Stier, Elizabeth A MD*; Goldstone, Stephen E MD†; Berry, J Michael MD‡; Panther, Lori A MD§; Jay, Naomi PhD‡; Krown, Susan E MD∥; Lee, Jeannette PhD¶; Palefsky, Joel M MD‡
Invasive anal cancers are a significant cause of morbidity among HIV-infected individuals. Like cervical cancers, anal cancers are believed to arise from intraepithelial lesions caused by oncogenic human papillomavirus (HPV). The marked decrease in cervical cancer incidence and mortality rates over the past 4 decades has been attributed to detection of preinvasive high-grade squamous intraepithelial lesions (HSILs) by cervical cytology and to treatment of these lesions, thereby preventing progression to invasive cervical cancer.1 In recent years, methods used to detect preinvasive cervical lesions have been adapted to evaluate the anal epithelium, namely, anal cytology2 and high-resolution anoscopy (HRA).3 These sensitive methods have detected HSILs of the anal canal in up to 52% of HIV-infected men who have sex with men.4
Because cervical and anal epithelia show similar biology, it has been hypothesized that detection and treatment of HSIL of the anus may prevent progression to invasive anal cancer. The most effective treatment for HSIL of the cervix-excision of the transformation zone of the cervix with a conization procedure-is not an option for treatment of anal lesions, because excision of large sections of the anal canal may be associated with anal stenosis, abscess, anal spasm, and dyschezia.
Cervical dysplasia can also be treated with ablative procedures such as laser or cryotherapy. In many centers, targeted excision and/or ablation with a laser, cryotherapy, or cautery is also standard treatment for HSIL of the anus. Chang et al5 prospectively evaluated the surgical outcomes of patients with large-volume HSILs treated with a combination of excision and cauterization. Although tolerance of the procedure was reported to be generally good, approximately half of the patients reported “uncontrolled pain” for an average of 2.9 weeks. In addition, of the 29 HIV-infected patients, 23 (79%) showed recurrent HSILs after a mean of 12 months.
The infrared coagulator (IRC), which is approved by the US Food and Drug Administration for the treatment of hemorrhoids, tattoo removal, chronic rhinitis, and genital warts, has been suggested as an alternative method for outpatient treatment of anal HSIL.6 The IRC relies on the delivery of short pulses of a narrow beam of visible and infrared light through a small contact tip applicator that is applied directly to the target tissue, resulting in thermal coagulation and tissue necrosis. The depth of coagulation and tissue necrosis is precisely adjustable.
Goldstone et al6 reported a retrospective analysis of treatment with the IRC of anal HSILs in 68 HIV-infected men. Seventy-two percent of lesions treated with the IRC were without recurrence for at least 2 years, and no complications were reported.
Herein, we report the first prospective, multicenter, open-label pilot study of the IRC to treat anal HSILs in HIV-positive men and women. The aim of the study, which was developed by the HPV Working Group of the AIDS Malignancy Consortium (AMC), was to determine the safety and tolerability of the IRC and to estimate the activity of the IRC as a treatment for HSILs of the anal canal.
MATERIALS AND METHODS
Eligible patients were adult men and women at least 18 years old with serologic documentation of HIV infection and biopsy-proven internal anal HSIL documented by HRA. Patients were required to have no more than 3 discrete intra-anal HSIL lesions with positive margins and a maximal diameter of 1 cm. Subjects whose biopsies showed evidence of microscopic invasion or those with a history of anal cancer were excluded.
Patients were required to be on a stable highly active antiretroviral therapy (HAART) regimen for at least 12 weeks before study enrollment. Female patients of child-bearing potential were required to have a negative pregnancy test result within 72 hours of study entry. Patients with history of severe photosensitivity reactions were excluded.
Subjects were enrolled at AMC sites where the infrastructure was available to evaluate and treat anal intraepithelial neoplasia (AIN). Treating clinicians at each site were certified according to standards developed by the HPV Working Group of the AMC. Certification included a site visit by a designated member of the AMC HPV Working Group to observe the clinician performing HRA and using the IRC on appropriate patients. HRA was performed in accordance with the technique described by Jay et al.3
All patients gave written informed consent in accordance with human experimental guidelines of the Department of Health and Human Services and the human investigations committees at each of the participating sites.
