This study assessed trends in HCV
incidence in the different HIV transmission groups in the Swiss HIV Cohort
Study (SHCS) in order to optimize surveillance and prevention......In
MSM, history of inconsistent condom use (aHR: 2.09, 1.33-3.29) and past
syphilis (2.11, 1.39-3.20) predicted HCV seroconversions.....IDUs are the
prevalent group with HCV in HIV+ in the Swiss Cohort but the high increasing
incident rate among HIV+ MSM (increased 18-fold in the last 13 years) reflects
a trend observed and reported in Western Europe in major cities & reported
at conferences over the past few years....."In
MSM, history of inconsistent condom use (aHR: 2.09, 1.33-3.29) and past
syphilis (2.11, 1.39-3.20) predicted HCV seroconversions.....In the SHCS, the large majority of recent HCV
infections occurred in MSM"......"suggests that 3-monthly screening with liver
function tests is costeffective in settings with HCV-incidences above 1.25%.
Accordingly, our results should provide important information for establishing
cost-effective screening programs in HIV-infected patients. However, routine
serological screening does not prevent risk behaviour, and coordinated efforts
between physicians and public-health representatives are needed to control the
ongoing HCV infection epidemic in HIV-infected MSM.....It
is crucial that HIV-infected MSM are counselled with regard to the risks of
sexual activities that involve traumatic mucosal sex, and that condoms are used
consistently in sexual risk situations [28]. The example of IDUs demonstrates
that it is possible to reduce the incidence of HCV-infections through improved
screening and preventive interventions......Accordingly, the intervals of HCV-screening will be
reduced to 1 year in the SHCS. This should allow identifying patients with new
HCV-infections earlier in order to optimize counselling with regard to HCV
transmission risks, and to improve the response to HCV-therapy by initiating
treatment during acute infection. The revised screening practice in the SHCS is
in line with recent guidelines which recommend yearly screening for all
HIV-infected MSM who still engage in any risk behavior [33-35]. Furthermore, a
recent study demonstrated that yearly serological HCV screening and 6-monthly
liver function testing is cost-effective [36]. Importantly, this analysis
demonstrated that the optimal screening strategy strongly depends on
HCV-incidence"
"It is
crucial that HIV-infected MSM are counselled with regard to the risks of sexual
activities that involve traumatic mucosal sex, and that condoms are used
consistently in sexual risk situations [28]. The example of IDUs demonstrates
that it is possible to reduce the incidence of HCV-infections through improved
screening and preventive interventions.....This underscores the need for improved HCV
surveillance and prevention among HIV-infected MSM. Accordingly, the intervals
of HCV-screening will be reduced to 1 year in the SHCS. This should allow
identifying patients with new HCV-infections earlier in order to optimize
counselling with regard to HCV transmission risks, and to improve the response
to HCV-therapy by initiating treatment during acute infection"
"Of 4,629 MSM, 2,678
IDU and 4,530 HET screened for HCV infection, 3,333 (72%), 123 (5%) and 3,078
(68%) had a negative HCV serology at baseline and available serological
follow-up. Over 23,707 person-years (py) for MSM, 733 py for IDU and 20,752 py
for HET, 101 (3%), 41 (33%) and 25 (1%) seroconverted, respectively. The IR
of HCV infections in MSM increased from 0.23 (95% CI: 0.08-0.54) per 100py in
1998 to 4.09 (2.57-6.18) in 2011. The IR decreased in IDU and remained
below 1 per 100 py in HET. In MSM, history of inconsistent condom use (aHR:
2.09, 1.33-3.29) and past syphilis (2.11, 1.39-3.20) predicted HCV
seroconversions.
Conclusions. In the SHCS, HCV incidence
decreased in IDU, remained stable in HET and increased 18-fold in MSM in the
last 13 years. These observations underscore the need for improved HCV
surveillance and prevention among HIV-infected MSM."
Most
demographic and behavioural characteristics were similar in HCV-seroconverters
compared to patients who did not seroconvert (Web-Table 1). Median age was 34.9
years in seroconverters and 37.7 years in non-seroconverters, and the
proportion of Caucasians was 93% and 92%, respectively. In both groups, over
90% of patients started ART before or during follow-up. However, previous HBV
exposure, a past history of syphilis and inconsistent condom use were more
common in HCV-seroconverters (Web-Table 1). The incidence of HCV infections was
higher in patients who reported inconsistent condom use, as well as in those
who had a previous episode of syphilis or exposure to HBV compared to the other
patients (Figure 3). In adjusted Cox regression analyses, only inconsistent
condom use (adjusted Hazard Ratio (aHR): 2.09,
95% CI 1.33-3.29) and previous diagnosis of syphilis (2.11,
1.39-3.20) were significantly associated with HCV seroconversions (Table
2).
In the SHCS, the large majority
of recent HCV infections occurred in MSM. This underscores the need for
improved HCV surveillance and prevention among HIV-infected MSM. Accordingly,
the intervals of HCV-screening will be reduced to 1 year in the SHCS. This
should allow identifying patients with new HCV-infections earlier in order to
optimize counselling with regard to HCV transmission risks, and to improve the
response to HCV-therapy by initiating treatment during acute infection. The
revised screening practice in the SHCS is in line with recent guidelines which
recommend yearly screening for all HIV-infected MSM who still engage in any
risk behavior [33-35]. Furthermore, a recent study demonstrated that yearly
serological HCV screening and 6-monthly liver function testing is
cost-effective [36]. Importantly, this analysis demonstrated that the optimal
screening strategy strongly depends on HCV-incidence.
For instance, it suggests that 3-monthly
screening with liver function tests is costeffective in settings with
HCV-incidences above 1.25%. Accordingly, our results should provide important
information for establishing cost-effective screening programs in HIV-infected
patients. However, routine serological screening does not prevent risk
behaviour, and coordinated efforts between physicians and public-health
representatives are needed to control the ongoing HCV infection epidemic in
HIV-infected MSM. Although most reported HCV-transmissions among MSM occurred
in HIV-infected patients, it is conceivable that HCV-incidence among
HIV-negative MSM is underestimated as these individuals are not in regular
medical care. Clinicians and patients should be aware of the risk of acute HCV
infection in MSM, and intensified prevention and counselling should be
performed. Despite the fact that antiretroviral treatment is highly efficient
in preventing HIV-transmission it does not have any effect on prevention of
other sexually transmitted diseases. Condom use declined in recent years in MSM
with suppressed HIV viral load in the SHCS [37], which likely contributes to
the increasing incidence of HCV-infections in this population. It is crucial
that HIV-infected MSM are counselled with regard to the risks of sexual
activities that involve traumatic mucosal sex, and that condoms are used consistently in sexual risk
situations [28]. The example of IDUs demonstrates that it is possible to reduce
the incidence of HCV-infections through improved screening and preventive
interventions.
Abstract
Background.  Hepatitis C Virus (HCV)
infection has a growing impact on morbidity and mortality in HIV-infected
patients. We assessed trends in HCV incidence in the different HIV transmission
groups in the Swiss HIV Cohort Study (SHCS) in order to optimize surveillance
and prevention.
Methods. HCV infection incidence was
assessed from 1998, when routine serial HCV screening was introduced in the
SHCS, until 2011. All HCV-seronegative patients with at least one follow-up
serology were included. Incidence rates (IR) of HCV infections were compared
between men who have sex with men (MSM), injection drug users (IDU) and
heterosexuals (HET). Predictors of HCV seroconversion in MSM were estimated.
Results. Of 4,629 MSM, 2,678 IDU and
4,530 HET screened for HCV infection, 3,333 (72%), 123 (5%) and 3,078 (68%) had
a negative HCV serology at baseline and available serological follow-up. Over
23,707 person-years (py) for MSM, 733 py for IDU and 20,752 py for HET, 101
(3%), 41 (33%) and 25 (1%) seroconverted, respectively. The IR of HCV
infections in MSM increased from 0.23 (95% CI: 0.08-0.54) per 100py in 1998 to
4.09 (2.57-6.18) in 2011. The IR decreased in IDU and remained below 1 per 100
py in HET. In MSM, history of inconsistent condom use (aHR: 2.09, 1.33-3.29)
and past syphilis (2.11, 1.39-3.20) predicted HCV seroconversions.
Conclusions. In the SHCS, HCV incidence
decreased in IDU, remained stable in HET and increased 18-fold in MSM in the
last 13 years. These observations underscore the need for improved HCV
surveillance and prevention among HIV-infected MSM.
Hepatitis C Virus (HCV) infection is a
major cause of morbidity and mortality in HIV-infected patients [1-3]. For many
years, HCV-infections occurred almost exclusively in injection drug users (IDU)
or hemophiliacs. In a previous analysis in the Swiss HIV Cohort study
(SHCS), HCV seroprevalence was 33% overall and 90% among individuals reporting
IDU [4]. Recently, there have been several outbreaks of HCV infections
among HIV-infected men who have sex with men (MSM), predominantly in large
cities in Europe, Australia and the USA [5-14]. These clusters of HCV
infections involved high-risk sexual practices within confined social networks.
However, the impact of these localized epidemics on the incidence of HCV
infections in general HIV-infected populations and the long-term trends in HCV
incidence rates in different transmission risk groups are largely unknown. A
better understanding of changes in HCV-incidence rates and transmission
patterns is urgent because of the increasing burden of HCV-related disease in
HIV-infected patients [1-3].
The SHCS offers an ideal platform for
studying changes in HCV incidence in a general HIV-infected population and in
diverse HIV-transmission risk groups, as all HCV seronegative patients are
screened routinely at baseline and during follow-up since 13 years.
Furthermore, it allows a thorough evaluation of the association of HCV
seroconversions with different risk factors, as detailed patient history and
clinical data is collected every six months.
We assessed changes in the HCV infection
incidence in the three main HIV transmission groups (IDU, MSM and heterosexuals
(HET)) in the SHCS during the last 13 years, with the expectation that our
results would shed light on the magnitude and changes of the HCV epidemic in a
general and representative HIV-infected population.
METHODS
Swiss HIV cohort Study
The Swiss HIV Cohort Study (SHCS, www.shcs.ch) is a prospective cohort study with
ongoing enrolment of HIV-infected adults in Switzerland since 1988. It covers
at least 45% of the cumulative number of HIV infections declared to the Swiss
public health authorities, 69% of all patients living with AIDS and 75% of
patients receiving antiretroviral therapy in Switzerland [15]. Representativity
remained stable over the years. Detailed information on demographics, mode of
HIV acquisition, risk behavior, clinical events, co-infections, and treatment
is collected using a standard protocol at registration and then at intervals of 6 months. Local
ethical committees of all participating study sites have approved the study and
written consent has been obtained from all participants.
Inclusion criteria and
definitions
To compare HCV incidence between risk
groups, we categorised patients as IDU, MSM or HET according to the most
probable HIV-transmission mode. In order to minimize erroneous classifications,
we excluded MSM and HET patients who reported the use of injection drugs at any
occasion during follow-up. Since 1998, all SHCS patients have been routinely
screened for HCV infection: serology is performed every second year of
follow-up, independently of HIV transmission mode. Thus, all HCV-seronegative
patients who had at least one follow-up HCV antibody measurement after July
1998 were included in the analyses. Positive results by third-generation ELISA
are confirmed by immunoblot. Patients with a positive HCV serology at entry,
and those who seroconverted before 1998 were excluded. Individual follow-up
ended at the time of the first positive or last negative HCV serology. A
detailed patient flow chart is shown in Figure 1. Previous syphilis was defined
as a positive Treponema-pallidumhemagglutination-assay (TPHA) screening test
and hepatitis B virus (HBV) exposure by the presence of a positive anti-HBc
antibody test prior to HCV seroconversion, or before
the last negative HCV-serology in those
without HCV seroconversion. In this context, the term HBV exposure refers to
both resolved and chronic HBV infections. A subset of HCV seroconverters was
found to have a unique positive HCV serology followed by one or several
negative serologies. These patients could have cleared HCV antibodies after
spontaneous HCV clearance, as described before [16-18]. Alternatively, false
positive results cannot be excluded despite confirmation by HCV-immunoblot.
Therefore, the main analyses were repeated after exclusion of these
cases.
Statistical analyses
HCV incidence rates were assessed from
introduction of routine HCV-screening from July 1998 until November 2011.
Patients were considered to be at risk for HCV infection since the date of
their first negative HCV serology or since 01.07.1998 if the first measurement
preceded this date. Individual follow-up ended at the time of the first
positive or last negative HCV serology. Baseline characteristics were compared
between the three risk groups using analysis of variance (ANOVA) and
chi-squared tests for continuous and categorical variables, respectively. To
calculate the incidence of HCV infections in the different risk groups, events
were right-censored so that the date of seroconversion corresponded to the date
of the first positive HCV serology. Yearly HCV infection incidence rates (IR)
were obtained using Bayesian Poisson regression models with non-parametric
smoothing priors [19]. IR are reported as number of cases per 100 person-years
(py) and shown on a log scaled graph, by transmission group (MSM, IDU,
HET).
In sensitivity analyses, the yearly rates
of HCV infection in MSM were compared to the results obtained using two
alternate censoring approaches: (1) Mid-point censoring, in which the date of
seroconversion was set to the mid-point between the last negative and first
positive HCV serology, and (2) a non-parametric maximum likelihood estimation
method for interval-censored data [20]. Interval-censoring takes into account
that the exact date of seroconversion is not known but lies between the last
negative and the first positive serology. However, due to the low number of
seroconversions in IDU and HET, this method could only be applied to the MSM
risk group.
