Squamous cell carcinoma of the anus is thought to arise from a precancerous lesion. The etiology of this precancerous lesion is thought to involve integration of HPV into the patient’s genome. Similar to cervical cancer, a Bethesda staging criteria has been devised for precursor anal lesions (13). AIN1 is thought to be low grade squamous intraepithelial lesion (LSIL) whereas AIN 2, 3 are high grade squamous intraepithelial lesion (HSIL). Similar to cervical cancer, treatment is recommended for high grade precancerous (HSIL) anal lesions.

Studies have identified additional risk factors in the development of AIN. Wilkin et al (2004) studied the risk of developing AIN in HIV+ men (14). Almost three-quarters of men with abnormal anal cytology had co-infection with a high risk HPV serotype (HPV 16>>52>18>45) (14). Multivariate analysis indicated that abnormal cytology was more likely in patients with a history of RAI and no HAART treatment. AIN histology on biopsy was more likely in patients with history of RAI, history of no HAART use, young age (<40) and low CD4 count (<350). CD4 count was the most significant prognostic factor. Patient who were on HAART and had persistent low CD4 counts were also more likely to have AIN. The relationship between AIN and HAART use, CD4 count, and viral load is probably confounded as patients with lower CD4 counts are more likely to have high viral loads and to be started on HAART

Recently physicians have begun to advocate anal cancer screening in the high risk HIV+ population. Studies support the high incidence of precancerous lesions in the HIV+ population. Wilkin et al. (2004) analyzed the prevalence of anal precursor lesions in HIV+ men and reported that almost ½ of patients had abnormal cytology on the anal Pap smear and subsequently 40% of these patients were found to have AIN histology by biopsy (14). Kreuter et al (2010) prospectively examined a population of 400 HIV+ MSM over a period of 5 yrs and determined that over 70% had some degree of AIN (10). 35% had high grade AIN and 2.5% had anal cell cancer detected on screening. More importantly Kreuter et al (2010) demonstrated that untreated AIN can progress to anal cancer in as little as 8 months (10). Previous studies in the mid 1990’s had showed AIN progressing to anal cancer over 3-5 years (14, 15). Also studies indicate that that the incidence of AIN has increased with the widespread use of HAART in the HIV+ population (15).

The feasibility of screening for anal cancer has been research extensively over the past decade. New York State has established screening guidelines for anal cancer in HIV+ patients (16). They recommend that all HIV+ patients undergo screening for anal cancer and propose a similar screening scheme to cervical cancer. Initially patients should have an annual visual and digital rectal exam plus an anal PAP. If PAP reveals abnormal anal cytology then a highresolution anoscopy (HRA) should be performed similar to the colposcopy in cervical cancer. One caveat to anal cancer screening is that while the test is sensitive it is not specific. Both Palefsky et al (1997) and Goldstone et al (2001) showed that over 70%-90% of HIV+ MSM had some abnormal cytology on anal pap (17, 18). The correlation of abnormal pap with HSIL biopsy was poor. Therefore, all lesions noted on HRA should be biopsied. If HSIL is detected treatment should be offered, either medical ablation or surgical excision. If LSIL is detected the recommendation is to have repeat anal pap smears in 3-6 months. If persistent abnormal pap, these patients should have yearly HRA.

Mount Sinai implemented this practice in 2007 for all HIV+ patients (19). Researchers believe such practices are both cost effective and efficacious. Goldie et al (1999) performed a cost analysis on screening for AIN and found that screening increased quality-adjusted life expectancy for all HIV+ patients (20). Goldie et al (1999) calculated that screening with anal pap tests every year around time of diagnosis of HIV resulted in an incremental cost-effectiveness ratio of $16,600 per quality-adjusted life year saved (20). Screening more frequently than yearly did not provide any additional benefit.