Study Design and Treatment
Patients underwent a complete history and physical examination within 4 weeks of study enrollment. In men, this included evaluation of the penis and scrotum for lesions, discharge, tenderness, testicular asymmetry, and masses. A complete genital examination in women included a baseline colposcopic evaluation of the cervix, vagina, and vulva. All patients underwent a digital rectal examination, which included assessment of anal sphincter tone.
HRA of the anal canal was performed at baseline to document lesion size and location. Locations on the anal canal were described and charted by position (anterior, posterior, left lateral, right lateral, right anterior, left anterior, right posterior, and left posterior.) Because the circumference of the anus is roughly 12 cm, designation of lesion location by these 8 sectors permitted reliable tracking of lesions at each follow-up visit.
The following laboratory tests were obtained within 30 days before treatment: anal biopsy diagnostic of an HSIL, swab for anal cytology and anal HPV test, complete blood cell count (CBC) with differential and platelet count, and HIV RNA level and CD4 T-cell count and percentage.
At the initial study visit and all subsequent follow-ups, HRA was performed and HSILs were reidentified using photographs and charts from the patient's prior visit. Ablation of intra-anal lesions with the IRC was performed using the technique described by Goldstone et al.6 Briefly, each HSIL was infiltrated with up to 1 mL of a 1% lidocaine solution with epinephrine, first above and then distal to the dentate line. After achieving adequate anesthesia, the IRC light guide was applied to the lesion with gentle pressure and fired in 1.5-second pulses. Successful tissue necrosis was marked by tissue blanching. The light guide was then repositioned over another portion of the lesion until the entire lesion had been treated. Debridement was then done to remove the eschar, and the lesion was retreated with the IRC. These steps were continued until reaching the level of the submucosal vessels. The vessels were then coagulated with the IRC. All lesions were treated at a single visit. No patients received intravenous anesthesia, and antibiotics were prescribed only for patients at risk for endocarditis.
Study subjects were discharged from the office and instructed to take stool softeners and over-the-counter pain relievers or oral prescription narcotics as needed. Patients were advised to avoid anal receptive intercourse for 2 weeks. A telephone nursing evaluation was done 4 weeks after the IRC procedure.
Study subjects were re-examined every 3 months for 1 year. At each visit, the clinician took a history and performed a focused physical examination, a Papanicolaou smear for anal cytology, and HRA. Any apparent HSILs were biopsied. If no abnormal lesions were noted, a biopsy of the treated site(s) was performed at the 6- and 12-month visits. Anal HPV testing was done at the 6- and 12-month visits. A CBC, CD4 and CD8 T-cell count, and measurement of the HIV-1 plasma RNA level were performed at each visit. For the latter, the lower limit of sensitivity used at some study sites was 75 copies/mL, and this level was used to distinguish detectable from undetectable HIV RNA levels.
Response to treatment was assessed by cytology and histology at each follow-up evaluation. If the biopsy and cytology results were normal or showed a low-grade squamous intraepithelial lesion, the lesion response was considered complete at that visit. If the biopsy showed an HSIL from a treated area at the first post-IRC treatment evaluation, the lesion was classified as persistent. If a treated area showed a complete response at a follow-up visit and then showed an HSIL at a later evaluation, the lesion was classified as recurrent. If an HSIL developed in a previously untreated area, that lesion was classified as metachronous.
Patients with persistent HSILs at the 3-month follow-up evaluation were offered a second treatment with the IRC within 4 weeks.
Patients who had a biopsy-proven HSIL at a previously treated area of the anus at any visit at 6 months or later were classified as treatment failures.
Metachronous lesions were treated at the discretion of the investigator. These lesions were not followed for response and were not evaluated in this study.
This study utilized the Cancer Therapy Evaluation Program Common Terminology Criteria for Adverse Events (CTCAE), version 3.0, for toxicity and adverse event reporting.7 A grading (severity) scale was provided for each adverse event term.
HPV testing from anal swab specimens was performed as described previously using a polymerase chain reaction with L1 consensus primers and probes specific for 29 individual HPV types and a mixture of 10 other types.8 HPV viral load was determined in a validated semiquantitative manner by estimating the size of the dot-blot signal with consensus probes and type-specific probes (Palefsky et al, personal communication).
