ORIGINAL ARTICLE 227. Mitigating the burden of hepatitis C virus among people who inject drugs in Belgium. Methods. Introduction

Transcription

1 Reference number to be mentioned by correspondence : AG/Gastro-4609.R0-mathei- ORIGINAL ARTICLE 227 Mitigating the burden of hepatitis C virus among people who inject drugs in Belgium Catharina Matheï 1, Stefan Bourgeois 2, Sarah Blach 3, Christian Brixko 4, Jean-Pierre Mulkay 5, Homie Razavi 3, Geert Robaeys 6,7,8 (1) Department of Public Health and Primary Care, University Hospitals Leuven, KU Leuven, Leuven, Belgium ; (2) ZNA Campus Stuivenberg, Antwerp, Belgium ; (3) Center for Disease Analysis, Louisville, CO, USA ; (4) Department of Gastroenterology and Digestive Oncology, CHR Citadelle, Liege, Belgium ; (5) Hepatogastroenterology, CHU Saint-Pierre, Brussels, Belgium ; (6) Department of Gastroenterology and Hepatology, Ziekenhuis Oost-Limburg, Genk, Belgium ; (7) Faculty of Medicine and Life Sciences, Hasselt University, Diepenbeek, Belgium ; (8) Department of Hepatology, UZ Leuven, Leuven, Belgium. Abstract Background and Aims : In 2010, there were an estimated PWID in Belgium and 43% (34%-57%) were HCV infected. Understanding HCV transmission dynamics in high-risk populations and assessing the potential impact of improved HCV treatment strategies requires robust epidemiological data and mathematical modeling. Methods : HCV transmission was modeled using cohorts to track HCV incidence and prevalence among active PWID in the general PWID population, OST and NSP. Model assumptions were derived from published literature and expert consensus. The relative impact of increasing the number of PWID treated with new oral DAAs was considered. Results : If the current transmission paradigm continues, there will be 2645 HCV-infected PWID in Annually treating 30 (1% of 2015 population) or 120 (4% of 2015 population) HCV-infected PWID with oral DAAs will result in 5% and 25% reductions, respectively, in HCV-infected PWID by Treating 370 PWID annually (12.5% of 2015 population) will result in a > 90% reduction by Conclusion : Treating a small number of PWID can result in substantial reduction in HCV prevalence in this population ; however, high levels of treatment are necessary to reduce the viral pool and thus the risk of secondary infections. This analysis supports implementation of a screening and treatment strategy among PWID when combined with an expansion of harm reduction programs. (Acta gastro enterol. belg., 2016, 79, ). Key words : hepatitis C, PWID, injecting drug users, transmission, incidence, prevalence. Introduction Injection drug use is the primary mode of hepatitis C virus (HCV) transmission in high-income countries (1). In Belgium, the number of people who reported ever injecting drugs remained stable from 2000 to 2013 (2,3). Harm reduction services including needle and syringe programs (NSP) have been offered in the French community since 1994 and in the Flemish community since 2001 (4). NSP services are offered through low-threshold facilities (LTF), as well as mobile outreach facilities and pharmacies (4). The provision of methadone and buprenorphine for opiate substitution therapy (OST) have been available in Belgium since 1994, but they have only been legally recognized since 2002 (4). Understanding HCV transmission dynamics among high-risk populations requires robust epidemiological data and country-specific mathematical modeling to assess the potential impact of improved HCV treatment strategies. Recent therapeutic advances promise greater convenience (oral therapies) with higher efficacy (> 90% sustained viral response) and shorter duration of treatment ; however, the impact of these therapies on the HCV incidence and prevalence in Belgium is yet unknown. The aim of this analysis was to determine the size and distribution of the HCV-infected population of people who inject drugs (PWID) and to assess the impact of treatment on the total number of HCV infections as well as the number of secondary infections following a HCV cure. While a majority of transmission occurs through the sharing of needles and syringes, transmission may still occur via the sharing of injecting paraphernalia (5). This analysis focuses on PWID who are currently injecting (within the last 12 months), as this is the population at primary risk of transmitting HCV or becoming infected. Methods Modeling the PWID population The HCV transmission model was developed as an add-on module to a previously described HCV disease burden model (6-8). Among PWID, the population was defined by HCV infection [HCV-RNA (+) or HCV-RNA (-)] and risk-behavior status (sharing needles or not sharing). At the most basic level, PWID were modeled to move across segments in a limited number of directions. Behavior changes resulting in needle sharing (among those previously not sharing) or in a discontinuation of needle sharing (among those previously sharing) could occur, or PWID could move into or out of NSP and OST. Additionally, PWID who share needles and were previously not infected with HCV could become infected. And finally, PWID who were HCV infected (regardless Disclosure of Interest Statement: This study was sponsored by Gilead Sciences, Inc. Correspondence to : Catharina Matheï, Department of Public Health and Primary Care, KU Leuven Address : Kapucijnenvoer 33 Blok J Bus 7001, 3000 Leuven, Belgium. catharina.mathei@med.kuleuven.be Submission date : 30/04/2016 Acceptance date : 05/05/2016

