Hepatitis C burden falls sharply among French drug users as antivirals scale up

A two-decade surveillance effort in France has documented one of the clearest national pictures yet of how hepatitis C virus (HCV) transmission among people who use drugs responds to treatment and prevention policy. In a study published in Nature Communications, a team led by Chiara E. Sabbatini of Santé publique France analyzed three cross-sectional surveys conducted in 2004, 2013 and 2023, recruiting people who use drugs through harm reduction services across the country. The results show a substantial decline in both the prevalence of active infection and the rate of new infections between 2013 and 2023, a period that coincided with the nationwide rollout of direct-acting antiviral (DAA) therapies and sustained investment in harm reduction. Yet the findings also make clear that transmission has not stopped, and that the World Health Organization’s 2030 elimination targets remain out of reach without renewed, targeted interventions.

The study rests on the ANRS-Coquelicot survey series, a rare longitudinal framework in which comparable biological and behavioral data are collected at roughly ten-year intervals. Participants were recruited at harm reduction and addiction care structures, and the researchers gathered both questionnaire data on drug use practices and biological samples that allowed them to test for HCV antibodies, which indicate past exposure, and for HCV RNA, which signals current, potentially transmissible infection. This dual measurement is technically important: antibody prevalence reflects the cumulative history of infection in the population, whereas RNA prevalence captures the share of people who remain infected and capable of transmitting the virus. Distinguishing the two is essential for judging whether a decline reflects cured infections, prevented infections, or both.

To extract trends from the three survey waves, the team employed multivariate regression models that estimated the prevalence of HCV antibodies and HCV RNA as functions of both age and calendar time. This approach matters because the population of people who use drugs is not static; it ages, turns over, and changes in composition. If the population is simply getting older, crude prevalence figures can rise or fall for reasons that have nothing to do with transmission dynamics. By modeling age and time simultaneously, the researchers could separate the demographic aging of the population from genuine epidemiological change. Incidence, the rate of new infections per unit of time, was estimated using two-state transmission models, which infer the force of infection from the pattern of antibody positivity across age and time.

The demographic picture that emerged is itself striking. The mean age of survey participants rose from 35 years in 2004 to 44 years in 2023, reflecting the aging of an established cohort of people who inject or have injected drugs, alongside changes in who is entering harm reduction services. Meanwhile, the proportion of participants reporting injection remained stable, standing at 61.7 percent in 2023. This combination, an older population with a persistently high share of injectors, means that the population studied remains one in which blood-borne transmission risk is structurally embedded, even as the overall viral burden has fallen. The stability of injection prevalence also suggests that the observed declines in HCV metrics cannot be attributed simply to fewer people injecting.

The headline epidemiological result concerns the decade between 2013 and 2023. Among all people who use drugs surveyed, the prevalence of HCV RNA, the marker of active infection, fell from 19.4 percent (95 percent confidence interval 16.3 to 23.0) to 9.5 percent (7.2 to 12.3), a reduction of more than half. Among the subset who inject drugs, the decline was even steeper in absolute terms: active infection prevalence dropped from 29.6 percent to 14.5 percent. These figures indicate that by 2023, roughly one in seven people who inject drugs in France still carried detectable HCV, down from nearly one in three a decade earlier. Because RNA positivity tracks current infection rather than past exposure, this drop points to both successful treatment of existing infections and a slowing of new ones.

The incidence estimates reinforce that interpretation. Among people who inject drugs, the estimated rate of new HCV infections fell from 6.5 per 100 person-years (95 percent confidence interval 4.1 to 8.0) in the earlier period to 2.4 per 100 person-years (1.4 to 4.2) by 2023. In practical terms, this means that in the earlier era, roughly six to seven new infections occurred for every hundred people who inject drugs followed for a year, whereas by the most recent survey that figure had fallen to between one and four. The authors attribute these reductions to the combined impact of DAA scale-up, which since the mid-2010s has allowed oral, well-tolerated cures for the vast majority of treated patients, and France’s long-standing harm reduction policies, including needle and syringe programs and opioid agonist maintenance treatment.

