on the ability to estimate the transmission parameter. We illustrate these results by classifying HIV trees with a classifier that we trained on simulated trees from different networks, infection rates and turnover rates. Our results point to the importance of correctly estimating and modelling contact networks with dynamics when using phylodynamic tools to estimate epidemiological parameters.

HCV seroconversion in a cohort of people who use drugs followed in a mobile harm reduction unit in Madrid: Breaking barriers for HCV elimination

Background and aims

Harm reduction strategies have been shown to decrease the incidence of human immunodeficiency virus (HIV) infection in people who inject drugs (PWID), but the results have been inconsistent when it comes to prevention of hepatitis C virus (HCV) infection. We aimed to examine the rate of HCV seroconversion among people who use drugs (PWUD) followed at a mobile harm reduction unit (MHRU) to evaluate if a low-threshold methadone substitution program (LTMSP) is associated with a low HCV seroconversion rate and subsequently identify barriers for elimination.

Materials and methods

A cohort of PWUD have been followed at a MRHU in Madrid between 2013 and 2016. Individuals who were negative for HCV antibodies at baseline and who had at least one retest for HCV antibodies were eligible. Kaplan-Meier methods were employed to estimate the global incidence density.

Results

During the study period, 946 PWUD were screened for HCV at least once. At baseline 127 PWUD were negative for HCV antibodies and had at least one follow-up HCV antibodies test. The baseline HCV prevalence was 33%. After a median 0.89 (IQR 0.3–1.5) years of follow-up and 135 person-years of risk for HCV infection, 28 subjects seroconverted. The incidence density for HCV seroconversion for this sample was 20.7 cases (95% CI: 14.3–29.7) per 100 person-years. Injecting drugs in the last year was strongly associated to HCV seroconversion (AHR 15.5, 95%CI 4.3–55.8, p < 0.001). Methadone status was not associated to HCV seroconversion.

Conclusions

A high incidence of HCV infection was found among PWUD at a MHRU in Madrid. In this setting opiate substitutive treatment (OST) as a LTMSP does not appear to protect against HCV seroconversion.

Economically Efficient Hepatitis C Virus Treatment Prioritization Improves Health Outcomes

BACKGROUND

The total cost of treating the 3 million Americans chronically infected with hepatitis C virus (HCV) represents a substantial affordability challenge requiring treatment prioritization. This study compares the health and economic outcomes of alternative treatment prioritization schedules.

METHODS

We developed a multiyear HCV treatment budget allocation model to evaluate the tradeoffs of 7 prioritization strategies. We used optimization to identify the priority schedule that maximizes population net monetary benefit (NMB). We compared prioritization schedules in terms of the number of individuals treated, the number of individuals who progress to end-stage liver disease (ESLD), and population total quality-adjusted life years (QALYs). We applied the model to the population of treatment-naive patients with a total annual HCV treatment budget of US$8.6 billion.

RESULTS

First-come, first-served (FCFS) treats the fewest people with advanced fibrosis, prevents the fewest cases of ESLD, and gains the fewest QALYs. A schedule developed from optimizing population NMB prioritizes treatment in the first year to patients with moderate to severe fibrosis who are younger than 65 years, followed by older individuals with moderate to severe fibrosis. While this strategy yields the greatest population QALYs, prioritization by disease severity alone prevents more cases of ESLD. Sensitivity analysis indicated that the differences between prioritization schedules are greater when the budget is smaller. A 10% annual treatment price reduction enabled treatment 1 year sooner to several patient subgroups, specifically older patients and those with less severe liver fibrosis.

CONCLUSION

In the absence of a sufficient budget to treat all patients, explicit prioritization targeting younger people with more severe disease first provides the greatest health benefits. We provide our spreadsheet model so that decision makers can compare health tradeoffs of different budget levels and various prioritization strategies with inputs tailored to their population.

