Screening for anal cancer recommended for HIV positive men
Simon Collins, HIV i-Base
A plenary session at EACS reported that new evidence supports screening for anal cancer in HIV positive men. 
This is important, because the increased risk of anal cancer in gay men, and particularly in HIV positive gay men, has been highlighted for many years. Screening is safe and treatment is effective, especially when diagnosed early and the importance of establishing a screening programme has been repeatedly raised by community-based advocates for many years. Controversially, NHS guidelines do not recommend screening for anal cancer in the two largest risk groups: HIV positive people and gay men.
The presentation, by Professor Mark Bower from the Chelsea and Westminster Hospital, London, who also chaired the panel for the BHIVA malignancy guidelines , clearly supported a review of the NICE decision. He also highlighted that current healthcare resources could not cope adequately with any significant increase in demand for screening.
The lecture started by highlighting the difficulties that are inherent in proving the clinical benefit for any screening programme, including for those that are now an integrated part of NHS care (such as cervical and breast cancer). Interpreting data is dependent on the choice of endpoints, control groups and inherent biases that generally will always support the benefits from screening, even if reduced mortality cannot be inferred.
For example, the incidence of a cancer, or advanced cancer, or even using reductions in cancer specific mortality as an endpoint, does not necessarily provide the data to prove the benefit of screening due to three important inherent biases in evaluating any screening programme are just as relevant in the context of anal cancer and HIV.
Firstly, lead time bias refers to greater survival time after diagnosis. This may be driven by the earlier diagnosis commonly resulting from any availability of broader screening and so survival time since diagnosis does not necessarily have any impact on final mortality. Awareness of a diagnosis for longer is dependent on effective treatment to translate into better prognosis and longer survival. Secondly, lag time bias refers to the tendency for a greater proportion of cancers picked up in a screening programme to be more slowly progressing and less aggressive compared to symptomatic cancers picked up in any control group. In this case, slow growing cancers have a longer screening time in which to be detected and this will translate to an apparent improvement in survival.
Finally, over-diagnosis bias relates to picking up cancers in screening programmes which are never going to progress, or in patients who will die from unrelated or natural causes. This translates to a higher incidence of diagnosis in a screening population but a lower incidence of cancer-related mortality.
Despite the scientific difficulties associated with proving the effectiveness of a screening programme, the presentation outlined why anal cancers screening is now appropriate, based on proven incidence in this population and the effectiveness of treatment.
Although anal cancer was not included as an AIDS-defining malignancy in the US CDC 1993 definitions, unlike cervical cancer that has a similar incidence and etiology, a meta-analysis of major cohort studies has suggested that anal cancer is 20-50 times more common in HIV-positive people than age and gender matched general population.  This is an enormous relative risk: by comparison, tobacco smoking is associated with an approximate 17-fold increased risk for lung cancer.
The HIV effect is also more than a direct result of a weakened immune system: transplant patients who have artificially reduced immunosupression, only have a 4-5 fold increased risk of anal cancer.
Part, or much, or this increased incidence in HIV positive gay men, may be related to the increased risk that was reported in MSM in pre-HIV data. 
Results from 11 linked HIV and cancer registries estimated a relative risk of 59 for HIV positive MSM, but the same study also highlighted a 6-fold increased relative risk in HIV positive compared to HIV-negative IDUs. 
Population studies now estimate the incidence of anal cancer at 1.5 per 100,000 in the general population, but at 35 and 70 per 100,000 in the general population, gay men and HIV positive gay men respectively. One cohort, in San Diego, reported an even higher rate of 224 in HIV positive MSM.  This compares to an incidence of cervical cancer before the introduction of screening or 15 per 100,000.
Anal cancer does not appear related to CD4 count, and some studies have suggested that the incidence post-HAART may be increasing.  This can be balanced by evidence that suggests that, due to effective treatment, survival rates in the HAART era now approximate to that in HIV negative cohorts (of around 75% at two years). 
A limited number of studies support a similar etiology between cervical and anal cancer, with 5% patients diagnosed with AIN2/3 and a similar proportion of patient after surgery for anal warts, progressing to anal cancer. A UCSF study published in 1997 in HIV positive MSM suggested a progression rate of 20% from normal cytology to HSIL and a 60% progression rate for men diagnosed with LSIL to HSIL (with 5% reverting to normal). In 21 patients with invasive anal cancer (from the same UCSF cohort of 1700 men), the median time to progression from a diagnosis of HSIL was 47 months (range 4-139 months).
However, natural history studies should now be considered unethical, as they would for cervical cancer, given the clear link between AIN2/3 and risk of progression to anal cancer and the availability of effective treatment.
Anal cytological screening is easy, well tolerated and acceptable to patients. Results show either normal cytology or one of three diagnosis: ASCUS, LSIL or HSIL (Abnormal Squamous Cells of Undetermined Significance; Low-grade Squamous Intraepithelial Lesions; or High-grade Squamous Intraepithelial Lesions).
In the screening algorithm developed by Joel Palefsky in the US, normal results should lead to repeat annual screening as routine follow-up. ASCUS, LSIL and HSIL should prompt high-resolution anoscopy, with repeat annual anoscopy for AIN1 and treatment for AIN2/3.  This is supported by several studies reporting acceptable rates of 34-83% sensitivity and 38-72% specificity for anal cytology compared to histology.
Finally, the recent availability of reasonable and established treatments for AIN2/3 (infrared coagulation with clearance rates 50-60% at one year; topical trichloroacetic acid; imiquimod with 40% resolution vs 8% control; and surgical anal mucosectomy), that argue for the reevaluation of anal cytological screening.