The primary aim of this pilot study was to evaluate the safety of the IRC. A secondary aim was to determine whether there was sufficient evidence for the efficacy of the IRC to warrant further study. Because published studies of the treatment of HSILs in HIV-infected patients reported recurrence rates of 70% or more, we considered that a well-tolerated office procedure that achieved a recurrence rate of <50% would be an improvement. A sample size of 15 patients was chosen to test the null hypothesis that the recurrence rate at 1 year is ≥80% versus the alternative hypothesis that it is ≤50% at the 1-sided 0.06 significance level with a power of 0.85. With 15 patients, the null hypothesis would be rejected if ≤9 patients had a recurrence. To ensure that 15 patients could be assessed for efficacy, the study planned to enroll 18 patients. Descriptive statistics were used to describe the patient characteristics, disease outcomes, incidence of adverse events, and HPV test results.
Twenty patients enrolled in the study from January through June 2004. Two patients (patients 7 and 9) were ultimately found to be ineligible, did not receive treatment, and were excluded from all analyses. The 18 treated patients, 16 men and 2 women, were enrolled at 4 AMC sites. Baseline characteristics are shown in Table 1. The mean age was 44 (range: 32 to 53) years, and 67% of the patients were white. The median HIV RNA level was 75 (range: <50 to 500,000) copies/mL, and the mean CD4 count was 581 (range: 192 to 1065) cells/μL. Twelve of the 18 patients had HIV RNA levels <75 copies/mL. Sixteen patients had been on their current HAART regimen for more than 12 months, and the other 2 patients had started their current regimen within the past 4 to 6 months. Eight of the patients had previously received treatment for anal dysplasia, including 6 patients who had surgical treatment (with excision, cautery, and/or laser), 1 patient who received the experimental SGN-00101 Hsp-E7 therapeutic vaccine, and 2 patients who were treated with topical imiquimod (1 in conjunction with surgery).
The IRC was generally well tolerated. Twelve patients reported mild or moderate adverse events related to the procedure (Table 2). Four patients reported grade 1 anal discomfort, and 6 reported grade 2 discomfort. Eleven patients reported grade 1 anal/rectal bleeding, and 1 patient had grade 2 bleeding. Two patients reported short-lived grade 1 anal incontinence within a few days to several weeks after the procedure; neither patient was troubled by these episodes, which resolved spontaneously. There were no grade 3 or 4 procedure-related adverse events reported.
Response of HSILs
Of the 16 patients who were followed for 1 year, 6 (37.5%) had recurrent or persistent disease (95% confidence interval: 15.2% to 64.6%). Thus, the null hypothesis that the recurrence rate is 80% is rejected (1-sided P = 0.0002; Fig. 1 Table 3).
Of the 18 patients, 8 showed a complete response at the 3-month visit and 10 had persistent HSILs and were candidates for a second ablative treatment with the IRC. Eight of the patients with persistent HSILs were retreated, and 2 were lost to follow-up.
Overall, at 1 year, 10 (62.5%) of 16 patients showed a complete response. Of the remaining 6, 4 had persistent HSILs after the second treatment and 2 had recurrences after an initial response (at 6 and 13 months). If the 2 patients who were lost to follow-up were treated as treatment failures, 10 (56%) of 18 patients showed a complete response.
Forty-four HSILs were treated on study. On average, 2.2 HSILs per patient were treated with the IRC at the initial visit. Of these HSIL lesions, 15 (34%) showed persistent HSILs at the 3-month visit (ie, complete response rate of 66% at 3 months on a per-lesion basis).
Five of 18 subjects had concordance between the Papanicolaou cytology and histologic pathology findings at the baseline visit (ie, cytology and histology showed HSILs). Although all patients had HSILs by biopsy, most patients had minimally abnormal Papanicolaou test results (6 atypical squamous cells of undetermined significance [ASCUS], 4 low-grade squamous intraepithelial lesions), 1 patient had normal cytology, and 2 patients had minimally abnormal Papanicolaou test results that were considered suspicious for a higher grade lesion.
There was no correlation between changes in HIV RNA viral load or CD4 cell count throughout the study and patient response (data not shown). Of the 3 patients with CD4 counts <250 cells/μL at enrollment, 1 showed a complete response and 2 patients who showed persistent HSILs at the 3-month visit were lost to follow-up.
