Given the comparative rarity of anal cancer, screening
the general public is not considered necessary. But for those at higher risk
(gay men with HIV, possibly all gay men and women who have anal sex), cervical
screening is a good precedent for the value of anal screening. In the UK, cervical
screening is offered to women aged 25 to 65. The death rate due to cervical
cancer in women under 45 went down by nearly two-thirds between 1988, when
screening was introduced, and 2002, despite there being an increase in genital
wart diagnoses at the same time.14
So surely we should be trying to do the same for anal
cancer?
Professor Mark Bower is a consultant at London’s
Chelsea and Westminster Hospital,
specialising in HIV-related cancers. Though in favour of people with HIV having
regular anal screening, he says that the case for it being routine is
surprisingly hard to make.
That’s partly because it’s still relatively rare. In the Chelsea and Westminster
cohort, they have seen 60 cases in 11,112 patients (one per 188 patients)
throughout the clinic’s history, but this includes patients coming to the
hospital specifically to see HPV and anal cancer specialists. In patients
attending the Chelsea and Westminster’s general HIV clinic, they see
fewer than one new case a year.
This may seem odd, given that rates of AIN are very high.
For instance, one study of HIV-positive men found that despite AIN grades 2 or 3 being
found at least once in 133 of the 247 patients in the study (54%), there were
only two cases of anal cancer in three years.15
We don’t know
exactly why some anal (or cervical) lesions turn into an invasive cancer, and
others don’t. Bower has evaluated the cases of nearly 1000 HIV-positive men who
have sex with men seen over the last ten years at the Chelsea
and Westminster.
“These guys’
AIN grade goes up and down and up again,” he says. “A lot of them have been
coming here for ten years and show no signs of progressing.”
This is
partly due to the natural history of HPV and the fact that infections regress
as often as they recur. Most AIN grade 1 lesions simply disappear and only a
minority progress to higher grades. We don’t even know the rate at which
high-grade AIN lesions change into anal cancer: estimates vary hugely from 0.2
to 12.5% a year (the consensus is between 1 and 5%). The thing that keeps
lesions coming back in gay men is not persistent HPV infection but reinfection;
in HIV-positive gay men, persistent infection adds to the risk.
Or incidence
of anal cancer may be lower than expected because, in many patient cohorts, gay
men with HIV are already being screened regularly. Even in cervical cancer, it
has been difficult to calculate the benefit of national screening because so
much ad hoc screening was being done before the national programme began.
“Maybe it’s
because of our excellent interventions,” says Bower, “or maybe it’s because
progression to cancer just doesn’t happen in most people with AIN.” There has
never actually been a randomised controlled trial of cervical cancer screening,
and there couldn’t ethically be one of an HPV-associated cancer now: would you allow your doctor to ignore pre-cancerous
cell changes to see if they turned into cancer?
Another
problem is cost-effectiveness.
There have
been two studies in the US,
showing that screening would be relatively cost-effective in both HIV-negative
and HIV-positive gay men. In the cost-effectiveness study in HIV-positive gay
men, the cost per quality-adjusted life-year (QALY) saved was $16,000 with
annual screening and $13,000 if done every two years.16 In
HIV-negative gay men, the cost was considerably greater if you screened annually
($34,800) but comparable if done biennially ($15,100).17
However, a UK
cost-effectiveness model found that national screening of gay men (with or
without HIV) was unlikely to be cost-effective, with an average cost per QALY
gained of £39,405, which is way beyond the usually quoted NICE (National
Institute for Health and Clinical Excellence) threshold of £30,000.18
It was actually more cost-effective to screen all gay men in this study, rather
than just the HIV-positive ones.
This model,
however, contained a number of different assumptions from the US models. In
the US,
it was assumed that annual rate of transition from high-grade AIN to anal
cancer was high: from 3.6 to 5% a year. Actual surveys suggest a lower rate of
progression. The UK
study assumed a much lower rate: about one case of anal cancer per 500 cases on
untreated AIN grades 2 or 3 (0.2%), or one case per 2500 treated cases. This is
probably on the low side, and there have been a number of other criticisms
levelled at the UK
paper, such as the assumption of a high rate of regression from AIN 1 to
asymptomatic.
Screening gay men for anal cancer
and its precursors has not been recommended in UK guidelines. The British HIV Association’s cancer
guidelines of 2008 state: “there is little evidence for routine [screening] as
the early detection of lesions still poses substantial difficulties and single
biopsies may miss areas of AIN, with histology and cytology yielding some
discordant results.”19
In complete contrast, US guidelines
– such as those from New York State20 – recommend “anal cytology at baseline and annually”, especially for men with HPV or
anal warts, and the European AIDS Clinical Society (EACS) guidelines recommend
a rectal examination and/or smear every one to three years for gay men.21
Anoscopy would be reserved for people with abnormal cytology results, and the
New York guidelines estimate that this would be less than 30% of the screened
population.