Anal cancer is a growing problem in HIV-positive patients and the prevalence has been rising following the introduction of effective therapy for HIV, a consequence of increased longevity . In the Chelsea, UK, and Westminster, UK, cohort of 6127 patients from which most of the patients in this study were drawn, the incidence of anal squamous carcinoma was 141 per one million patients per annum for the years 2002–2007, with an average of four cancers a year . Ninety percent of this cohort is male and most are MSM. These data lend urgency to the need to define optimum treatment modalities for anal intraepithelial neoplasia (AIN).
The reported outcomes for most published treatment studies have been disappointing [2,3]. Imiquimod as a candidate treatment option has two distinct advantages: the possibility of self-application and usage in patients with extensive disease. A placebo-controlled study using imiquimod twice weekly for 16 weeks for vulvar intraepithelial neoplasia (VIN2/3) showed clearance at 12 months in 35% of those on active drug, and improvement in 81%, with no improvement in controls .
Imiquimod is not licensed for use in or around the anus, but prior to commencing this study, the authors had experienced some success in the treatment of anal canal warts with imiquimod and this treatment appeared to be reasonably well tolerated. Sanclemente et al.  subsequently reported reasonable efficacy in the use of imiquimod to treat perianal warts in HIV-positive patients, with greater responses after 20 weeks than after 16 weeks of treatment. If too much cream is inserted too far into the anorectum, systemic absorption of the drug can lead to transient flu-like symptoms the following day. There has also been an uncontrolled study of using imiquimod in HIV-positive MSM with predominantly external AIN . This paper will be discussed below.
By analogy with cervical lesions, some spontaneous resolution of low-grade squamous intraepithelial lesion (LSIL) is to be expected, but spontaneous resolution of high-grade squamous intraepithelial lesion (HSIL) has been regarded as a rare event in HIV positives, even when effective immune reconstitution has occurred. We decided for these reasons to focus our attention on the treatment of HSIL. The aim of this study was to determine whether imiquimod was more effective than placebo for the treatment of anal canal HSIL.
HIV-positive MSM who were receiving follow-up for histologically proven HSIL were recruited from two anoscopy clinics. The local research ethics committees of the two hospitals approved the study where the study was conducted. All patients were required to have been on HAART for a minimum of 3 months prior to recruitment, to have a CD4 cell count of 100 cells/μl or higher and to have no history of previous imiquimod use within the anal canal. It was initially planned to recruit 120 patients with a one-to-one randomization to imiquimod or placebo. This data was based on a sample size calculation: a regression to low-grade or normal in 25% of patients in the treatment group and 5% in the control group would require 116 patients in order to achieve a 5% significance and 80% certainty. If regression occurred in 40% of patients on imiquimod only, 50 patients would be required to achieve 5% significance and 90% certainty. If 30% of recruited patients dropped out or were lost to follow-up, then the power to detect a difference of the same magnitude would still have been acceptable at 80%.
Matching packs of imiquimod and placebo were supplied by 3M Health Care Ltd, Loughborough, UK. On recruitment, patients were allocated the next randomization number in the study. The randomization sequence was known only to 3M Health Care and the study was not unblinded until all patients had completed the randomized stage. The ongoing presence of HSIL was confirmed and mapped visually using the technique of high-resolution anoscopy (HRA) and confirmed histologically just prior to commencement. Patients were instructed on how to self-apply the cream three times a week by inserting a finger no further than 2 cm into the anal canal and not using more than half a sachet to reduce the possibility of significant systemic absorption. A diary was kept of treatment applications and any side-effects. Patients experiencing soreness were instructed not to apply further cream until symptom free, and where symptoms were marked, the dosage was reduced. The duration of treatment was 4 months, and at the end of this period microscopic visualization using HRA was repeated, with mapping of any remaining lesions. Anal cytology, HRA with mapping and biopsy as indicated, were repeated 8 weeks after the end of treatment. The method of cytology, HRA and histology were as previously described . The key diagnostic slides were seen by a single pathologist and reviewed by a second pathologist. Where there was disagreement in classification, the most abnormal finding was used. Improvement was defined as clearance of AIN or downgrading to LSIL by cytology and HRA at 6 months following commencement, and sustained one year of follow-up. All participants were offered ongoing 6-monthly HRA and cytology. Partial and nonresponders were all offered 4 months of open-label treatment with imiquimod and ongoing 6-monthly surveillance.
Statistical methods the baseline demographics and clinical responses of the imiquimod and placebo groups were compared using the unpaired t-test for geometrically distributed data; the Mann–Whitney U test was used where the data were hyper geometric, and the χ2 test with Yates' correction was used for qualitative data.