Effects on HPV
HPV 16 was the most common HPV type identified and was detected in 11 patients during at least 1 visit (see Table 3). HPV 16 was detected at any time in 4 of 6 nonresponders and in 5 of 10 responding patients. There was no clear correlation between HPV type or intensity and persistence, clearance, or recurrence of HSILs. There was no consistent change in HPV type or viral load in patients before and after treatment with the the IRC.
Although the incidence of several AIDS-associated cancers has declined since the introduction of HAART, the incidence of anal cancer has continued to rise in the HAART era.9,10 Screening with anal cytology and HRA frequently detects HSILs, which have the potential to progress to invasive anal cancer. Standard procedures for the treatment of HSILs are associated with significant postoperative pain and a high risk of recurrent lesions, and therapeutic methods with less morbidity and increased efficacy are needed.
In this study, we confirm the report by Goldstone et al6 that treatment of HSILs of the anal canal with the IRC in HIV-infected patients is safe and well tolerated. Reported adverse events were mild or moderate and were described by patients as being of minimal concern.
Our study differed in several important respects from that of Goldstone et al.6 That analysis was retrospective, patients were re-evaluated by HRA every 6 months (although they did undergo standard anoscopy at 3 months after the procedure), repeat biopsies were done only of lesions that appeared suspicious on HRA, metachronous lesions were treated, and patients could be retreated more than once. In our prospective study, patients were evaluated by HRA every 3 months, repeat biopsies were performed at treated sites to assess for recurrence even in the absence of any apparent lesion, lesions could only be retreated once at the first 3-month follow-up visit, and metachronous lesions were not followed. Given these differences, the efficacy of treatment at 3 months is the only point that can be directly compared between the 2 studies. In our study, 66% of lesions showed a complete response to the initial IRC treatment at 3 months, which is nearly identical to the 70% response rate reported by Goldstone et al6 on a per-lesion basis.
The variance between cytology and histology illustrates that although cytology is a sensitive test to detect AIN in HIV-positive men, the concordance with respect to actual grading of lesions on cytology and histology is poor, as it is in the cervix.1,2 For this reason, we have advocated using cytology as a screening test to determine who needs referral for HRA-guided biopsy, because the treatment plan is always based on the latter. Only 11% (3 of 28) of patients had normal cytology with a documented HSIL, which is consistent with the earlier literature. After IRC treatment, we would recommend that patients be followed with a combination of cytology and HRA, but we cannot yet recommend screening intervals or duration of screening based on this pilot study.
Given that the IRC only ablates the lesions and not the apparently normal anal mucosa, it is not surprising that we observed no change in HPV type or intensity after IRC treatment. This study only included patients with discrete lesions of limited size, and as such, only a small proportion of the anal mucosa was ablated. Because HPV also resides in histologically normal tissue, we would expect that only the HPV associated with the ablated lesions might have been eradicated. Because anal dysplasia and cancer are caused by HPV, the persistence of HPV, even in the absence of anal dysplasia, necessitates long-term follow-up, because these patients continue to be at risk for developing metachronous HSILs. Data on metachronous lesions were not collected in this study; we plan to look at this in a future study with a larger number of participants.
This study is also significant because it represents the first treatment study of anal HSILs that was not performed primarily by surgeons. Six patients were treated by a general surgeon, 4 by a gynecologist, and 8 by nonsurgically trained clinicians. All these clinicians were trained and certified in HRA and the IRC. There were no obvious differences in adverse events or outcomes among clinicians in this small clinical trial. The demonstration that these procedures can be successfully and safely performed by nonsurgeons, given adequate training and supervision, could be important for ensuring patient compliance with treatment recommendations and follow-up, however, because this procedure allows clinicians performing HRA to treat their patients without referral to a surgeon.
The IRC is not the only potential office-based nonsurgical treatment for HSILs of the anus. Additional possibilities include liquid nitrogen ablation or topical applications of trichloroacetic acid or imiquimod,11 but large studies evaluating the efficacy of these methods have not been published.
In conclusion, the IRC is a well-tolerated method for treating discrete anal canal HSILs in HIV-infected patients. Practitioners from a variety of disciplines were able to perform this procedure safely with appropriate training. Because this pilot study met the predetermined criterion of a response rate of >50% at 1 year, a larger study of the IRC for the treatment of HSILs in HIV-infected patients is warranted to evaluate the efficacy of treatment.
© 2008 Lippincott Williams & Wilkins, Inc.