Combination interventions to prevent HCV transmission among people who inject drugs: modeling the impact of antiviral treatment, needle and syringe programs, and opiate substitution therapy

Abstract

Background: Interventions such as opiate substitution therapy (OST) and high-coverage needle and syringe programs (HCNSP) cannot substantially reduce hepatitis C virus (HCV) prevalence among people who inject drugs (PWID). HCV antiviral treatment may prevent onward transmission. We project the impact of combining OST, HCNSP, and antiviral treatment on HCV prevalence/incidence among PWID.

Methods: An HCV transmission model among PWID was used to project the combinations of OST, HCNSP, and antiviral treatment required to achieve different prevalence and incidence reductions within 10 years for 3 chronic prevalence scenarios and the impact of HCV treatment if only delivered through OST programs. Multivariate and univariate sensitivity analyses were performed.

Results: Large reductions (>45%) in HCV chronic prevalence over 10 years require HCV antiviral treatment. Scaling up OST and HCNSP substantially reduces the treatment rate required to achieve specific HCV prevalence reductions. If OST and HCNSP coverage were increased to 40% each (no coverage at baseline), then annually treating 10, 23, or 42 per 1000 PWID over 10 years would halve prevalence for 20%, 40%, or 60% baseline chronic HCV prevalences, respectively. Approximately 30% fewer treatments are necessary with new direct-acting antivirals. If coverage of OST and HCNSP is 50% at baseline, similar prevalence reductions require higher treatment rates for the same OST and HCNSP coverage.

Conclusions: Combining antiviral treatment with OST with HCNSP is critical for achieving substantial reductions (>50%) in HCV chronic prevalence over 10 years. Empirical studies are required on how best to scale up antiviral treatment and combine treatment with other interventions.

Figure 1.

Combinations of annual treatment rates per 1000 injectors and coverage of opiate substitution therapy (OST) and high-coverage needle and syringe programs (HCNSP) required to reduce prevalence by 50% within 10 years. Results shown for 3 baseline chronic prevalence settings (20%, 40%, and 60%). A and B, Assumes no intervention coverage at baseline with OST and HCNSP scale-up to 0%, 20%, 40%, or 60% of each and using pegylated interferon and ribavirin (peg-IFN + RBV) (A) and interferon (IFN)-free direct-acting antivirals (DAAs) (B). C, Assumes 50% coverage of OST and HCNSP at baseline with OST and HCNSP scale-up to 50%, 60%, 70%, or 80% of each using peg-IFN + RBV. The box-plots signify the uncertainty (middle line is the median, limits of the boxes are 25% and 75% percentiles and whiskers are 2.5% and 97.5% percentiles) in the impact projections due to uncertainty in the intervention effect estimates.

Figure 2.

Contour maps of the relative reductions in prevalence (%) at 10 years with combinations of antiviral treatment (y-axis) and opiate substitution therapy/high-coverage needle and syringe program (OST and HCNSP) (x-axis) scale-up with no baseline coverage of OST, HCNSP, or treatment. Results shown for 3 baseline hepatitis C virus chronic prevalence settings (20%, 40%, and 60%) with pegylated interferon and ribavirin (pegIFN + RBV) (A–C) and IFN-free direct-acting antivirals (D–F). Projections used the median estimates for efficacy of OST, HCNSP, and peg-IFN + RBV from Table 1.

Figure 3.

Minimum coverage of opiate substitution therapy/high-coverage needle and syringe programs (OST and HCNSP) required if antiviral treatment (pegylated interferon and ribavirin [peg-IFN + RBV]) is delivered alongside OST. Figures show the minimum coverage of OST and HCNSP required (y-axis) for various desired relative prevalence reductions at 10 years (x-axis) with the 20%, 40%, and 60% baseline hepatitis C virus (HCV) chronic prevalence settings if all infected people who inject drugs (PWID) on OST are treated annually, limited by HCV prevalence (A) or 5% of PWID on OST are treated annually (B). Projections used the median estimates for efficacy of OST, HCNSP, and peg-IFN + RBV from Table 1.