The mechanistic logic behind this dual intervention effect is well established in viral epidemiology. HCV transmission among people who inject drugs depends on the probability that a shared injection instrument contacts the blood of someone with active infection. Treating infected individuals rapidly reduces the prevalence of RNA-positive people in the network, lowering the chance that any given sharing event exposes someone to the virus, an effect often described as treatment as prevention. Harm reduction tools independently reduce the frequency of blood contact. When both operate together, the reproduction number of the virus can fall below one in parts of the network, producing the sustained declines documented here. France’s decision to remove restrictions on DAA access, making treatment available regardless of liver disease stage or ongoing drug use, positioned it to realize this effect at population scale.

Nevertheless, the study’s authors emphasize that ongoing transmission persists, and the numbers bear this out. An incidence of 2.4 per 100 person-years among people who inject drugs remains far above the levels implied by the WHO’s elimination framework, which calls for an 80 percent reduction in new infections by 2030 relative to a 2015 baseline. The persistence of transmission despite broad treatment access suggests that barriers remain that antivirals alone cannot overcome. These may include undiagnosed infection among people not in contact with services, reinfection after successful cure among those with continued exposure, uneven geographic coverage of testing and treatment, and populations, such as younger or more marginalized drug users, who are less reached by existing harm reduction and care structures. The aging of the surveyed population also raises questions about whether newer cohorts of drug users are being captured and served effectively.

From a methodological standpoint, the Coquelicot series offers lessons for other countries pursuing elimination. Repeated, nationally coordinated bio-behavioral surveys anchored in harm reduction settings provide direct, empirically grounded estimates of both prevalence and incidence, rather than relying on modeled extrapolations from case notifications, which notoriously undercount infections in this population because many cases are asymptomatic and never clinically detected. The use of age-period models and two-state transmission modeling demonstrates how surveillance data can be converted into actionable epidemiological parameters. For policy makers, the key indicators to monitor are RNA prevalence, which measures the infectious reservoir, and incidence, which measures whether that reservoir is still generating new infections; the French data show both moving in the right direction but neither approaching zero.

The overall message is one of substantial but incomplete progress. Twenty years of surveillance show that a combination of curative therapy and prevention infrastructure can halve the burden of active hepatitis C among people who use drugs, a population that has historically borne the heaviest share of HCV in high-income countries. The decline from 19.4 percent to 9.5 percent RNA prevalence among all surveyed drug users, and from 6.5 to 2.4 new infections per 100 person-years among injectors, represents thousands of infections averted or cured. But with one in fourteen surveyed drug users still actively infected and transmission continuing at a measurable rate, the final stretch toward the 2030 elimination goals will require precisely the kind of targeted public health action the authors call for: expanded testing, rapid re-treatment of reinfections, and strengthened reach into the populations that current services still miss.

Subject of Research: Trends in hepatitis C virus incidence and prevalence among people who use drugs in France from 2004 to 2023

Article Title: HCV incidence and prevalence trends among people who use drugs in France 2004 to 2023

Article References: Sabbatini, C. E., Le Strat, Y., Bonaldi, C., Brouard, C., Audran, M., Saboni, L., Richard, J.-B., David, A., Gaudy-Graffin, C., Chevaliez, S., Barin, F., Stefic, K., & Jauffret-Roustide, M. (2026). HCV incidence and prevalence trends among people who use drugs in France 2004 to 2023. Nature Communications. https://doi.org/10.1038/s41467-026-78598-3

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78598-3

Keywords: hepatitis C, HCV incidence, HCV prevalence, people who inject drugs, direct-acting antivirals, harm reduction, France, epidemiology, ANRS-Coquelicot survey, WHO elimination targets, public health surveillance, transmission modeling