Once HSIL histology is confirmed, there are a couple of treatment options. However there is still debate on the most efficacious treatment for precursor anal lesions. A study by Chang et al (2002) showed that surgery for treatment of AIN was effective in preventing recurrence of AIN in 8 HIV negative men studied but only effective in 23/29 HIV positive patients (21). Singh et al (2009) evaluated the efficacy of nonsurgical treatments (trichloroacetic acid, i.e. TCA) (22). Overall TCA worked well in younger patients (<48yo). For HIV+ patients specifically, those with two or fewer HSIL lesions responded the best. 32% had no residual lesion on follow up. HIV negative patients had a much better chance of clearance of AIN lesions than HIV+ patients. 75% of HIV+ patients had recurrence after clearance of the initial AIN lesion treated with TCA within 6 months suggesting that close follow up is needed in this high risk population (22). The treatment of AIN with surgery or with non surgical methods such as TCA is not without morbidity. Studies do show a low incidence of morbidity with possible side effects such as fibrosis and anal sphincter stenosis (3).

The standard treatment for squamous cell carcinoma of the anus is concurrent mitomycin C (MMC), 5-fluorouracil (5-FU), and RT. There have been multiple prospective randomized trials that have shown improvement in local control, disease free survival, and sphincter preservation with the addition of chemotherapy to RT (27-30, 33). There have been 4 randomized trials that have established concurrent MMC and 5-FU with RT as the standard of care. The initial UKCCR (United Kingdom Coordinating Committee of Cancer Research) trial (ACT I) compared concurrent MMC and 5-FU with RT to RT (27). RT was prescribed to 45Gy given over 4 to 5 weeks with the inguinal lymph nodes and anus included in the target. A boost was given 6 weeks later if the patient had a > 50% response or a complete response. The boost was given either with external beam (15Gy) or Ir192 implant (25Gy). Clinical response was measured at 2 months after completion of the boost. The combined modality arm consisted of the same RT schedule with 2 cycles of chemotherapy; bolus MMC (12-10mg/m2) and continuous venous infusion 5FU (1000-750mg/m2/day for 4 days) commencing at the start of RT. An update was recently reported with a median follow up of 13 years. The hazard ratio for local regional relapse (HR=0.46, p<0.001), disease free survival (HR=0.70, P<0.001), anal cancer mortality (HR=0.61, P<0.001), and colostomy free survival (HR=0.76, P=0.004) all favored the combined modality arm. The hazard ratio for overall survival was 0.86 and not statistically significant (P=0.12). There was an increase in non-anal cancer deaths in the first 5 years in the combined modality arm which almost disappeared at 10 years. Also there were more deaths due to second malignancy in the combined modality arm (P=0.03). Again acute toxicity was higher in the concurrent arm but there were no differences in late toxicity (28).

The EORTC (European Organization for Research and Treatment of Cancer) confirmed the results in a similar study comparing concomitant RT with 5FU and MMC versus RT alone (29). The locoregional control rate was improved by almost one-fifth and the colostomy free rate improved by onethird in the combined modality arm at 5 yrs. Both results were statistically significant. There was no difference between the two arms in overall survival. Again acute toxicity was worse in the combined modality arm (29).

The importance of MMC in the treatment of anal cancer was confirmed in the RTOG 87-04 (Radiation Therapy Oncology Group) trial which randomized patients to RT with MMC and 5FU versus RT and 5FU alone (30). RT total doses ranged from 45 to 54Gy. This trial showed that the MMC arm had superior colostomy free survival (71%vs 59%), disease free survival (73% vs. 51%), and fewer post treatment positive biopsies (7.7% vs. 15%) at 4years. All were statistically significant. Grade 4 and 5 acute toxicity were greater in the MMC arm (23% versus 7%) (30).

MMC is known to be toxic and the question was tested if a potentially less toxic agent, cisplatin, could be substituted. Specifically, the hemolytic uremic syndrome and thrombocytopenia are not uncommon life-threatening toxicities that result from use of MMC (31, 32). RTOG 98-11 tested concurrent MMC and 5FU versus induction cisplatin and 5FU followed by concurrent RT and cisplatin and 5FU. RT total doses ranged from 45-59 Gy. At five years the cisplatin arm had statistically significant inferior colostomy rate of 19% versus 10%. Other end points measured such as disease free survival, overall survival, local regional recurrence, and distant metastasis were not statistically different. Cisplatin was associated with less hematologic toxicity (33). The ACT II trial from the UK which also analyzed if cisplatin could replace MMC showed that there was no difference in clinical complete response or OS with either MMC or cisplatin (34). The ACT II trial also studied if maintenance therapy after definitive chemoradiation would be of benefit and reported that maintenance therapy did not improve overall survival or recurrence free survival. Followup for the ACT II trial is only at 3 yrs so further study is needed to determine if cisplatin could substitute for MMC (34). Currently concurrent 5FU + MMC with radiation is still the standard of care.