2 228 C. Matheï et al. of sharing behaviors) could be cured of their HCV through treatment with antivirals. PWID exited the model due to mortality (annual rate of 2% assumed for Belgium based on analog data from Luxembourg and expert consensus) or cessation of injection drug use. The rate of cessation was calculated on an annual basis to match historical trends in the overall PWID population size. The number of new PWID was calculated to balance the number of PWID leaving the population. Additionally, participation in high-coverage needle and syringe programs (NSP), opiate substitution therapy (OST), and NSP & OST was modeled. After 2015, all inputs were modeled to remain constant. Modeling new infections New infections were calculated using the probability of infection multiplied by the size of the PWID population that was sharing needles but not currently HCVinfected. The estimated probability of infection was calculated using the average annual number of unsafe injections, HCV prevalence in the injecting population, and the probability of HCV transmission from a contaminated needle [1.8% (Range : 0.0%-10.3%)] (9-11). Although transmission of HCV may still occur via sexual transmission or non-injection drug use, for the purposes of this analysis, the risk of transmission among non-injecting equipment-sharing PWID was assumed to be negligible. Engagement in NSP or OST was associated with a 50% reduction in unsafe injections, and simultaneous involvement in NSP and OST was associated with a 75% reduction in unsafe injections (12-17). Estimating the size of the HCV-infected PWID population A 2013 study to estimate the size of the PWID population employed a benchmark- multiplier method using data from the Belgian HIV/AIDS register as a benchmark and a sero-behavioral study among 1122 PWID in contact with drug treatment services or imprisoned as a multiplier (2). This study estimated (95% CI : ) ever-injectors living in Belgium, and approximately 41% were estimated to be currently injecting in 2010 (2). This translates to approximately (Range : ) active PWID. The anti-hcv prevalence among PWID was 43% (Range : 34%- 57%) (18). Accounting for a spontaneous clearance rate of 25%, the viremic prevalence in the PWID population was 32%, compared with a 0.7% viremic prevalence in the Belgian adult population (8,19,20). PWID behavior and harm reduction programs The number of individuals enrolled in OST in Belgium nearly doubled between 2003 (9000 participants) and 2012 ( participants) (18). Among these, 23% were estimated to continue to inject while participating in OST (21). The remaining 77% of the OST population were assumed to not be at risk of becoming infected with HCV or transmitting the disease via injection drug use, and were not included in the model. In 2012, syringes were distributed both through NSPs in the French and Flemish communities, as well as through pharmacies in the French community (18). Assuming an average of one sterile needle distributed per participant per day, the calculated number of PWID engaged in NSP was The term general PWID population was used for PWID who were not engaged in any harm reduction programs. Estimates for the proportion of PWID sharing injecting equipment ranged from 14.2% to 40% (18). A sharing rate of 31% was modeled to achieve an average time to infection of 2.7 years in the general PWID population (22). Scenarios After the model was populated and calibrated to reflect historical trends in the number of HCV+ PWID, scenarios were developed to assess the impact of providing HCV treatment with direct-acting antivirals (DAAs). Projected sustained viral response (SVR) rates in Belgium have been described previously (8). Treatment of 1%, 4%, 10% and 12.5% of the total HCV+ PWID population was modeled beginning in 2016, and the resulting impact on total HCV infections as well as secondary infections was calculated. In the absence of behavioral changes, cured PWID remained susceptible to secondary infection with HCV, with an average time to infection of 2.7 years. Sensitivity Analysis The uncertainty in inputs (number of PWID, percent injecting, percent sharing needles, etc.) was captured as a range using a Beta-PERT distribution. Monte Carlo simulations were run to determine 95% uncertainty intervals using Crystal Ball, an Excel add-in by Oracle. Results Baseline In 2015, there were an estimated 9080 PWID in Belgium, 47% of whom were engaged in harm reduction efforts including NSP and/or OST. Although individuals were enrolled in OST, only 3230 continued to inject. NSP services were accessed by 1880 PWID, and approximately 810 PWID on NSP were also being treated in OST (Fig. 1). Overall, 2970 PWID were HCV infected, with an estimated 160 new HCV infections occurring in The highest annual number of new infections occurred in the general PWID population (115 new cases), with 15 to 25 cases occurring in NSP or OST programs and 5 cases occurring among PWID engaged in both NSP and OST (Fig. 2b). In the absence of treatment, there