While cost-effectiveness is always a factor in screening programmes the presentation made the following points:
- The first cost effectiveness study reported that cervical screening (3-yearly) in HIV negative women was estimated at costing USD$180,000 per life year saved. This compared to approximately USD$11,000 for anal cytology screening in HIV positive men. 
- A more detailed and recent analysis from the same group estimated costs of USD $16,600 and $13,000 per Quality Adjusted Life Year saved (QALY) for annual or two-yearly screening in HIV positive MSM. For HIV negative MSM, these costs were $34,800 and $15,100 respectively. 
- A recent UK study determined screening in HIV positive men is estimated to cost UKP£39,400, and that this is apparently not cost-effective. 
While the BHIVA/BAASH guidelines have stated that the benefit of screening is not yet proven, a more positive set of guidelines from New York State has recommended screening in HIV positive MSM, HIV positive CIN/VIN, and HIV positive people with a history of genital warts, although these recommendations are unlikely to be running in practice due to the demand this would place on anoscopy services.
For any screening to be effective it will be dependent on providing timely anoscopy follow-up for patient with abnormal cytology results.
Although the National Institute for Health Research (NIHR) has commissioned a new review to look at screening for anal cancer in the UK (currently in press), the degree to which HIV is addressed is unclear. 
This is difficult because an earlier review highlighted important reasons for making screening available to HIV positive people and MSM. This earlier report was produced in 2003 and has not been reviewed since. 
The UK cost-utility study from Karnon and colleagues is worrying for many reasons, especially if this is used as evidence against initiating a pilot screening programme. This paper has a complicated methodology, and some assumptions in their modeling are not fully explained.
- Mortality associated with anal cancer is based on data from 1996.
- All patients diagnosed with HIV before 1990 are assumed to have died by 2005.
- The disutility weighting for false-positive results is not clearly explained.
- It is odd to find worse outcomes from any screening programme, especially given the high relative risk in HIV-positive MSM. In this paper, no scenario produced better results from a screening programme compared to no screening.
- The paper uses triangular distribution and markedly different results could come from different distributional forms.
- In the explanation of calibration methodology, the rating scale used is not generally seen as most methodologically robust, and time trade off and standard gamble may be more appropriate.
- In the discussion, the authors recognise that they assumed higher HIV mortality than the US study, but they do not specify the rate used. They concede that this would reduce the effectiveness of screening, so this is an important figure to elaborate. They also do not state basis of expected mortality data, though it is used in several parts of the model, including reducing the effectiveness of screening in older MSM as they have fewer QALYs to gain.
- Finally, the paper makes a strange point about how MSM in the UK may not use a screening programme. This is supported by a reference to a study that reported that people get safer sex information from friends rather than clinics. The rationale for this is not clear. As people primarily use clinics for monitoring and treatment, accessing prevention information seems an inappropriate marker to use.
Unless stated otherwise, all references are to the Programme and abstracts from the 12th European AIDS Conference (EACS), 11-14 November 2009, Cologne.
- Bower M. Screening for non-AIDS malignancies: if and how? 12th EACS, 11-14 November, 2009, Cologne. Webcast:
- Bower M et al. British HIV Association guidelines for HIV-associated malignancies (2008).
- Grulich AE et al. Incidence of cancers in people with HIV/AIDS compared with immunosuppressed transplant recipients: a meta-analysis. Lancet 2007; 370: 59-67. (7 July 2007). doi: 10.1016/S0140-6736(07)61050-2.
- Daling JR, Weiss NS, Hislop TG et al. Sexual practices, sexually transmitted diseases, and the incidence of anal cancer. N Engl J Med 1987, 317:973.
- Frisch et al. Human papillomavirus-associated cancers in patients with HIV and AIDS. Journal of the National Cancer Institute, Vol. 92, No. 18, 1500-1510, September 20, 2000. Free full text:
- Diamond C et al. Increased incidence of squamous cell anal cancer among men with AIDS in the era of highly active antiretroviral therapy. Sex Transm Dis. 2005 May;32(5):314-20.
- Bower M et al Highly active antiretroviral therapy and the incidence of non-AIDS- defining cancers in people with HIV infection. J Clin Oncol 2009. 27(6);884-90.
- Biggar et al. Survival after cancer diagnosis in persons with AIDS. JAIDS 2005;39:(3):293-299. Free full text:
- Palefsky J et al. Progression of high-grade anal intraepithelial neoplasia to invasive anal cancer among HIV+ men who have sex with men. 16th CROI, 2009. Poster abstract 867.
- Palefsky J et al. Natural history and clinical management of anal human papillomavirus disease in men and women infected with HIV. Clin Inf Dis 2002;35:1127-1134.
- Goldie SG et al. The clinical effectiveness and cost-effectiveness of screening for anal squamous intraepithelial lesions in homosexual and bisexual HIV-positive men. JAMA. 1999;281:1822-1829.
- Goldie at al. Cost-effectiveness of screening for anal squamous intraepithelial lesions and anal cancer in human immunodeficiency virus-negative homosexual and bisexual men. Am J Med 2000; 108(8): 634-41. (1 June 2000).
- Karnon J et al. Cost-utility analysis of screening high-risk groups for anal cancer. Journal of Public Health 2008 30(3):293-304
- Czoski-Murray C. What are the pros and cons of screening high risk populations for anal cancer? NIHR NTA Study.
- Structured review for the UK National Screening Committee – appraising the viability, effectiveness and appropriateness of an anal cancer screening programme (2003). Direct download: