Opioid substitution and syringes exchange programs have drastically reduced HCV spread in France, while HCV sexual transmission in men who have sex with men (MSM) has recently arose as a significant phenomenon. Epidemiological data such as prevalence and incidence rates can quantify an epidemic at its chronic stage but are less meaningful at its early stages or if the transmission of the pathogen only occurs in a subgroup of individuals. Phylodynamic inferences use both pathogen phylogenies based on genetic sequences and epidemiological data to describe infectious diseases transmission dynamic. We used a phylodynamic approach to estimate key epidemiological parameters such as the reproduction number (R0) and the infectious period duration of acute HCV infection (AHI) in French MSM.
A birth-death epidemiological model with 2 host types corresponding to respectively the ‘classic’ HCV epidemic (mostly IVDU-blood product recipients) and the ‘new’ epidemic in MSM was implemented. Two periods (< and >1997) were considered for the classic epidemic. 30,000 simulated phylogenies were first generated under a variety of parameter set. These simulations were then used to ‘feed’ a regression model and to infer epidemiological parameters using an approximate Bayesian computation approach. The model was then run on the true HCV phylogeny from AHI and chronic HCV infections, to infer R0, infectious period and assortativity estimates (the extent to which virus transmission is random or occurs mostly within groups) for both epidemics. The validity of the results was estimated using a parametric bootstrap approach.
213 NS5B sequences from HCV genotype 1a infections were analyzed (68 from AHI in MSM, 145 from chronic infections in non-MSM patients). Estimates of the beginning dates for the classic and the new epidemics were 1983 (95%CI 1981-1983) and 2000 (95%CI 1999-2002) respectively. Estimates of R0 for the classic epidemic >1997 and for the new epidemic were 1.5 (IQR 1.3-1.7) and 1.7 (IQR 1.4-2.1) respectively. Estimates for the infectious period duration for the classic and the new epidemics were 2.3 years (IQR 1.6-3.1 years) and 0.4 years (0.4-0.5 years) respectively.
AHI epidemic in French MSM was characterized by a similar R0, but a much shorter infectious period and a greater transmission rate per unit of time than the classic epidemic. These result shows how phylodynamic can help to understand the transmission dynamics of an epidemic spreading in different populations