3 HCV among PWID in Belgium 229 Fig. 1. Size of the OST, NSP and PWID populations, 2015 were no secondary infections. Over the next 15 years, the HCV-infected PWID population was estimated to decrease by 10%, likely due to an aging population and current harm reduction efforts. Sensitivity analysis showed that the uncertainties around the number of PWID in Belgium and the percent of PWID sharing needles account for more than 94% of the observed variability in HCV prevalence among PWID. Increased Treatment Treatment of 30 patients annually (1% of the PWID population) resulted in a 5% reduction in total HCV infections among PWID by 2030 (Fig. 3a). Secondary infections among treated and cured HCV patients were projected to reach 18 annually by Under this scenario, a total of 445 patients were treated between 2016 and 2030, with a resulting 203 secondary infections projected (Fig. 4). Treatment of 120 patients annually (4% of the PWID population) resulted in a 25% reduction in total HCV infections among PWID, with 70 secondary infections annually by 2030 (Fig. 3b). Under this scenario, a total of 1781 patients were treated between 2016 and 2030, with a cumulative 772 secondary infections projected (Fig. 4). Increasing treatment to 297 patients annually (10% of the PWID population) resulted in an 85% reduction in total HCV infections among PWID. Secondary *Gen pop refers to the PWID population not in NSP and/or OST. Fig. 2. a) Distribution of HCV-infected PWID by cohort, ; b) Distribution of new HCV infections among PWID by cohort,

4 230 C. Matheï et al. Fig. 3. Annual viremic cases, and new secondary infection, by percent of the total HCV+ PWID population treated, infections were projected to peak in 2024 at 119, before decreasing to 82 annually by Under this scenario, treatment of 297 patients annually could only be sustained until 2027, after which point the model ran out of patients and treatment numbers were reduced to 273 by From 2016 to 2030, a total of 4400 patients were treated and a cumulative 1401 secondary infections were projected (Fig. 4). Treatment of 370 patients annually (12.5% of the PWID population) resulted in a greater than 99% reduction in total HCV infections among PWID. Secondary infections were projected to peak in 2022 at 125, before decreasing to nearly 0 by Under this scenario, treatment of 370 patients annually could only be sustained until 2024, after which point the model ran out of patients and treatment numbers were reduced to 24 by From 2016 to 2030, a total of 4500 patients were treated with 1145 cumulative secondary infections projected (Fig. 4). Discussion The aim of this analysis was to determine the size and distribution of the HCV-infected PWID population and to assess the impact of treatment on the total number of HCV infections as well as the number of secondary infections following a HCV cure. It was found that lowlevel treatment of at-risk PWID will reduce HCV prevalence, but in the absence of behavioral changes, the number of secondary infections will increase. If, however, the viral pool is depleted (HCV-RNA prevalence near zero), the number of secondary infections will also decline. In this study, treatment of > 10% of HCV+ PWID was required in order to reduce the viral pool and prevent new infections. While this concept was demonstrated across the total PWID population (general PWID population, NSP, OST, NSP & OST), it could also be applied to a closed network of injectors, in order to achieve a zero prevalence locally. Overall in Belgium, there were viremic HCV infections in As identified in this analysis, there were approximately 2970 viremic infections among PWID, suggesting the remaining infections were among individuals who formerly injected or who have no history of injecting. Without intravenous contact with HCV-infected blood, the risk of HCV transmission is assumed to be negligible. Thus, this population of individuals can be treated without an immediate increased risk of secondary infections from injecting. This includes individuals who formerly injected who are engaged in OST and abstaining from injecting, as well as substance users who have switched to other methods of use (including smoking, snorting, etc.). Furthermore, a subset of the HCV-infected PWID population refrains from sharing injection equipment and is at a negligible risk of transmitting HCV or being infected. If, however, a formerly injecting individual relapses into injecting or new sharing behaviors are adopted, that individual would again be at risk of secondary infection. Thus, expanding harm reduction efforts to reduce the number of unsafe injections is a key measure to ensure the prevention of new HCV infections. The ability to model and estimate the size of the PWID population by segment does not necessarily reflect the challenges associated with expanding harm reduction services or treatment to a hard-to-reach population. There is currently a surplus of individuals on OST who no longer inject, who will need to be treated. Capacity to treat could become an issue if there are insufficient resources (human or otherwise) to follow up with these patients. Among active injectors, a variety of factors, including housing instability, psychiatric co-morbidities and chaotic injecting behaviors may preclude HCV treatment. Thus, the primary value of this exercise is in the ability to model the dynamics of the population and run what-if scenarios. Understanding the current size of the at-risk as well as the low-risk populations allows for budgeting and goal setting, and will provide a baseline for future efforts.