Population Health and Cost-Effectiveness Implications of a "Treat All" Recommendation for HCV: A Review of the Model-Based Evidence

The World Health Organization HCV Guideline Development Group is considering a "treat all" recommendation for persons infected with hepatitis C virus (HCV). We reviewed the model-based evidence of cost-effectiveness and population health impacts comparing expanded treatment policies to more limited treatment access policies, focusing primarily on evaluations of all-oral directly acting antivirals published after 2012. Searching PubMed, we identified 2,917 unique titles. Sequentially reviewing titles and abstracts identified 226 potentially relevant articles for full-text review. Sixty-nine articles met all inclusion criteria-42 cost-effectiveness analyses and 30 models of population-health impacts, with 3 articles presenting both types of analysis. Cost-effectiveness studies for many countries concluded that expanding treatment to people with mild liver fibrosis, who inject drugs (PWID), or who are incarcerated is generally cost-effective compared to more restrictive treatment access policies at country-specific prices. For certain patient subpopulations in some countries-for example, elderly individuals without fibrosis-treatment is only cost-effective at lower prices. A frequent limitation is the omission of benefits and consequences of HCV transmission (i.e., treatment as prevention; risks of reinfection), which may underestimate or overestimate the cost-effectiveness of a "treat all" policy. Epidemiologic modeling studies project that through a combination of prevention, aggressive screening and diagnosis, and prompt treatment for all fibrosis stages, it may be possible to virtually eliminate HCV in many countries. Studies show that if resources are not available to diagnose and treat all HCV-infected individuals, treatment prioritization may be needed, with alternative prioritization strategies resulting in tradeoffs between reducing mortality or reducing incidence. Notably, because most new HCV infections are among PWID in many settings, HCV elimination requires unrestricted treatment access combined with injection transmission disruption strategies. The model-based evidence suggests that a properly constructed strategy that substantially expands HCV treatment could achieve cost-effective improvements in population health in many countries.

Assessing uncertainty in the burden of hepatitis C virus: Comparison of estimated disease burden and treatment costs in the UK

Hepatitis C virus (HCV) is a major and growing public health concern. We need to know the expected health burden and treatment cost, and understand uncertainty in those estimates, to inform policymaking and future research. Two models that have been important in informing treatment guidelines and assessments of HCV burden were compared by simulating cohorts of individuals with chronic HCV infection initially aged 20, 35 and 50 years. One model predicts that health losses (measured in quality-adjusted life-years [QALYs]) and treatment costs decrease with increasing initial age of the patients, whilst the other model predicts that below 40 years, costs increase and QALY losses change little with age, and above 40 years, they decline with increasing age. Average per-patient costs differ between the models by up to 38%, depending on the patients' initial age. One model predicts double the total number, and triple the peak annual incidence, of liver transplants compared to the other model. One model predicts 55%-314% more deaths than the other, depending on the patients' initial age. The main sources of difference between the models are estimated progression rates between disease states and rates of health service utilization associated with different disease states and, in particular, the age dependency of these parameters. We conclude that decision-makers need to be aware that uncertainties in the health burden and economic cost of HCV disease have important consequences for predictions of future need for care and cost-effectiveness of interventions to avert HCV transmission, and further quantification is required to inform decisions.

Treatment of hepatitis C with new fixed dose combinations

INTRODUCTION

The advent of oral direct-acting antivirals (DAA) has revolutionized the hepatitis C virus (HCV) therapeutic landscape providing cure rates over 90%. However, a subset of patients remains at higher risk for treatment failure, including those infected with: i) genotype 3 and cirrhosis; ii) resistance-associated substitutions (RAS) occurring either as natural polymorphisms or selected after prior DAA failure; and iii) poor drug adherence associated with social disabilities (homeless, psychiatric illnesses, injection drug use, alcoholism, etc.). Whereas discovery of new DAA with increased antiviral activity across all genotypes and over RAS may enhance efficacy, development of fixed dose combinations (FDC) may be the best way to improve drug adherence in difficult-to-treat HCV populations. Areas covered: Three FDC regimens are in the last steps of clinical development for treating hepatitis C. Two distinct nucleotide analogues that inhibit the HCV polymerase, sofosbuvir and uprifosbuvir, are part of the FDC from Gilead and Merck, respectively. The AbbVie dual FDC does not include a polymerase inhibitor. All three new FDC include second-generation NS3 protease inhibitors and NS5A inhibitors active across all HCV genotypes and over common RAS. Expert opinion: Hepatitis C cure rates over 95% are expected with all three next-coming DAA, even in the most difficult-to-treat and/or cure patient populations. These regimens would be particularly needed for the growing number of prior DAA failures. Co-formulations and 8-week shorter treatment lengths will help to overcome drug adherence challenges in certain populations.