Recruitment commenced in October 2002 and was terminated in April 2005 owing to inability to obtain further supplies of randomized study drug after 64 patients had consented. The clinical trial finished in June 2006. There were 11 patients who did not complete the study and defaulted from follow-up. One is known to have developed an anal squamous carcinoma during the treatment phase on the placebo arm. The others were all contacted to ascertain their reason for withdrawal. Five had not commenced therapy and the remaining five (three of whom were on placebo) took only a few weeks treatment and then stopped. Only one patient discontinued because of side-effects. The commonest reason for withdrawal was unhappiness with the 50% probability of receiving placebo.
Of the 53 patients who completed the study, 28 were on active drug and 25 on placebo.
There were no statistically significant differences in baseline demographics between the two groups (Table 1). The mean duration of follow-up from completion of treatment was 36 months (Table 2).
In the imiquimod group, four patients resolved and remained disease free after a median 33 months of follow-up. One went on to develop incident warts and HSIL after 42 months of follow-up, probably because of a new human papillomavirus (HPV) infection. This patient after imiquimod was reintroduced showed a sustained improvement to LSIL and resolution of warts. Further, a patient who showed complete resolution at completion of treatment, and also 3 months later, was classed as a treatment failure because HSIL was detected after 8 months of follow-up.
Eight patients downgraded to LSIL, of whom one subsequently resolved and remained clear 48 months later. The other seven had no further HSIL after a median 33 months of follow-up, but one developed incident HSIL after 44 months. There were 16 nonresponders in the treatment group, of whom two were lost to follow-up after 6 months and received no further treatment. Following further treatment, only eight continued to have HSIL after a median 43 months.
In the placebo group, one patient resolved and remained clear 37 months later.
Imiquimod, therefore, was significantly associated with a positive outcome, combining those who cleared and those who downgraded, χ2 = 8.78 applying Yates' correction with one degree of freedom, P = 0.003. There was no significant difference in CD4 cell count rise in patients who responded compared with those who did not. The mean CD4 cell count at the time of diagnosis of ACIN was 391 cells/μl in the responder group and at the time of response was 459 cells/μl, whereas in the nonresponders, the mean CD4 cell count at diagnosis was 380 cells/μl and at the end of follow-up it was 486 cells/μl. Virtually all subjects had HIV RNA viral loads below 50 copies throughout the period of study.
Twenty-one patients opted to use open-label imiquimod, of whom 12 were from the placebo arm. Five of these patients showed complete resolution and four downgraded to LSIL, of whom one subsequently cleared. One of the five patients who was clear at the end of open-label treatment had newly incident HSIL after 29 months. Further, a patient was apparently clear but was found to have HSIL after 12 months and was, therefore, counted as a treatment failure. Median follow-up post improvement or resolution was 36 months. Most patients who did not respond received further treatment with trichloroacetic acid (TCA) or laser and further courses of imiquimod, and another two downgraded to LSIL as a consequence.
In summary, of the 54 patients for whom there is follow-up data, two developed squamous carcinoma of the anus and another two exhibited spontaneous resolution of HSIL. Thirteen patients have cleared following treatment, of whom four had newly incident HSIL at 23, 29, 44 and 48 months. Sixteen have had a sustained downgrading to LSIL following treatment, one of whom showed incident HSIL after 29 months. The two patients who resolved apparently spontaneously may well have been stimulated to do so by the repeated biopsies. Putting aside for a moment the reasons for late reappearance of HSIL, the total number of patients who have exhibited prolonged absence of HSIL following treatment with imiquimod is 29 out of 47 patients (61%).
The common finding of all treatments for AIN has been high-recurrence rates in HIV-positive patients, a consequence of viral persistence. The only previously published study of using imiquimod for AIN focussed on the perianal region and showed a 61% end-of-study clearance after 16 weeks of treatment . These impressive results are not strictly comparable because nearly a third of the patients treated had LSIL only and few had internal disease. In fact, only four of the patients had high-grade anal intraepithelial neoplasia (HG-AIN). Three of them downgraded to LSIL, whereas the remaining patient who failed to improve was poorly adherent to the treatment regime. Overall 21% of the patients had poor adherence, and it is worth noting that a whole sachet of cream was being used per application, with a high rate of soreness recorded, in contrast to our study in which the dosage was halved. From the same study, 19 patients were followed up for a mean of 30 months . Twenty-six percent experienced recurrent HSIL after a mean time of 24 months, whereas a further 31% developed cytological evidence of LSIL in the anal canal. The virological data from the study is extremely interesting in that the majority of patients who initially responded showed a significant reduction rather than a clearance in the viral load of the detected oncogenic HPV types. These viral loads remained low throughout the extended follow-up, and most ‘recurrent’ disease was associated with the acquisition of new oncogenic HPV types. This phenomenon may also explain the late reappearance of HSIL in some of our patients who were otherwise doing well on their treatment for HIV.
Another observational cohort has also reported the long-term outcome of relentlessly treating AIN lesions with a mixture of modalities, predominantly laser, but including also imiquimod and surgical excision . Clearance was defined as being disease free for 12 months and on this basis 63% of patients were cleared after being treated for a median of 31.5 months. The median time to cure was longer for patients with HSIL (37.5 months) than with LSIL (30.4 months), but the difference was not statistically significant. The time to cure was significantly dependent on the area of abnormality and lesions took considerably longer to clear in HIV-positive patients (39 months) than in HIV-negative (25 months).
In the course of the current study, some interesting phenomena were observed. Patients treated with imiquimod for a combination of warts and HSIL sometimes cleared these different lesions at different times, and sometimes the HSIL cleared before the warts. Internal and external AIN do not necessarily clear simultaneously, and it should not be assumed that perianal treatment alone is sufficient. LSIL and HSIL commonly coexist, but just as with warts and HSIL, the two types of lesion can clear independently. This should not surprise us, as in the anus LSIL is more likely to be caused by nononcogenic than by oncogenic HPV . For this reason, the successful clearance of HSIL, leaving only LSIL, should constitute a considerable reduction in the risk of anal cancer.
This study has demonstrated that imiquimod can play a valuable role in the management of AIN, that it can be well tolerated when used in the manner described and that there is value in prolonged and repeated courses of treatments, especially when combined or perhaps alternated with other treatment modalities. Now that longer term follow-up studies have shown the efficacy of this approach, one of the barriers to advocating AIN screening for all HIV-positive patients has been removed. It is true that we cannot yet prove that these interventions will prevent anal cancer, but we at least know that precursor HG-AIN can be cured. This is certainly not going to prevent all anal cancers, but it would be a grave error to await the outcome of long-term natural history studies. Although much more research is needed, the time has come for all HIV centres to begin screening for and treating AIN.
The authors would like to thank 3M Health Care Ltd for providing matching boxes of imiquimod and placebo, and for supervising the randomization, the patients who volunteered, and Sundhiya Mandalia for her statistical assistance.
P.F., M.N., S.B. and M.B. were involved in the design of the study. P.F. and M.N. recruited the patients and collected all the samples. N.S. reported all the cellular pathology specimens from Homerton Hospital, reviewed independently by N.F. N.F. reported all the cellular pathology specimens from the Chelsea and Westminster Hospital, from where the biopsies were reviewed by J.W. and the cytology by G.D. P.F. wrote the paper, receiving contributions from M.N., S.B. and M.B.
1. Powles T, Robinson D, Stebbing J, Shamash J, Neslon M, Gazzard B, et al. Highly active antiretroviral therapy and the incidence of non AIDS defining cancers in people with HIV infection. J Clin Oncol 2009; 27:884–890.
2. Fox PA. Human papillomavirus and anal intraepithelial neoplasia. Curr Opin Infect Dis 2006; 19:62–66.
3. Fox PA. Anal cancer screening in men who have sex with men. Curr Opin HIV Med 2009; 4:64–67.
4. van Seters M, van Beurden M, Kate JW, Ewing P, Eijkemans M, Kagie M, et al. Treatment of vulvar intraepithelial neoplasia with topical imiquimod. N Engl J Med 2008; 358:1465–1473.
5. Sanclemente G, Herrera S, Tyring SK, Rady PL, Zuleta J-J, Correa L-A, et al. Human papillomavirus viral load and HPV type in the clinical outcome of HIV positive patients treated with imiquimod for anogenital warts and anal intraepithelial neoplasia. J Eur Acad Dermatol Venereol 2007; 21:1054–1060.
6. Wieland U, Brockmeyer NH, Weissenborn SJ, Hochdorfer B, Stücker M, Swoboda J, et al. Imiquimod treatment of anal intraepithelial neoplasia in HIV positive men. Arch Dermatol 2006; 142:1438–1444.
7. Fox PA, Seet JE, Stebbing J, Francis N, Barton SE, Strauss S, et al. The value of anal cytology and human papillomavirus typing in the detection of anal intraepithelial neoplasia: a review of cases from an anoscopy clinic. Sex Transm Infect 2005; 81:142–146.
8. Kreuter A, Potthoff A, Brockmeyer NH, Gambichler T, Stücker M, Altmeyer P, et al. Imiquimod leads to a decrease of human papillomavirus DNA and to a sustained clearance of anal intraepithelial neoplasia in HIV infected men. J Investig Dermatol 2008; 128:2078–2083.
9. Nathan M, Hickey N, Mayuranathan L, Vowler SL, Singh N. Treatment of anal human papillomavirus associated disease: a long term outcome study. Int J STD & AIDS 2008; 19:445–449.
10. De Vuyst H, Clifford GM, Nascimento MC, Madeleine MM, Franceschi S. Prevalence and type of human papillomavirus in carcinoma and intraepithelial neolasia of the vulva, vagina and anus: a meta-analysis. Int J Cancer 2009; 124:1626–1636.
© 2010 Lippincott Williams & Wilkins, Inc.