Demographic characteristics of
HCV-seroconverters in MSM were compared to those who did not seroconvert during
follow-up using Chi-square and Mann-Whitney tests. The difference in incidence
of HCV infections between patients of different age groups, condom use
patterns, hepatitis B exposure status and history of past syphilis were also
shown in log-scaled graphs. Risk factors for HCV infection in MSM were
evaluated using a multivariable Cox regression model. Analyses were adjusted
for age category (16-29, 30-39 and > =40 years), CD4 count category before
last HCV serology (<200, 200-499 and >=500 cells/ μl), education level
(no, basic and high-level professional education), region of SHCS follow-up
(Zurich or other), ethnicity (Caucasian or other), use of non-IDU drugs (yes or
no), sexual relationships (stable or occasional partners), condom use (always
or inconsistent), past history of syphilis (yes or no), HBV exposure (yes or
no) and antiretroviral therapy (ART) status (yes or no). All statistical
analyses were performed using Stata 12 (Stata Corp, College Station, USA) and
R-2.14.1 [21] using add-on packages INLA [22] and Icens [23].
RESULTS
Baseline
characteristics
Of 4,629 MSM, 2,678 IDU and 4,530 HET
screened for HCV infection, 147 (3.2%), 2,468 (92.2%) and 513 (11.3%) had a
positive HCV serology at baseline (Figure 1). 6,534 patients, of which 3,333
MSM, 123 IDU and 3,078 HET, were included in the HCV incidence analyses. Median
age at first HCV screening test was slightly higher in MSM (38 years,
inter-quartile range (IQR) 32-44) and HET (36, 30-44), compared to IDU (33,
28-40), whereas MSM were more likely to have a high-level education than
patients in the two other groups (Table 1). Almost half of the MSM and IDU, but
only 29% of HET had their regular medical follow-up in Zurich, Switzerland’s
largest urban center. Finally, HET were much more likely to be non-Caucasians
(40.7%), compared to MSM (7.7) and IDU (8.9). HCV incidence
Over a total follow-up time of 23,707
person-years (py) for the MSM group, 733 py for the IDU group and 20,752 py for
the HET group, 101 (3.0%), 41 (33.3%) and 25 (0.8%) patients experienced an HCV
seroconversion during follow-up, respectively. Between 1998 and 2011, the IR of
HCV infections in MSM increased from 0.23 (95% CI: 0.08-0.54) per 100 py to
4.09 (2.57-6.18) in 2011in2011, with 51 cases observed in the last 3 years
(Figure 2). There was a similar increase in patients followed in Zurich
compared to the rest of the country (IR in 2011 were 2.61 (1.19-5.13) and 4.10
(2.18-7.16) per 100 py, respectively). Increases in HCV-incidence in MSM were
similar when right-censoring, midpoint-censoring or interval-censoring were
used to estimate yearly incidence rates (Web-Figure 1).
In IDU, HCV infection IR decreased from
13.89 (95% CI: 8.20-22.39) per 100 py in 1998 to 2.24 (0.55-10.66) in 2011,
with only 3 incident cases in the last 3 years. For comparison, the yearly
IR of HCV infections in the HET group remained below 0.5 per 100 py, with only
2 cases in 2011 (IR 0.43 per 100py, 95% CI: 0.12-1.29; Figure 2). Twenty-five
(15.0%) HCV-seroconverters had a unique positive serology followed by negative
tests. Of those, six patients had their HCV infection confirmed by HCV RNA
testing within one month of the positive serology. In another eight patients,
HCV RNA testing was negative,, which may be reflective of spontaneous HCV
clearance and loss of HCV-antibodies over time. Eleven patients did not have an
HCV RNA available to confirm the HCV infection. After exclusion of these
patients, the incidence rates of HCV infection remained similar to the results
described above (1998: MSM: 0.11 per 100 py, 95% CI 0.03-0.35; IDU: 14.36,
8.36-23.39; HET: 0.09, 0.02-0.35. 2011: MSM: 3.56, 2.19-5.53; IDU: 2.42,
0.51-13.36; HET: 0.47, 0.14-1.30).
Predictors of HCV infections in
MSM
Most demographic and behavioural
characteristics were similar in HCV-seroconverters compared to patients who did
not seroconvert (Web-Table 1). Median age was 34.9 years in seroconverters and
37.7 years in non-seroconverters, and the proportion of Caucasians was 93% and
92%, respectively. In both groups, over 90% of patients started ART before or
during follow-up. However, previous HBV exposure, a past history of syphilis
and inconsistent condom use were more common in HCV-seroconverters (Web-Table
1). The incidence of HCV infections was higher in patients who reported
inconsistent condom use, as well as in those who had a previous episode of
syphilis or exposure to HBV compared to the other patients (Figure 3). In
adjusted Cox regression analyses, only inconsistent condom use (adjusted Hazard
Ratio (aHR): 2.09,
95% CI 1.33-3.29) and previous diagnosis of syphilis (2.11,
1.39-3.20) were significantly associated with HCV seroconversions (Table
2). Age, education level, CD4+
T cell counts, ART status and the other variables included in the
multivariable model were not significantly associated with the main outcome. Of
note, having occasional sexual partners (aHR 0.97,
0.56-1.68) and being followed-up in Zurich (aHR 1.30,
0.87-1.97) did not predict HCV seroconversions.
Of the 63 (62%) HCV infections in MSM with
detectable HCV RNA and available genotyping results, the majority were caused
by HCV genotype 1 (42 cases, 66.7%), followed by genotypes 4 (12, 19.0%), 3 (8,
12.7%) and 2 (1, 1.6%).
DISCUSSION
We assessed trends in HCV incidence
between 1998 and 2011 in the main HIV transmission groups in the SHCS. In a
population of over 6,500 patients screened every 2 years for HCV infection, we
found that the yearly incidence rate had decreased in IDU, remained stable in
HET and dramatically increased in MSM. In the latter sub-population, 50% of all
HCV infections occurred in the last three years. In MSM, a history of
inconsistent condom use and a past episode of syphilis were significantly
associated with HCV seroconversion.
Since 2005, several local reports
described epidemics of HCV infections among HIV-infected MSM [7-12].
Phylogenetic and socio-demographic analyses have shown that these infections
occurred within confined groups of MSM with high-risk sexual behavior in large
urban centers [11, 24]. Therefore, incidence estimates in these cohorts were
not representative for the general HIV-infected population, in contrast to our
study. The systematic, nationwide screening of all HIV-infected patients at
risk since 1998 allowed us to accurately estimate the trends in HCV-incidence
among different risk groups. Since 1998, the incidence of HCV infection among
MSM increased 18-fold, reaching 4.1 cases per 100 py in 2011. This incidence
was higher compared to the large majority of previous studies among
HIV-infected MSM reporting estimates lower than one case per 100 py [5].
However, the most recent HCV-incidence in the SHCS was similar to estimates
from an international cohort collaboration of HIV-seroconverters, where
incidence estimates ranged from 2.3 to 5.1/100 py in 2007 [25]. In accordance
to previous clusters of HCV-infections in large cities, the HCV-incidence
increased markedly among individuals treated in Zurich, the largest urban
center in Switzerland which, to some extent, is comparable to other large
cities in Europe and the US. However, our study shows that HCV-incidence
increased similarly outside of the Zurich area, indicating a nationwide
increase in HCV-infections in HIV-infected MSM.