Treatment breaks due to toxicity lead to prolonged RT treatment duration and at times patients are unable to complete the planned RT. A report from Boston University Medical Center demonstrated that RT dose (>= 54Gy) was significantly associated with local control and overall survival and treatment time less than 40 days also showed a trend toward improved outcome (35). Other studies confirmed that greater overall treatment time was associated with worse outcomes and local failure (36). This is a clinical example of accelerated repopulation of residual tumor clonagens, in all likelihood (37). Roohipour et al 2008 in a multivariate analysis of 131 patients with anal squamous cell carcinoma showed the inability to complete definitive RT and disease stage at diagnosis were both predictive of relapse free survival (38). A recent analysis of 2 RTOG trials showed that for every 2 weeks of treatment interruption there was a 9.4% increase in the hazard ratio for colostomy failure (39). The success of sphincter sparing treatment is in part dependent upon getting the patient through treatment without interruptions for treatment side effects.

Early reports likely contributed to the perception that HIV+ patients do not tolerate aggressive combined modality treatment. Peddada et al (1997) in a limited study demonstrated that 8 HIV+ patients with anal cancer could be treated with concurrent 5FU/MMC but used a lower dose of RT, 30Gy instead of the standard of care > 50Gy (40). All 8 patients achieved a clinical complete response. One out of eight 8 had a grade 4 hematologic toxicity. 50% of the patients were alive and disease free. However the remaining four patients died from AIDS related illnesses in the 3 year follow up (40).

Around the same time Kim et al (George Washington University, 2001) showed that HIV+ patients had worse outcomes and tolerance than HIV negative patients in the treatment of anal cancer using the standard RT dose of 50-54 Gy and concurrent full dose chemotherapy (6). In this study, analyzed patients were from 1985 to 1998 a period of time before the advent of HAART. The HIV+ patients included patients with AIDS defining illnesses, and low CD4 counts. These patients tended to have a lower performance status. Overall 5/13 HIV+ patients analyzed failed initial definitive chemoradiation compared to 9/60 HIV negative patients. If the patients with known AIDS were removed from the analysis, the differences in treatment outcomes between the two groups are reduced. The only patients that required treatment breaks were the patients with AIDS. Each required an unscheduled 3-4 week treatment break due to grade 4 toxicity (1-skin ulcer, 1-thrombocytopenia). Furthermore, in the Kim analysis, late toxicity (poor skin healing) was higher in the HIV+ group 4/10 versus 5/33. Another early study also suggested that HIV+ patients with AIDS may not tolerate anal cancer treatment. Clinicians at the Beth Israel Medical Center in NY (1987-1991) commented on the results of treating 9 HIV+ patients, 3 of which had AIDS (41). The authors reported over 50% needed more than 2 weeks of unscheduled treatment break due to toxicity. Over 50% had grade 3-4 skin toxicity. However, 7/9 did have a clinical complete response despite the toxicity. At least one of the two patients that did not have a complete response also had AIDS. The author do mention that an early antiretroviral, zidovudine was given concurrently with chemoradiation in patients and was well tolerated.

Delineation of subgroups in the HIV+ population can help identify HIV+ patients that may not tolerate treatment from those that can tolerate standard combined modality therapy. In 1999, Hoffman et al (UCSF) published a report on a small cohort of HIV+ patients treated for anal cancer (42). Hoffman et al (1999) stratified patients based on CD4 count and showed that values greater than 200 portend to superior treatment outcomes and tolerance (42). Patients with higher CD4 counts were more likely to receive the standard of care in terms of chemotherapy and RT dosing. These authors suggested that fear of toxicity caused physicians to empirically alter chemotherapy regimens in the HIV+ population. The mean RT dose was similar between the 2 groups ~51 Gy. However even with an altered chemotherapy regimen the group with CD4 counts less than 200 still experienced more toxicity such as moist desquamation and hematologic suppression. In this group these side effects required longer treatment breaks and early treatment termination. An intact immune system appears to be critical to tolerating anal cancer treatment. A study from Emory University also found that HIV+ patients with CD4<200 did worse with anal cancer treatment (43). Of 17 HIV+ patients with anal cancer documented at Emory from 1994-2004, only those with CD4<200 were unable to complete treatment (43).

Antiretroviral drugs play a key role in controlling the HIV virus and helping bolster CD4 counts. Therapy for HIV changed dramatically in the mid 1990s with the implementation of HAART (highly active anti retroviral therapy). HAART therapy includes a combination of protease inhibitors (discovered/designed in 1995) and non nucleosidase reverse transcriptase inhibitors (1996). Widespread use of HAART came around 1999-2000. Papers discussing the use of HAART to aid in anal cancer treatment are thus limited. Hoffman et al (1999) at UCSF suggested that one patient in their cohort of 17 who initially had a CD4 count less than 200 tolerated the standard of care treatment for anal cancer due to the addition of a protease inhibitor which bolstered the CD4 count to greater than 200 (42). A later study done by Stadler et al (UT Southwestern 2004) demonstrated a trend toward improved efficacy of anal cancer treatment in HIV+ AIDS patient treated with HAART (44). Stadler et al (2004) compared outcomes in patients treated for anal cancer preHAART and post HAART (44). The UT Southwestern study differs from the other studies in that the chemotherapy used was 5FU/cisplatin instead of 5FU/MMC. The RT dose was similar at 54 Gy. In this study all patients had AIDS at time of diagnosis. Overall, 14 patients were analyzed, including 6 pre HAART and 8 on HAART. Stadler et al (2004) suggested a trend towards better treatment tolerability and outcome in patients treated with HAART. 2 year OS in patients on HAART was 67% vs. 17% in the pre-HAART era. 1yr and 3 yr mortality pre HAART was only 12% and 40% respectively compared to 67% and 80% for patients on HAART. The success of definitive treatment for HIV+ patients on HAART seems to fare the same as HIV negative patients in the randomized control trials. Moreover there was more toxicity in the preHAART patients (60%) compared to the HAART treated patients (50%) (44). It suggests that the HAART and increased CD4 count help patients tolerate treatment.

Recent single institutional studies have shown that as long as HIV+ patients can tolerate the standard of care treatment for anal cancer and do not have AIDS (i.e. CD4<200), the efficacy and durability of treatment is similar to immunocompetent patients. A group from Paris (Blazy et al 2005) reported on a cohort of 9 HIV+ men all on HAART treated with chemoradiation (45). They found no correlation between CD4 count and toxicity. Clinical outcome was similar to immunocompetent historical controls (45). Yet another single institutional study from St. Vincent’s (1997-2005) also reported that 32 HIV+ patients on HAART with mean CD4 count of 350 also did as well as immunocompetent patients (46). Similar to immunocompetent patients’ tumor size and treatment duration correlated with local regional control (46). A study at Case Western (1999-2007) compared treatment efficacy of immunocompetent and immunodeficient individuals (47). 14/36 patients were HIV+. The authors demonstrated similar efficacy of treatment and toxicity profile for both HIV+ and HIV negative patients. 3yr OS was 84-92%. The authors showed no correlation between CD4 count and response to treatment however the caveat being that 10/14 patients were on HAART and mean CD4 count was 190 (HIV1 RNA 16,670copies/ml) (47). Also HIV+ patients on RTOG 92-08 without treatment breaks did just as well as immunocompetent patients (48).

Another group from Germany (Fraunholz et al 2010) reported on a cohort of 21 HIV+ patients with anal cancer all on HAART (1997-2008) (49). While on HAART, all patients were able to complete the standard chemoradiation therapy for anal cancer. Only 5 cases required interruptions (median of 4 days). 81% had a complete clinical response, 5yr OS was 67%. Interestingly, this paper noted that CD4 counts dropped during treatment and one-third of patients had increases in HIV viral load (49). Both returned to baselines values at follow-up. It is unclear at this time what a transient increase in HIV viral load does to the overall disease progression in HIV+ patients. HAART appears to help HIV+ patients tolerate anal cancer treatment. However it has been observed that anal cancer treatment can cause immunosuppresion and patients need close monitoring during treatment. This immunosuppression may lead to the development of a specific pathologic subtype of anal cancer. The German study by Fraunholz et al (2010) noted that HIV+ patients on HAART had a large cell histology subtype of squamous cell carcinoma over 90% of the time compared to only 67% of HIV negative patients (49).