5 HCV among PWID in Belgium 231 Fig. 4. Cumulative secondary infections and treated patients, by percent of the HCV+ PWID population treated, Limitations The scenarios modeled here reflect treatment across all segments of the population proportionally (i.e., 4% of the general PWID, NSP, OST and NSP & OST populations) and do not consider the feasibility of providing treatment within each of these groups. Additionally, the model is unable to account for short-term changes in an individual s injecting status. After presenting to a center for an OST prescription, it may take as little as 2 weeks for an OST client to find a stable dose of methadone ; however, for the sake of the model, the individual will be counted as either in OST or in the general PWID population for a full year. On an individual level, this may represent a substantial difference in risk ; however, aggregated to a population level, the average risk should not be substantially impacted. Another limitation is that new infections were not modeled to occur in prison. PWID in prison were, however, eligible for treatment. In 2014, individuals were imprisoned (23), with approximately 8000 (68%) incarcerated for longer than one year (24). The HCV prevalence in the general prison population was 10-15% ; however, HCV prevalence was expected to be higher among incarcerated PWID (Expert Consensus). In 2015, approximately 1.7% of the HCV+ PWID population was incarcerated, or an estimated 340 incarcerated PWID. Furthermore, the risk of HCV infection was considered negligible among PWID who had discontinued injecting or who were injecting but not sharing needles. This approach may have underestimated the number of new cases occurring ; however, given the lower risk of transmission, it is unlikely that the number of new cases would be substantially increased. Transmission risk was limited to actively injecting PWID who were sharing needles because the availability of data for other populations was limited. Conclusion The primary insights from this analysis are as follows 1) Expanding harm reduction services to reach more users will decrease the risk of new infections ; 2) A majority of OST clients either are no longer injecting or are injecting safely without sharing needles. Per our analysis, these individuals have no contraindications for treatment on account of their injecting history, and can be treated with negligible risk of secondary infection ; 3) Treatment of greater than 10% of the HCV-infected PWID population will reduce the HCV prevalence among PWID by 99% and will nearly eliminate new infections by 2030 ; 4) An aggressive up-front treatment strategy in which 12.5% of the PWID population is treated would require a greater initial investment in treatment than treating 10% of the population ; however, the cumulative number of secondary infections by 2030 will be lower due to a more rapid depletion of the viral pool. References 1. GREBELY J., ROBAEYS G., BRUGGMANN P., AGHEMO A., BACKMUND M., BRUNEAU J., BYRNE J. et al. Recommendations for the management of hepatitis C virus infection among people who inject drugs. Int. J. Drug Policy, 2015, 26 : BOLLAERTS K., AERTS M., SASSE A. Improved benchmark-multiplier method to estimate the prevalence of ever-injecting drug use in Belgium, Arch. Public Health, 2013, 71 : PLETTINCKX E., ANTOINE J., BLANCKAERT P., DE RIDDER K., VANDER LAENEN F., LAUDENS F., CASERO L. et al. Belgian National Report on drugs 2014, New Developments and Trends. Brussels, Belgium, EUROPEAN MONITORING CENTRE FOR D, DRUG A. Country overview : Belgium. Key statistics on the drug situation in Belgium, DOERRBECKER J., BEHRENDT P., MATEU-GELABERT P., CIESEK S., RIEBESEHL N., WILHELM C., STEINMANN J. et al. Transmission of hepatitis C virus among people who inject drugs : viral stability and association with drug preparation equipment. J. Infect. Dis., 2013, 207 :

