Chapter 42. Cancer of the Anus

Cancers of the anus are rare problems with diverse histology. While squamous cell carcinoma (SCC) of the anal canal remains by far the most common of these neoplasms and the main focus of this chapter, the anus may also harbor tumors such as adenocarcinoma, melanoma, and basal cell carcinoma. The treatment of anal cancer has undergone dramatic changes in the past 30 years. Multimodality treatment consisting of radiation and chemotherapy has replaced abdominoperineal resection or wide local excision as the mainstay of therapy. Five-year survival rates now exceed 80% and radical surgery is reserved for cancers of the anal canal that do not respond to chemoradiation or that subsequently recur locally. Our understanding of the etiology and epidemiology of anal SCC and its precursor lesions has also profoundly changed in the past few decades, yielding new initiatives in both therapy and prevention that may further alter the future treatment of this disease. The importance of the surgeon's role in the detection and diagnosis of anal cancer remains undiminished. The surgeon is the clinician most likely to diagnose the disease, delegate treatment, and provide follow-up care. Anal cancer is clearly a disease that benefits from multidisciplinary intervention. Because of this, the treatment of anal cancer serves as a paradigm for the multimodality treatment of cancer.

The anal canal extends from the top of the anorectal ring (a palpable convergence of the internal sphincter, deep external sphincter, and puborectalis muscle) to the anal verge (the junction of the anal canal and the hair-bearing keratinized skin of the perineum). The lining of the anal canal is comprised of columnar cells, transitional epithelium, and non–hair bearing squamous epithelium. Tumors distal or beyond the verge have been historically been termed anal margin tumors (Fig. 42-1A).

The anal transition zone, or transformation zone (ATZ) is a unique anatomic region, which has a variable histologic makeup. It is a 1- to 2-cm region which begins at the dentate line and extends proximally. This zone, similar to the transformation zone of the cervix, contains a transitional epithelium containing columnar cells with variable amounts of squamous metaplasia. These metaplastic cells may be found as high as 6–10 cm proximal to the dentate line. This may, in part, explain the existence of rare “intra-anal” SCCs that have been found in the mid-low rectum. Tumors arising in the anal canal above and within the ATZ are typically nonkeratinizing SCCs. Those originating below this level are generally keratinizing.1

Because of the complex gross and histologic anatomy of this region, classification of anal neoplasms has been confusing and inconsistent. According to the World Health Organization (WHO) classification, anal canal lesions consist of squamous cell (cloacogenic) variants, including keratinizing, nonkeratinizing, and basaloid tumors. Other anal canal neoplasms include adenocarcinoma, carcinoid, lymphoma, and melanoma.2 Anal margin tumors include SCC, giant condyloma (verrucous carcinoma), and basal cell carcinoma.2

A more current clinical classification scheme is more progressively in use which is more broadly understood by all practitioners who may treat patients with anal pathology.3 This classification divides the area into three anatomic regions: intra-anal, perianal, and skin (Figs. 42-1B and C). Intra-anal lesions are contained fully within the anal canal and cannot be seen on external view. Lesions may be made partially visible with gentle distraction of the area. Perianal lesions are fully visible, and lie within a 5-cm radius of the anal opening. Skin lesions lie outside this 5-cm radius. Henceforth in this chapter, we will use the updated terms to refer to the location of these tumors.

The dentate line provides an anatomic reference point for lymphatic drainage of the anal canal and margin. Above the dentate line, drainage is primarily via the superior rectal lymphatics to the inferior mesenteric nodes and laterally along the middle and inferior rectal vessels to the internal iliac nodal basin. Lesions distal to the dentate line drain to the inguinal and femoral lymphatics. Tumors in the ATZ may follow both lymphatic routes. Patients with unexplained inguinal lymphadenopathy should undergo a careful examination of the anal canal.