In line with previous reports [4, 8,
11], unprotected anal sex was an important risk factor for HCV infections in
MSM. In the SHCS, inconsistent condom use doubled the risk of HCV infection in
MSM. Furthermore, a previous syphilis infection independently increased the
risk of an HCV seroconversion approximately twofold. Of note, we recently
observed a significant increase in syphilis acquisition among MSM in the SHCS,
paralleling the current HCV epidemic [26]. The shared route of transmission
of the two infections is the most likely explanation for this association.
Alternatively, it is also conceivable that the mucosal disruption caused by
syphilitic ulcers facilitates HCV infections. These findings underscore the
importance of the link between high risk sexual behavior and HCV transmissions.
As yearly syphilis screening and baseline HBV serology testing are routine
for all MSM in the SHCS, a positive test result should be an important warning
sign and lead to intensified counseling on the prevention of sexually
transmitted HCV-infections in these high-risk patients. Our results support
recent guidelines [27] to screen HIV-infected MSM with high-risk sexual
behaviors or concomitant ulcerative sexually transmitted diseases for hepatitis
C, and to screen for syphilis in MSM with acute hepatitis C. Although the use
of non-injection drugs has been associated with HCV infections among
HIV-infected MSM [28], we found no association between non-injection drug use
and HCV seroconversions in MSM. However, we cannot exclude that our
analysis underestimates the effect of non-injection drug use due to
underreporting of this behavior. Neither immunological status nor HIV-RNA
was associated with HCV-seroconversion: the large majority (96%) of
HCV-infections occurred at CD4-T cell counts above 200/ul and most cases
(93%) were on ART.
The incidence of HCV infections in
HIV-infected IDU in the SHCS has decreased in recent years, underscoring the
considerable success of preventive interventions such as methadone substitution
and needle exchange programs in reducing HCV-infections in IDUs followed in
routine HIV care. Furthermore, Switzerland’s long-term heroin prescription
program likely contributed to the decreasing incidence of HCV seroconversions
in this population. Besides a reduction in risk behavior, it is also possible
that protective genetic markers have been enriched during the course of the
epidemic in uninfected IDUs, as has been demonstrated recently [29]. However,
the large majority (90%) of IDU were excluded from the incidence analyses
because they already had a positive HCV serology at entry, limiting the size of
the IDU population available for the incidence calculations. In line with
previous studies [30, 31], the incidence of HCV infection in HET has remained
very low, and we cannot exclude that the few incident cases were related to
undisclosed IDU and/or MSM-related sexual activity. This is corroborated by
recent phylogenetic studies within the SHCS which revealed that
approximately 11% HIV-pol sequences from
heterosexuals were linked to transmission clusters of MSM
[32].
The major strength of our study is the
long-term routine HCV-screening at baseline and during follow-up in a
representative, nationwide HIV-population and in different transmission groups.
Routine serological surveillance of HCV infection is paramount to the diagnosis
of new infections, as the majority of individuals experiencing such an event
remain asymptomatic, and transaminase elevations can be transient and
unrecognized. Furthermore, we could minimize transmission group
misclassification through the availability of detailed longitudinal information
on sexual activity and drug abuse. An important limitation of our analysis resides in the
fact that we could not estimate the exact HCV seroconversion dates between two
serological screening tests. Furthermore, a limited number of patients might
have experienced HCV-reinfections which cannot be recognized by serological
testing, possibly leading to the underestimation of the true overall incidence
of HCV infections. Similarly, the incidence might have been underestimated
through patients with late or absent seroconversion, as described by Thomson et
al. [13]. A further limitation lies in the sub-optimal sensitivity and
specificity of HCV antibody assays. Although positive ELISA tests were
routinely confirmed by immunoblot, false positive results cannot be entirely
excluded. Finally, some patients who experienced an HCV seroconversion could
have lost their antibodies within a few months after the infection in the
setting of spontaneous HCV-clearance [16-18]. In the SHCS, 25 seroconverters
had a unique positive HCV serology, followed by further negative tests.
However, incidence rates remained similar after excluding these patients. In
the SHCS, the large majority of recent HCV infections occurred in MSM.
This underscores the need
for improved HCV surveillance and prevention among HIV-infected MSM.
Accordingly, the intervals of HCV-screening will be reduced to 1 year in the
SHCS. This should allow identifying patients with new HCV-infections earlier in
order to optimize counselling with regard to HCV transmission risks, and to
improve the response to HCV-therapy by initiating treatment during acute
infection.
The revised screening practice in the SHCS is in line with recent guidelines
which recommend yearly screening for all HIV-infected MSM who still engage in
any risk behavior [33-35]. Furthermore,
a recent study demonstrated that yearly serological HCV screening and 6-monthly
liver function testing is cost-effective [36]. Importantly, this analysis
demonstrated that the optimal screening strategy strongly depends on
HCV-incidence.
For instance, it suggests that 3-monthly
screening with liver function tests is costeffective in settings with
HCV-incidences above 1.25%. Accordingly, our results should provide important
information for establishing cost-effective screening programs in HIV-infected
patients. However, routine serological screening does not prevent risk
behaviour, and coordinated efforts between physicians and public-health
representatives are needed to control the ongoing HCV infection epidemic in
HIV-infected MSM. Although most reported HCV-transmissions among MSM occurred
in HIV-infected patients, it is conceivable that HCV-incidence among
HIV-negative MSM is underestimated as these individuals are not in regular
medical care. Clinicians and patients should be aware of the risk of acute HCV
infection in MSM, and intensified prevention and counselling should be
performed. Despite the fact that antiretroviral treatment is highly efficient
in preventing HIV-transmission it does not have any effect on prevention of
other sexually transmitted diseases. Condom use declined in recent years in MSM
with suppressed HIV viral load in the SHCS [37], which likely contributes to
the increasing incidence of HCV-infections in this population. It is crucial that HIV-infected MSM are
counselled with regard to the risks of sexual activities that involve traumatic
mucosal sex, and that condoms are used consistently in sexual risk situations
[28]. The example of IDUs demonstrates that it is possible to reduce the
incidence of HCV-infections through improved screening and preventive
interventions.
Clinical Infectious Diseases Advance Access published
August 14, 2012
Gilles
Wandeler1,2,
Thomas Gsponer 2,
Andrea Bregenzer3,
Huldrych F. Günthard4,
Olivier
Clerc5,
Alexandra Calmy6,
Marcel Stöckle7,
Enos Bernasconi8,
Hansjakob
Furrer1,
Andri Rauch1
and the Swiss HIV Cohort Study
1Clinic
for Infectious Diseases, Bern University Hospital and University of Bern,
Bern,
Switzerland,
2Institute
of Social and Preventive Medicine, University of Bern, Bern,
Switzerland,
3Cantonal
Hospital, St.Gallen, Switzerland, 4Division
of Infectious Diseases
and
Hospital Epidemiology, University Hospital Zurich, University of Zurich,
Zurich,
Switzerland,
5University
Hospital Lausanne, Lausanne, Switzerland, 6University
Hospital
Geneva,
Geneva, Switzerland, 7University
Hospital Basel, Basel, Switzerland, 8Regional
Hospital,
Lugano, Switzerland
incidence in the different HIV transmission groups in the Swiss HIV Cohort
Study (SHCS) in order to optimize surveillance and prevention......In
MSM, history of inconsistent condom use (aHR: 2.09, 1.33-3.29) and past
syphilis (2.11, 1.39-3.20) predicted HCV seroconversions.....IDUs are the
prevalent group with HCV in HIV+ in the Swiss Cohort but the high increasing
incident rate among HIV+ MSM (increased 18-fold in the last 13 years) reflects
a trend observed and reported in Western Europe in major cities & reported
at conferences over the past few years....."In
MSM, history of inconsistent condom use (aHR: 2.09, 1.33-3.29) and past
syphilis (2.11, 1.39-3.20) predicted HCV seroconversions.....In the SHCS, the large majority of recent HCV
infections occurred in MSM"......"suggests that 3-monthly screening with liver
function tests is costeffective in settings with HCV-incidences above 1.25%.
