Clinical trials
traditionally report outcomes using an intention-to-treat analysis,
meaning
final outcomes for all participants are included. If people who were
cured and reinfected are counted as treatment failures, therapy can
appear less effective
than it really is.
The multinational C-EDGE CO-STAR trial enrolled 301 people with previously
untreated chronic hepatitis C, with HCV genotypes 1, 4 or 6. About
three-quarters were men, most were white, the median age was 48 years, 21% had
liver cirrhosis and 7% had HIV co-infection. Participants were required to be
on stable OST using methadone or buprenorphine for at least three months at
study entry, but they were not excluded if urine tests showed they continued to
use illicit drugs during the study, as many of them did.
Participants were randomly assigned to receive either
the grazoprevir/elbasvir co-formulation or a placebo once daily for 12 weeks
(immediate treatment arm), at which point the study was unblinded and placebo
recipients were also given active treatment on an open-label basis (deferred
treatment arm).
As recently reported, grazoprevir/elbasvir was well
tolerated and cured 92 to 94% of study participants with HCV genotypes 1 or 4,
though the SVR rate fell to just 20% for the small number of people with
genotype 6.
Based on viral population sequencing and phylogenetic analysis,
there were five probable reinfections after achieving undetectable viral load
in the immediate treatment group at both 12 and 24 weeks post-treatment, and one
apparent late reinfection occurring between post-treatment weeks 12 and 24 in
the deferred treatment group.
Interestingly, three of the six people with probable
reinfection cleared the recurrent virus without further treatment. This is
consistent with other research suggesting that the immune system gets better
at controlling HCV with repeated attempts. But this does not mean that reinfection
is benign, as reinfection in someone
with cirrhosis on the verge of decompensation could be a large risk.
Reinfection incidence rates were 10.5
per 100 PY for the immediate treatment group through follow-up week
12, and 4.6 per 100 PY for the immediate and deferred groups combined through
follow-up week 24. Excluding those with spontaneous clearance, the chronic or
persistent reinfection rate fell to around 3%, Dr Dalgard said.
If these people with reinfection were
counted as treatment successes, this would push SVR12 rates up to 96% for the
immediate treatment group and 97% for the deferred group (SVR24 94 and 97%,
respectively) – and up to 60% for genotype 6. Although only nine study
participants had genotype 6, reinfection occurred in two of them, both men in Thailand.
“Further follow-up is required to determine the
natural course of HCV reinfection in the setting of interferon-free HCV
treatment and the impact of viral persistence following reinfection on
long-term response rates in this population,” the researchers concluded in
their study abstract.
“We have to accept that reinfection will happen after
treatment to some extent,” Dr Dalgard said.
Dr Midgard talked more about the consequences of reinfection,
noting that even low rates of reinfection could be a concern over time,
reducing individual and population-level treatment benefits, changing
cost-effectiveness calculations, and having implications for hepatitis C
prevention and harm reduction strategies. Slow treatment scale-up among people
who inject drugs, for example, could create a growing pool of people
susceptible to reinfection.
Dr Midgard ended by stressing the importance of
acknowledging the occurrence of HCV reinfection without stigma and
discrimination, which could drive people away from care. Education and counselling
about the possibility of reinfection is needed, as are ongoing post-treatment
surveillance and harm reduction efforts.
“If reinfections don't occur, you are treating the
wrong population,” he concluded.