Incidence and Epidemiology

In the last two decades of the twentieth century, the incidence of anal cancer nearly doubled; in the year 2000 approximately 3400 new cases were reported in the United States. During the ensuing 9 years, this figure has risen even faster—in 2009, it is estimated that there will be 5290 new cases, reflecting a trend that mirrors increases in human immunodeficiency virus (HIV) infection.4–7 Although this number represents only 1–2% of all large bowel cancers, the rise in incidence underscores a significant and serious change in the epidemiology of the problem.5 Squamous cell cancers of the anus are thought to have a viral etiology that is similar to that of cervical cancer. There is much evidence to suggest that high-risk sexual activity in the era of the HIV is responsible for the potentiation of the viruses that cause anal SCC and that the rise in incidence is directly linked to this phenomenon.

Until the past decade, the highest rates of anal SCC were described in women with numbers increasing after 30 years of age to plateau at an incidence of 5.0/100,000 after age 85.8 The ratio of females to males affected was approximately 2:1.9 However, in the past decade men under the age of 45 who have sex with men have constituted the group with both the greatest number of reported cases as well as the greatest increase in disease incidence. Currently in the United States, anal cancer occurs more frequently in males than in females.9

Considered as a group, men who practice anoreceptive intercourse have an incidence of anal SCC of 35/100,000—a rate identical to that of cervical cancer prior to routine cervical cytological screening.9 Although not yet listed as an acquired immunodeficiency syndrome (AIDS)-defining illness like cervical cancer, an argument may be made that anal SCC should have similar emphasis. The United States AIDS-Cancer registry is a survey that linked AIDS-related cancer registries in 11 states or metropolitan areas for the period of time between 1995 and 1998 and included over 309,000 HIV-infected patients.10 The relative risk of SCC-type anogenital cancers in this population was much higher than that of the general population. The relative risks for cervical, vulvar or vaginal, and penile cancers were 5.4, 5.8, and 3.7, respectively, while the risk for anal cancer in women was 6.8 and for men 37.9.10 Subset analysis of affected individuals revealed that those less than 30 years of age, compared to the healthy, HIV-negative, human papillomavirus- (HPV) negative population, had dramatically elevated relative risks of anal cancer of 134 for women and 162.7 for men. Analyzing the data by HIV exposure history showed that homosexual contact resulted in the highest relative risk of anal SCC, with other categories such as intravenous (IV) drug abuse, heterosexual contact among women, and blood transfusion somewhat less.11

Etiology, Pathogenesis, and Risk Factors

Human Papillomavirus

HPV is a double-stranded papova DNA virus with a predilection for mucoepithelial tissues. More than 100 HPV strains have been identified, but only approximately 30 have been isolated in cancers of the anogenital region.12 The majority of those exposed clear the virus, however, chronic HPV infection results in either anogenital warts (condyloma acuminata) or squamous intraepithelial lesions (SILs).12 Condylomata are generally associated with HPV 6 and 11 and their subtypes and consist of fleshy growths that harbor and generate infectious viruses and have virtually no malignant potential.12

It is estimated that between 10 and 40% of HIV-positive males who have sex with males (MSM) will develop chronic infection with HPV strains with malignant potential. The most commonly isolated oncogenic HPV viruses are HPV 16, 18, 31, 33, and 35, which are strongly associated with invasive cancer and are commonly found in both anal and cervical cancer.13 In a case-control study of 388 patients with anal cancer from Denmark and Sweden, 88% of anal cancers harbored HPV DNA.14

HPV infection is the most common sexually transmitted viral disease. Transmission is not prevented by safer-sex practices. These oncogenic viruses may lead to premalignant changes and uncontrolled cellular proliferation, via integration in host-cell genome and loss of cell-cycle regulation.15

Human Immunodeficiency Virus Infection

There is an increased incidence of both anal SCC as well as its precursor lesion, high-grade squamous intraepithelial lesion (HSIL) in patients with HIV infection. Progression to high-grade dysplasia is accelerated in patients with a low CD4 count (<200).16 Data collected in case-control studies among homosexual men and heterosexual women with high-risk behaviors show a direct correlation between HIV seropositivity, HPV prevalence, and anal cancer and its precursors. Anal HPV infection is found in nearly 92% of HIV-positive MSM compared to 66% in HIV-negative MSM. Early epidemiologic evidence among homosexual men in the San Francisco Bay area documents a dramatic rise in anal SCC between 1973 and 1999, when the relative risk increased from 3.7 to 20.6.17 Similar studies conducted in New York city between the years 1979 and 1985 show a 10-fold increase in anal SCC in men 20–49 years of age that coincided with the explosion of HIV in this population.18