6 232 C. Matheï et al. 6. RAZAVI H., ELKHOURY A.C., ELBASHA E., ESTES C., PASINI K., POYNARD T., KUMAR R. Chronic hepatitis C virus (HCV) disease burden and cost in the United States. Hepatology, 2013, 57 : RAZAVI H., WAKED I., SARRAZIN C., MYERS R.P., IDILMAN R., CALINAS F., VOGEL W. et al. The present and future disease burden of hepatitis C virus (HCV) infection with today s treatment paradigm. J. Viral. Hepat., 2014, 21 Suppl 1 : VANDIJCK D., MORENO C., STARKEL P., VAN DAMME P., VAN VLIERBERGHE H., HINDMAN S.J., RAZAVI H. et al. Current and future health and economic impact of hepatitis C in Belgium. Acta Gastroenterol. Belg., 2014, 77 : CENTERS FOR DISEASE C., PREVENTION. Recommendations for follow-up of health care workers after occupational exposure to hepatitis C virus. MMWR, 1997, 46 : MITSUI T., IWANO K., MASUKO K., YAMAZAKI C., OKAMOTO H., TSUDA F., TANAKA T. et al. Hepatitis C virus infection in medical personnel after needlestick accident. Hepatology, 1992, 16 : HASAN F., ASKAR H., AL KHALIDI J., AL SHAMALI M., AL KALAOUI M., AL NAKIB B. Lack of transmission of hepatitis C virus following needlestick accidents. Hepatogastroenterology, 1999, 46 : TSUI J.I., EVANS J.L., LUM P.J., HAHN J.A., PAGE K. Association of opioid agonist therapy with lower incidence of hepatitis C virus infection in young adult injection drug users. JAMA Intern. Med., 2014, 174 : TURNER K.M., HUTCHINSON S., VICKERMAN P., HOPE V., CRAINE N., PALMATEER N., MAY M. et al. The impact of needle and syringe provision and opiate substitution therapy on the incidence of hepatitis C virus in injecting drug users : pooling of UK evidence. Addiction, 2011, 2011/05/28 : VAN DEN BERG C., SMIT C., VAN BRUSSEL G., COUTINHO R., PRINS M., AMSTERDAM C. Full participation in harm reduction programmes is associated with decreased risk for human immunodeficiency virus and hepatitis C virus : evidence from the Amsterdam Cohort Studies among drug users. Addiction, 2007, 102 : WHITE B., DORE G.J., LLOYD A.R., RAWLINSON W.D., MAHER L. Opioid substitution therapy protects against hepatitis C virus acquisition in people who inject drugs : the HITS-c study. Med. J. Aust., 2014, 201 : HAGAN H., POUGET E.R., DES JARLAIS D.C. A systematic review and meta-analysis of interventions to prevent hepatitis C virus infection in people who inject drugs. J. Infect. Dis., 2011, 204 : MACARTHUR G.J., VAN VELZEN E., PALMATEER N., KIMBER J., PHARRIS A., HOPE V., TAYLOR A. et al. Interventions to prevent HIV and Hepatitis C in people who inject drugs : a review of reviews to assess evidence of effectiveness. Int. J. Drug Policy, 2014, 25 : BMCDDA. Belgian national report on drugs, Brussels, Belgium : Belgian Monitoring Center for Drugs and Drug Addiction, Report No. : D/2013/ GOWER E., ESTES C., BLACH S., RAZAVI-SHEARER K., RAZAVI H. Global epidemiology and genotype distribution of the hepatitis C virus infection. Journal of Hepatology, 2014, 61 : S45- S VAN DAMME P., LALEMAN W., STARKEL P., VAN VLIERBERGHE H., VANDIJCK D., HINDMAN S.J., RAZAVI H. et al. Hepatitis C Epidemiology in Belgium. Acta Gastroenterol. Belg., 2014, 77 : EMCDDA. High risk drug use- Injecting - % injecting in treatment : European Monitoring Centre for Drugs and Drug Addiction, MATHEI C., SHKEDY Z., DENIS B., KABALI C., AERTS M., MOLENBERGHS G., VAN DAMME P. et al. Evidence for a substantial role of sharing of injecting paraphernalia other than syringes/needles to the spread of hepatitis C among injecting drug users. J. Viral. Hepat., 2006, 13 : BELGIUM S. Prison Population. In. Statistics Belgium, INTERNATIONAL CENTRE FOR PRISON S. Belgium - Pre-trial/remand prison population : trend, We would like to ask you to check on the author s names, as this is how they are going to be published on PubMed and in the contents! Mitigating the burden of hepatitis C virus among people who inject drugs in Belgium. C. Matheï, S. Bourgeois, S. Blach, C. Brixko, J.P. Mulkay, H. Razavi, G. Robaeys