Accordingly, our results should provide important information for establishing
cost-effective screening programs in HIV-infected patients. However, routine
serological screening does not prevent risk behaviour, and coordinated efforts
between physicians and public-health representatives are needed to control the
ongoing HCV infection epidemic in HIV-infected MSM.....It
is crucial that HIV-infected MSM are counselled with regard to the risks of
sexual activities that involve traumatic mucosal sex, and that condoms are used
consistently in sexual risk situations [28]. The example of IDUs demonstrates
that it is possible to reduce the incidence of HCV-infections through improved
screening and preventive interventions......Accordingly, the intervals of HCV-screening will be
reduced to 1 year in the SHCS. This should allow identifying patients with new
HCV-infections earlier in order to optimize counselling with regard to HCV
transmission risks, and to improve the response to HCV-therapy by initiating
treatment during acute infection. The revised screening practice in the SHCS is
in line with recent guidelines which recommend yearly screening for all
HIV-infected MSM who still engage in any risk behavior [33-35]. Furthermore, a
recent study demonstrated that yearly serological HCV screening and 6-monthly
liver function testing is cost-effective [36]. Importantly, this analysis
demonstrated that the optimal screening strategy strongly depends on
HCV-incidence"
"It is
crucial that HIV-infected MSM are counselled with regard to the risks of sexual
activities that involve traumatic mucosal sex, and that condoms are used
consistently in sexual risk situations [28]. The example of IDUs demonstrates
that it is possible to reduce the incidence of HCV-infections through improved
screening and preventive interventions.....This underscores the need for improved HCV
surveillance and prevention among HIV-infected MSM. Accordingly, the intervals
of HCV-screening will be reduced to 1 year in the SHCS. This should allow
identifying patients with new HCV-infections earlier in order to optimize
counselling with regard to HCV transmission risks, and to improve the response
to HCV-therapy by initiating treatment during acute infection"
"Of 4,629 MSM, 2,678
IDU and 4,530 HET screened for HCV infection, 3,333 (72%), 123 (5%) and 3,078
(68%) had a negative HCV serology at baseline and available serological
follow-up. Over 23,707 person-years (py) for MSM, 733 py for IDU and 20,752 py
for HET, 101 (3%), 41 (33%) and 25 (1%) seroconverted, respectively. The IR
of HCV infections in MSM increased from 0.23 (95% CI: 0.08-0.54) per 100py in
1998 to 4.09 (2.57-6.18) in 2011. The IR decreased in IDU and remained
below 1 per 100 py in HET. In MSM, history of inconsistent condom use (aHR:
2.09, 1.33-3.29) and past syphilis (2.11, 1.39-3.20) predicted HCV
seroconversions.
Conclusions. In the SHCS, HCV incidence
decreased in IDU, remained stable in HET and increased 18-fold in MSM in the
last 13 years. These observations underscore the need for improved HCV
surveillance and prevention among HIV-infected MSM."
Most
demographic and behavioural characteristics were similar in HCV-seroconverters
compared to patients who did not seroconvert (Web-Table 1). Median age was 34.9
years in seroconverters and 37.7 years in non-seroconverters, and the
proportion of Caucasians was 93% and 92%, respectively. In both groups, over
90% of patients started ART before or during follow-up. However, previous HBV
exposure, a past history of syphilis and inconsistent condom use were more
common in HCV-seroconverters (Web-Table 1). The incidence of HCV infections was
higher in patients who reported inconsistent condom use, as well as in those
who had a previous episode of syphilis or exposure to HBV compared to the other
patients (Figure 3). In adjusted Cox regression analyses, only inconsistent
condom use (adjusted Hazard Ratio (aHR): 2.09,
95% CI 1.33-3.29) and previous diagnosis of syphilis (2.11,
1.39-3.20) were significantly associated with HCV seroconversions (Table
2).
In the SHCS, the large majority
of recent HCV infections occurred in MSM. This underscores the need for
improved HCV surveillance and prevention among HIV-infected MSM. Accordingly,
the intervals of HCV-screening will be reduced to 1 year in the SHCS. This
should allow identifying patients with new HCV-infections earlier in order to
optimize counselling with regard to HCV transmission risks, and to improve the
response to HCV-therapy by initiating treatment during acute infection. The
revised screening practice in the SHCS is in line with recent guidelines which
recommend yearly screening for all HIV-infected MSM who still engage in any
risk behavior [33-35]. Furthermore, a recent study demonstrated that yearly
serological HCV screening and 6-monthly liver function testing is
cost-effective [36]. Importantly, this analysis demonstrated that the optimal
screening strategy strongly depends on HCV-incidence.
For instance, it suggests that 3-monthly
screening with liver function tests is costeffective in settings with
HCV-incidences above 1.25%. Accordingly, our results should provide important
information for establishing cost-effective screening programs in HIV-infected
patients. However, routine serological screening does not prevent risk
behaviour, and coordinated efforts between physicians and public-health
representatives are needed to control the ongoing HCV infection epidemic in
HIV-infected MSM. Although most reported HCV-transmissions among MSM occurred
in HIV-infected patients, it is conceivable that HCV-incidence among
HIV-negative MSM is underestimated as these individuals are not in regular
medical care. Clinicians and patients should be aware of the risk of acute HCV
infection in MSM, and intensified prevention and counselling should be
performed. Despite the fact that antiretroviral treatment is highly efficient
in preventing HIV-transmission it does not have any effect on prevention of
other sexually transmitted diseases. Condom use declined in recent years in MSM
with suppressed HIV viral load in the SHCS [37], which likely contributes to
the increasing incidence of HCV-infections in this population. It is crucial
that HIV-infected MSM are counselled with regard to the risks of sexual
activities that involve traumatic mucosal sex, and that condoms are used consistently in sexual risk
situations [28]. The example of IDUs demonstrates that it is possible to reduce
the incidence of HCV-infections through improved screening and preventive
interventions.
Abstract
Background.  Hepatitis C Virus (HCV)
infection has a growing impact on morbidity and mortality in HIV-infected
patients. We assessed trends in HCV incidence in the different HIV transmission
groups in the Swiss HIV Cohort Study (SHCS) in order to optimize surveillance
and prevention.
