A programme that provides pre-exposure prophylaxis (PrEP)
through primary care referrals in San Francisco has not seen a single HIV
infection in people using PrEP since it started in 2012. This is despite high levels
of sexually transmitted infections (STIs) other than HIV and, in about
40% of users, less use of condoms.
In an editorial on the findings of the study, Kimberley
Koester of UCSF and Bob Grant, principal investigator of the pioneering iPrEx
PrEP study, comment that it is important to keep one’s eyes on the main prize –
no new HIV infections – as “an expected and most desired outcome”.
However, they note, it is also “Time for a vigorous
conversation about sexually transmitted infections, too long eclipsed by fear
of HIV infection.” They suggest that PrEP may offer more opportunities both
within clinics and within the community to discuss and test for other STIs.
Kaiser Permanente is one of the USA’s largest healthcare providers,
providing medical care to over nine million people nationwide and 170,000
adults within the city of San Francisco. The present study includes all Kaiser
Permanente clients living in San Francisco who were evaluated for PrEP between
July 2012, when the US Food and Drug Administration approved
tenofovir/emtricitabine (Truvada) for
PrEP, until February 2015.
The Kaiser programme generally refers patients to a
specialist PrEP programme through their primary care physician, at the patient’s
request or physician’s suggestion. Patients are always screened for medical
contra-indications and tested for HIV during referral. While taking PrEP they are
tested for HIV and STIs at least quarterly (for some people as often as once a
month). In addition, since July 2014, patients have been sent a confidential questionnaire,
six months after starting PrEP, about whether there have been any changes in
their sexual behaviour.
During the study period, 1045 people were referred for PrEP. Of
these 80% were evaluated for PrEP by specialist providers and 657 actually
started PrEP (63% of those referred and 81% of those evaluated). Twenty
individuals stopped and restarted PrEP in the study period – they were not
classed as ‘initiators’ – and 144 individuals were evaluated but chose not to
As the researchers note, there was a distinct expansion in
interest and uptake after September 2013 (see graphic below – click to open a larger version) and again in May 2014
when the US Centers
for Disease Control and Prevention (CDC) recommended PrEP. As assessment and initiation lag
behind referral, we do not yet know if the referral peak in Jan-Feb 2015 has
been reflected in more initiations.
Of the 657 people who started PrEP, all were cisgender (born male) men
who had sex with men (MSM) except for three heterosexual women and one
transgender MSM. Their average age was 37. Only one person reported injecting
drugs and only 15 people (2%) reported using post-exposure prophylaxis (PEP)
in the three months before starting PrEP.
People who initiated PrEP, as opposed to people who decided
not to, were more likely to report multiple sexual partners (84% versus 69%).
Eight per cent of initiators had previously been in a PrEP study. Initiators
were not more likely to have HIV-positive regular partners than non-initiators.
The main reasons for not starting
PrEP in non-initiators were being at low risk either in their own or the
physician’s judgement (35%), concerns about cost (15%) and not wanting to do
all the required follow-up tests (10%). Only 3% cited concerns about side-effects
as a reason not to start.
The average length of time people actually stayed on PrEP
during the study period was seven months, but this includes a large number of people
who were still on PrEP in February 2015 when the study period ended.
There were no HIV infections in anyone who initiated PrEP
during the study period. However, 187 people (28.5%) were diagnosed
with another STI during follow-up, 78 of them (12%) more than once. The annual
incidence rate for any STI – i.e. the proportion of STIs diagnosed over the
course of a year on PrEP, in those who took it for a year or more – was 50%. Annual incidence was 33% for rectal infections, 33% for chlamydia, 28% for gonorrhoea
and 5.5% for syphilis.
One hundred and eighty-eight PrEP users were sent the six-month
behaviour-change survey and three-quarters of them completed it, so we know
about self-reported behaviour change in about 22% of PrEP users. In these 143
people, there was a decrease in condom use in a minority: condom use stayed the
same in 56% of PrEP users, decreased in 41%, and increased in 3%. The number of
sexual partners remained more or less unchanged: it stayed the same in 74% of
PrEP users, decreased in 15% and increased in 11%.
The researchers note that the level of STIs seen cannot be
compared to a matched control group not taking PrEP, so we do not know if STI
incidence is higher than in non-PrEP users. In addition, increases in STI
diagnoses in PrEP users might have been as much due to higher rates of testing as it
was to higher rates of infection, especially as some STIs are asymptomatic and
self-limiting; without regular testing some may pass without notice.
The survey also did not collect data on HIV status, viral
load and PrEP use among partners, so we do not know if PrEP users’ condom choices
were related to the perceived HIV status or PrEP status of others. The authors
comment that a more ‘refined’ concept of risk compensation is needed that can capture
whether increases in condomless sex actually do translate into higher STI risk
in PrEP users: clearly they do not translate into higher HIV risk.
In their commentary, Koester and Grant discuss how control of
HIV via PrEP and antiretroviral therapy may lead to more discussion of other
STIs. They comment that the surprisingly large number of MSM willing to
request and be referred for PrEP within primary care is welcome, as it had been
assumed MSM would more willingly discuss PrEP in community settings.
However, they add that the lack of STI testing within primary
care is a missed opportunity, as is the general lack of discussion and disclosure of
STIs between MSM. PrEP, they comment, is best combined with a parallel plan for
STIs that may include condoms, frequent STI testing and discussion with
partners. Consideration should also be given to STI self-testing and
They add: “Members of the gay community historically led the
innovations around safer sex practices in the face of HIV/AIDS. If the STI
burden in the context of PrEP use becomes too great, communities can and will make