Not all reports state that HIV+ patients on HAART do fine with treatment. There are a couple of reports that show that HIV+ patients on HAARRT do worse than HIV negative patients. A multicenter cohort from Europe (Zurich, Paris, Geneva, Montreal, 1997-2006) reported on 40 HIV+ patients on HAART and 80 HIV negative patients (50). Overall there was >90% complete response. HIV+ patients on HAART had larger duration of treatment and more toxicities than immunocompetent patients. The 5 yr local control was only 38% for HIV+ patients on HAART compared to 87% for HIV negative patients (50). The Zurich author’s single institutional report stated that it was difficult for HIV+ patients to adhere to chemotherapy regimens and that chemotherapy had been decreased or changed in their cohort (51). This may be one explanation for the differences in outcomes. Northwestern (Hogg et al 2009) also suggested that HIV+ patients on HAART have more local recurrences and less response to therapy (52). Still there was >80% complete response and ~70% overall survival in their single institutional experience (52).

Trials using IMRT have been conducted to determine if this new technique still provides the same effective treatment outcome as conventional external beam RT while minimizing toxicities. Single institution studies seem to suggest encouraging results with IMRT. A study by Duke (Pepek et al 2010) demonstrated that out of 47 patients treated, the hematologic toxicity was 27%; there were no grade 3 skin toxicities and only 9% grade 3 GI toxicity (56). Only 18% of patients required treatment breaks. Again efficacy was in the 80% range (56). However long term follow up is lacking with a median follow up of only one year. Milano et al (2005) reported on 17 patients treated at the University of Chicago with similar results to the Duke trial (57). There were no treatment breaks from skin or GI toxicity and the authors were able to minimize toxicity to genitalia and small bowel. There was still 38% hematologic toxicity (57). These single institution studies show encouraging results with IMRT reducing acute toxicities and making treatment more tolerable.

Multi institutional trials using IMRT have also been conducted. A trial from the Mayo Clinic, University of California at San Diego, Emory University, Loyola University, and the University of Chicago was conducted to determine the efficacy of IMRT in treating anal cancer. 53 patients were analyzed, 8/53 were HIV+ (58). Treatment efficacy was similar to historical controls in that overall survival and colostomy free survival was in the 80% range (58). Median follow up time is only slightly over one year. In this study all HIV+ patients had a complete response. IMRT did help minimize GI toxicity (15% grade 3) however they still reported high rates of dermatologic (38%) and hematologic toxicity (34%). 41.5% of patients had to take treatment break (58). This trial recently reported the first volumetric study following IMRT implementation. There results suggest if the volume of small bowel receiving 30 Gy is less than 450cc then there is a 3 fold reduction in toxicity. IMRT had no effect on limiting bone marrow toxicity as 58% of patients had grade 3-4 leukopenia (59). This past year, data from a RTOG multiinstitutional phase II trial analyzing IMRT in the treatment of anal cancer were published in abstract form (60). Hong et al (2010) observed that IMRT was feasible and that IMRT decreased skin toxicity as well as high grade GI/GU toxicities more than 15% as compared to historic controls from the RTOG 9811 paper describing the standard of care for anal cancer (33,60).

Currently, the National Cancer Comprehensive Network (NCCN) guidelines state that patients should have a digital rectal exam 8-12 weeks after chemoradiation to determine response to treatment (61). If there is question of disease progression or no response, then a clinical biopsy is warranted. If biopsy is positive, the NCCN still recommends waiting an additional 4 weeks to assess response to chemoradiation. If persistent disease is biopsy positive then restaging is indicated. If disease is local (i.e. no distant metastases) then an APR with colostomy is recommended with/without groin dissection based on nodal positivity for salvage cure. Salvage surgery is effective approximately 50-60% of the time (62).