Methods. HCV infection incidence was
assessed from 1998, when routine serial HCV screening was introduced in the
SHCS, until 2011. All HCV-seronegative patients with at least one follow-up
serology were included. Incidence rates (IR) of HCV infections were compared
between men who have sex with men (MSM), injection drug users (IDU) and
heterosexuals (HET). Predictors of HCV seroconversion in MSM were estimated.
Results. Of 4,629 MSM, 2,678 IDU and
4,530 HET screened for HCV infection, 3,333 (72%), 123 (5%) and 3,078 (68%) had
a negative HCV serology at baseline and available serological follow-up. Over
23,707 person-years (py) for MSM, 733 py for IDU and 20,752 py for HET, 101
(3%), 41 (33%) and 25 (1%) seroconverted, respectively. The IR of HCV
infections in MSM increased from 0.23 (95% CI: 0.08-0.54) per 100py in 1998 to
4.09 (2.57-6.18) in 2011. The IR decreased in IDU and remained below 1 per 100
py in HET. In MSM, history of inconsistent condom use (aHR: 2.09, 1.33-3.29)
and past syphilis (2.11, 1.39-3.20) predicted HCV seroconversions.
Conclusions. In the SHCS, HCV incidence
decreased in IDU, remained stable in HET and increased 18-fold in MSM in the
last 13 years. These observations underscore the need for improved HCV
surveillance and prevention among HIV-infected MSM.
Hepatitis C Virus (HCV) infection is a
major cause of morbidity and mortality in HIV-infected patients [1-3]. For many
years, HCV-infections occurred almost exclusively in injection drug users (IDU)
or hemophiliacs. In a previous analysis in the Swiss HIV Cohort study
(SHCS), HCV seroprevalence was 33% overall and 90% among individuals reporting
IDU [4]. Recently, there have been several outbreaks of HCV infections
among HIV-infected men who have sex with men (MSM), predominantly in large
cities in Europe, Australia and the USA [5-14]. These clusters of HCV
infections involved high-risk sexual practices within confined social networks.
However, the impact of these localized epidemics on the incidence of HCV
infections in general HIV-infected populations and the long-term trends in HCV
incidence rates in different transmission risk groups are largely unknown. A
better understanding of changes in HCV-incidence rates and transmission
patterns is urgent because of the increasing burden of HCV-related disease in
HIV-infected patients [1-3].
The SHCS offers an ideal platform for
studying changes in HCV incidence in a general HIV-infected population and in
diverse HIV-transmission risk groups, as all HCV seronegative patients are
screened routinely at baseline and during follow-up since 13 years.
Furthermore, it allows a thorough evaluation of the association of HCV
seroconversions with different risk factors, as detailed patient history and
clinical data is collected every six months.
We assessed changes in the HCV infection
incidence in the three main HIV transmission groups (IDU, MSM and heterosexuals
(HET)) in the SHCS during the last 13 years, with the expectation that our
results would shed light on the magnitude and changes of the HCV epidemic in a
general and representative HIV-infected population.
METHODS
Swiss HIV cohort Study
The Swiss HIV Cohort Study (SHCS, www.shcs.ch) is a prospective cohort study with
ongoing enrolment of HIV-infected adults in Switzerland since 1988. It covers
at least 45% of the cumulative number of HIV infections declared to the Swiss
public health authorities, 69% of all patients living with AIDS and 75% of
patients receiving antiretroviral therapy in Switzerland [15]. Representativity
remained stable over the years. Detailed information on demographics, mode of
HIV acquisition, risk behavior, clinical events, co-infections, and treatment
is collected using a standard protocol at registration and then at intervals of 6 months. Local
ethical committees of all participating study sites have approved the study and
written consent has been obtained from all participants.
Inclusion criteria and
definitions
To compare HCV incidence between risk
groups, we categorised patients as IDU, MSM or HET according to the most
probable HIV-transmission mode. In order to minimize erroneous classifications,
we excluded MSM and HET patients who reported the use of injection drugs at any
occasion during follow-up. Since 1998, all SHCS patients have been routinely
screened for HCV infection: serology is performed every second year of
follow-up, independently of HIV transmission mode. Thus, all HCV-seronegative
patients who had at least one follow-up HCV antibody measurement after July
1998 were included in the analyses. Positive results by third-generation ELISA
are confirmed by immunoblot. Patients with a positive HCV serology at entry,
and those who seroconverted before 1998 were excluded. Individual follow-up
ended at the time of the first positive or last negative HCV serology. A
detailed patient flow chart is shown in Figure 1. Previous syphilis was defined
as a positive Treponema-pallidumhemagglutination-assay (TPHA) screening test
and hepatitis B virus (HBV) exposure by the presence of a positive anti-HBc
antibody test prior to HCV seroconversion, or before
the last negative HCV-serology in those
without HCV seroconversion. In this context, the term HBV exposure refers to
both resolved and chronic HBV infections. A subset of HCV seroconverters was
found to have a unique positive HCV serology followed by one or several
negative serologies. These patients could have cleared HCV antibodies after
spontaneous HCV clearance, as described before [16-18]. Alternatively, false
positive results cannot be excluded despite confirmation by HCV-immunoblot.
Therefore, the main analyses were repeated after exclusion of these
cases.
Statistical analyses
HCV incidence rates were assessed from
introduction of routine HCV-screening from July 1998 until November 2011.
Patients were considered to be at risk for HCV infection since the date of
their first negative HCV serology or since 01.07.1998 if the first measurement
preceded this date. Individual follow-up ended at the time of the first
positive or last negative HCV serology. Baseline characteristics were compared
between the three risk groups using analysis of variance (ANOVA) and
chi-squared tests for continuous and categorical variables, respectively. To
calculate the incidence of HCV infections in the different risk groups, events
were right-censored so that the date of seroconversion corresponded to the date
of the first positive HCV serology. Yearly HCV infection incidence rates (IR)
were obtained using Bayesian Poisson regression models with non-parametric
smoothing priors [19]. IR are reported as number of cases per 100 person-years
(py) and shown on a log scaled graph, by transmission group (MSM, IDU,
HET).
In sensitivity analyses, the yearly rates
of HCV infection in MSM were compared to the results obtained using two
alternate censoring approaches: (1) Mid-point censoring, in which the date of
seroconversion was set to the mid-point between the last negative and first
positive HCV serology, and (2) a non-parametric maximum likelihood estimation
method for interval-censored data [20]. Interval-censoring takes into account
that the exact date of seroconversion is not known but lies between the last
negative and the first positive serology. However, due to the low number of
seroconversions in IDU and HET, this method could only be applied to the MSM
risk group.
Demographic characteristics of
HCV-seroconverters in MSM were compared to those who did not seroconvert during
follow-up using Chi-square and Mann-Whitney tests. The difference in incidence
of HCV infections between patients of different age groups, condom use
patterns, hepatitis B exposure status and history of past syphilis were also
shown in log-scaled graphs. Risk factors for HCV infection in MSM were
evaluated using a multivariable Cox regression model. Analyses were adjusted
for age category (16-29, 30-39 and > =40 years), CD4 count category before
last HCV serology (<200, 200-499 and >=500 cells/ μl), education level
(no, basic and high-level professional education), region of SHCS follow-up
(Zurich or other), ethnicity (Caucasian or other), use of non-IDU drugs (yes or
no), sexual relationships (stable or occasional partners), condom use (always
or inconsistent), past history of syphilis (yes or no), HBV exposure (yes or
no) and antiretroviral therapy (ART) status (yes or no). All statistical
analyses were performed using Stata 12 (Stata Corp, College Station, USA) and
R-2.14.1 [21] using add-on packages INLA [22] and Icens [23].
RESULTS
Baseline
characteristics
Of 4,629 MSM, 2,678 IDU and 4,530 HET
screened for HCV infection, 147 (3.2%), 2,468 (92.2%) and 513 (11.3%) had a
positive HCV serology at baseline (Figure 1). 6,534 patients, of which 3,333
MSM, 123 IDU and 3,078 HET, were included in the HCV incidence analyses. Median
age at first HCV screening test was slightly higher in MSM (38 years,
inter-quartile range (IQR) 32-44) and HET (36, 30-44), compared to IDU (33,
28-40), whereas MSM were more likely to have a high-level education than
patients in the two other groups (Table 1). Almost half of the MSM and IDU, but
only 29% of HET had their regular medical follow-up in Zurich, Switzerland’s
largest urban center. Finally, HET were much more likely to be non-Caucasians
(40.7%), compared to MSM (7.7) and IDU (8.9). HCV incidence
Over a total follow-up time of 23,707
person-years (py) for the MSM group, 733 py for the IDU group and 20,752 py for
the HET group, 101 (3.0%), 41 (33.3%) and 25 (0.8%) patients experienced an HCV
seroconversion during follow-up, respectively. Between 1998 and 2011, the IR of
HCV infections in MSM increased from 0.23 (95% CI: 0.08-0.54) per 100 py to
4.09 (2.57-6.18) in 2011in2011, with 51 cases observed in the last 3 years
(Figure 2). There was a similar increase in patients followed in Zurich
compared to the rest of the country (IR in 2011 were 2.61 (1.19-5.13) and 4.10
(2.18-7.16) per 100 py, respectively). Increases in HCV-incidence in MSM were
similar when right-censoring, midpoint-censoring or interval-censoring were
used to estimate yearly incidence rates (Web-Figure 1).
In IDU, HCV infection IR decreased from
13.89 (95% CI: 8.20-22.39) per 100 py in 1998 to 2.24 (0.55-10.66) in 2011,
with only 3 incident cases in the last 3 years. For comparison, the yearly
IR of HCV infections in the HET group remained below 0.5 per 100 py, with only
2 cases in 2011 (IR 0.43 per 100py, 95% CI: 0.12-1.29; Figure 2). Twenty-five
(15.0%) HCV-seroconverters had a unique positive serology followed by negative
tests. Of those, six patients had their HCV infection confirmed by HCV RNA
testing within one month of the positive serology. In another eight patients,
HCV RNA testing was negative,, which may be reflective of spontaneous HCV
clearance and loss of HCV-antibodies over time. Eleven patients did not have an
HCV RNA available to confirm the HCV infection. After exclusion of these
patients, the incidence rates of HCV infection remained similar to the results
described above (1998: MSM: 0.11 per 100 py, 95% CI 0.03-0.35; IDU: 14.36,
8.36-23.39; HET: 0.09, 0.02-0.35. 2011: MSM: 3.56, 2.19-5.53; IDU: 2.42,
0.51-13.36; HET: 0.47, 0.14-1.30).
Predictors of HCV infections in
MSM
Most demographic and behavioural
characteristics were similar in HCV-seroconverters compared to patients who did
not seroconvert (Web-Table 1). Median age was 34.9 years in seroconverters and
37.7 years in non-seroconverters, and the proportion of Caucasians was 93% and
92%, respectively. In both groups, over 90% of patients started ART before or
during follow-up. However, previous HBV exposure, a past history of syphilis
and inconsistent condom use were more common in HCV-seroconverters (Web-Table
1). The incidence of HCV infections was higher in patients who reported
inconsistent condom use, as well as in those who had a previous episode of
syphilis or exposure to HBV compared to the other patients (Figure 3). In
adjusted Cox regression analyses, only inconsistent condom use (adjusted Hazard
Ratio (aHR): 2.09,
95% CI 1.33-3.29) and previous diagnosis of syphilis (2.11,
1.39-3.20) were significantly associated with HCV seroconversions (Table
2). Age, education level, CD4+
T cell counts, ART status and the other variables included in the
multivariable model were not significantly associated with the main outcome. Of
note, having occasional sexual partners (aHR 0.97,
0.56-1.68) and being followed-up in Zurich (aHR 1.30,
0.87-1.97) did not predict HCV seroconversions.
Of the 63 (62%) HCV infections in MSM with
detectable HCV RNA and available genotyping results, the majority were caused
by HCV genotype 1 (42 cases, 66.7%), followed by genotypes 4 (12, 19.0%), 3 (8,
12.7%) and 2 (1, 1.6%).
DISCUSSION
We assessed trends in HCV incidence
between 1998 and 2011 in the main HIV transmission groups in the SHCS. In a
population of over 6,500 patients screened every 2 years for HCV infection, we
found that the yearly incidence rate had decreased in IDU, remained stable in
HET and dramatically increased in MSM. In the latter sub-population, 50% of all
HCV infections occurred in the last three years. In MSM, a history of
inconsistent condom use and a past episode of syphilis were significantly
associated with HCV seroconversion.
Since 2005, several local reports
described epidemics of HCV infections among HIV-infected MSM [7-12].
Phylogenetic and socio-demographic analyses have shown that these infections
occurred within confined groups of MSM with high-risk sexual behavior in large
urban centers [11, 24]. Therefore, incidence estimates in these cohorts were
not representative for the general HIV-infected population, in contrast to our
study. The systematic, nationwide screening of all HIV-infected patients at
risk since 1998 allowed us to accurately estimate the trends in HCV-incidence
among different risk groups. Since 1998, the incidence of HCV infection among
MSM increased 18-fold, reaching 4.1 cases per 100 py in 2011. This incidence
was higher compared to the large majority of previous studies among
HIV-infected MSM reporting estimates lower than one case per 100 py [5].
However, the most recent HCV-incidence in the SHCS was similar to estimates
from an international cohort collaboration of HIV-seroconverters, where
incidence estimates ranged from 2.3 to 5.1/100 py in 2007 [25]. In accordance
to previous clusters of HCV-infections in large cities, the HCV-incidence
increased markedly among individuals treated in Zurich, the largest urban
center in Switzerland which, to some extent, is comparable to other large
cities in Europe and the US. However, our study shows that HCV-incidence
increased similarly outside of the Zurich area, indicating a nationwide
increase in HCV-infections in HIV-infected MSM.
In line with previous reports [4, 8,
11], unprotected anal sex was an important risk factor for HCV infections in
MSM. In the SHCS, inconsistent condom use doubled the risk of HCV infection in
MSM. Furthermore, a previous syphilis infection independently increased the
risk of an HCV seroconversion approximately twofold. Of note, we recently
observed a significant increase in syphilis acquisition among MSM in the SHCS,
paralleling the current HCV epidemic [26]. The shared route of transmission
of the two infections is the most likely explanation for this association.
Alternatively, it is also conceivable that the mucosal disruption caused by
syphilitic ulcers facilitates HCV infections. These findings underscore the
importance of the link between high risk sexual behavior and HCV transmissions.
As yearly syphilis screening and baseline HBV serology testing are routine
for all MSM in the SHCS, a positive test result should be an important warning
sign and lead to intensified counseling on the prevention of sexually
transmitted HCV-infections in these high-risk patients. Our results support
recent guidelines [27] to screen HIV-infected MSM with high-risk sexual
behaviors or concomitant ulcerative sexually transmitted diseases for hepatitis
C, and to screen for syphilis in MSM with acute hepatitis C. Although the use
of non-injection drugs has been associated with HCV infections among
HIV-infected MSM [28], we found no association between non-injection drug use
and HCV seroconversions in MSM. However, we cannot exclude that our
analysis underestimates the effect of non-injection drug use due to
underreporting of this behavior. Neither immunological status nor HIV-RNA
was associated with HCV-seroconversion: the large majority (96%) of
HCV-infections occurred at CD4-T cell counts above 200/ul and most cases
(93%) were on ART.
The incidence of HCV infections in
HIV-infected IDU in the SHCS has decreased in recent years, underscoring the
considerable success of preventive interventions such as methadone substitution
and needle exchange programs in reducing HCV-infections in IDUs followed in
routine HIV care. Furthermore, Switzerland’s long-term heroin prescription
program likely contributed to the decreasing incidence of HCV seroconversions
in this population. Besides a reduction in risk behavior, it is also possible
that protective genetic markers have been enriched during the course of the
epidemic in uninfected IDUs, as has been demonstrated recently [29]. However,
the large majority (90%) of IDU were excluded from the incidence analyses
because they already had a positive HCV serology at entry, limiting the size of
the IDU population available for the incidence calculations. In line with
previous studies [30, 31], the incidence of HCV infection in HET has remained
very low, and we cannot exclude that the few incident cases were related to
undisclosed IDU and/or MSM-related sexual activity. This is corroborated by
recent phylogenetic studies within the SHCS which revealed that
approximately 11% HIV-pol sequences from
heterosexuals were linked to transmission clusters of MSM
[32].
The major strength of our study is the
long-term routine HCV-screening at baseline and during follow-up in a
representative, nationwide HIV-population and in different transmission groups.
Routine serological surveillance of HCV infection is paramount to the diagnosis
of new infections, as the majority of individuals experiencing such an event
remain asymptomatic, and transaminase elevations can be transient and
unrecognized. Furthermore, we could minimize transmission group
misclassification through the availability of detailed longitudinal information
on sexual activity and drug abuse. An important limitation of our analysis resides in the
fact that we could not estimate the exact HCV seroconversion dates between two
serological screening tests. Furthermore, a limited number of patients might
have experienced HCV-reinfections which cannot be recognized by serological
testing, possibly leading to the underestimation of the true overall incidence
of HCV infections. Similarly, the incidence might have been underestimated
through patients with late or absent seroconversion, as described by Thomson et
al. [13]. A further limitation lies in the sub-optimal sensitivity and
specificity of HCV antibody assays. Although positive ELISA tests were
routinely confirmed by immunoblot, false positive results cannot be entirely
excluded. Finally, some patients who experienced an HCV seroconversion could
have lost their antibodies within a few months after the infection in the
setting of spontaneous HCV-clearance [16-18]. In the SHCS, 25 seroconverters
had a unique positive HCV serology, followed by further negative tests.
However, incidence rates remained similar after excluding these patients. In
the SHCS, the large majority of recent HCV infections occurred in MSM.
This underscores the need
for improved HCV surveillance and prevention among HIV-infected MSM.
Accordingly, the intervals of HCV-screening will be reduced to 1 year in the
SHCS. This should allow identifying patients with new HCV-infections earlier in
order to optimize counselling with regard to HCV transmission risks, and to
improve the response to HCV-therapy by initiating treatment during acute
infection.
The revised screening practice in the SHCS is in line with recent guidelines
which recommend yearly screening for all HIV-infected MSM who still engage in
any risk behavior [33-35]. Furthermore,
a recent study demonstrated that yearly serological HCV screening and 6-monthly
liver function testing is cost-effective [36]. Importantly, this analysis
demonstrated that the optimal screening strategy strongly depends on
HCV-incidence.
For instance, it suggests that 3-monthly
screening with liver function tests is costeffective in settings with
HCV-incidences above 1.25%. Accordingly, our results should provide important
information for establishing cost-effective screening programs in HIV-infected
patients. However, routine serological screening does not prevent risk
behaviour, and coordinated efforts between physicians and public-health
representatives are needed to control the ongoing HCV infection epidemic in
HIV-infected MSM. Although most reported HCV-transmissions among MSM occurred
in HIV-infected patients, it is conceivable that HCV-incidence among
HIV-negative MSM is underestimated as these individuals are not in regular
medical care. Clinicians and patients should be aware of the risk of acute HCV
infection in MSM, and intensified prevention and counselling should be
performed. Despite the fact that antiretroviral treatment is highly efficient
in preventing HIV-transmission it does not have any effect on prevention of
other sexually transmitted diseases. Condom use declined in recent years in MSM
with suppressed HIV viral load in the SHCS [37], which likely contributes to
the increasing incidence of HCV-infections in this population. It is crucial that HIV-infected MSM are
counselled with regard to the risks of sexual activities that involve traumatic
mucosal sex, and that condoms are used consistently in sexual risk situations
[28]. The example of IDUs demonstrates that it is possible to reduce the
incidence of HCV-infections through improved screening and preventive
interventions.
Clinical Infectious Diseases Advance Access published
August 14, 2012
Gilles
Wandeler1,2,
Thomas Gsponer 2,
Andrea Bregenzer3,
Huldrych F. Günthard4,
Olivier
Clerc5,
Alexandra Calmy6,
Marcel Stöckle7,
Enos Bernasconi8,
Hansjakob
Furrer1,
Andri Rauch1
and the Swiss HIV Cohort Study
1Clinic
for Infectious Diseases, Bern University Hospital and University of Bern,
Bern,
Switzerland,
2Institute
of Social and Preventive Medicine, University of Bern, Bern,
Switzerland,
3Cantonal
Hospital, St.Gallen, Switzerland, 4Division
of Infectious Diseases
and
Hospital Epidemiology, University Hospital Zurich, University of Zurich,
Zurich,
Switzerland,
5University
Hospital Lausanne, Lausanne, Switzerland, 6University
Hospital
Geneva,
Geneva, Switzerland, 7University
Hospital Basel, Basel, Switzerland, 8Regional
Hospital,
Lugano, Switzerland
RSS Feed