Since the advent of highly active antiretroviral therapy (HAART), morbidity and mortality in the HIV population have decreased dramatically, both due to opportunistic infections and malignancies. Thus, it might be hypothesized that the incidence of HPV-related diseases and anal SCC would decrease in this group. However the opposite appears to be true; increasing incidence of anal cancer in the HIV-positive population in the HAART era have been shown by multiple groups. A report by Piketty et al using the French Hospital database on HIV19 showed an increase in incidence, especially in the MSM population. In another recent publication, a Surveillance, Epidemiology and End Results (SEER)-based analysis showed that over the period from 1992 to 2003, anal cancer was the only malignancy increasing in incidence among HIV-positive individuals in the United States.20

However, HPV, HSIL, and anal cancer do not seem to be phenomena linked exclusively to homosexual men. Similar findings occur in HIV-infected male heterosexual IV drug users who deny anal-receptive sex. In this cohort, a high rate of HPV infection coincides with an elevated rate of HSIL as well as anal cancer.21 Heterosexual women who are HIV-positive or have progressed to AIDS have high rates of HSIL as well.10 When HIV-positive and HIV-negative cohorts (both male and female) with similar HPV risk factors are compared, the rates of both HSIL and anal cancer are dramatically increased in the HIV-positive groups.10,11,18,22

Although this trend includes both men and women with HIV, HIV-positive MSM is the highest risk group with an incidence of anal cancer higher than 78/100,000 person years.23

Smoking

Cigarette smoking is a well-known risk factor for anal SCC that is independent of sexual practices. The risk increases two- to fivefold over that of the general population.18,24 It is speculated based on data demonstrating an increased incidence in premenopausal women of 5.6 with a 6.7% linear increase per pack-year, that smoking may have some antiestrogenic effect permissive for the disease in the estrogen-sensitive tissues of the anal canal.25 This hypothesis is supported by the finding that no-risk increase was demonstrated by this study in either postmenopausal female or male smokers.

Chronic Inflammation

At one time, benign anorectal conditions such as hemorrhoids, fissures, and fistulas were thought to predispose to the development of SCC. The etiology or common mechanism was presumed to be prolonged exposure of the anal epithelium to chronic inflammatory conditions. Patients with inflammatory bowel disease were believed to be at increased risk, particularly when anal fistulas were present. In 1994, Frisch examined this issue in a large population and found no evidence to support a causal relationship between benign anorectal conditions and anal cancer up to 13 years after resolution of the benign condition.26 In another large population study, Frisch and Johansen identified 9602 Danish patients with a diagnosis of either Crohn's disease or ulcerative colitis with a mean follow-up of 10 years.27 Only two patients developed anal SCC during this time. Both patients had the disease longer than 15 years. Although long-term irritable bowel disease patients may be at slightly increased risk of anal SCC, short- and mid-term risk is not significantly different from that of the general population.27

Anal Intraepithelial Neoplasia or Squamous Intraepithelial Neoplasia

Diagnosis and Treatment

Anal intraepithelial neoplasia (AIN) is widely believed to be the precursor lesion for SCC of the anus. The terms carcinoma in situ (CIS), Bowen's disease, AIN, anal dysplasia, and squamous intraepithelial lesion (SIL) have all been used to refer to the same spectrum of pathology; nomenclature varies by pathologist. There is a growing effort to make this pathological definition more uniform, and prognostically, lesions can be lend themselves to be divided into normal, low-grade squamous epithelial lesions (LSILs), HSIL, and invasive cancer.3 In the United States, the Bethesda criteria for anal intraepithelial lesions (AIN) lists two dominant categories—HSIL and LSIL.13 In the European literature, HSIL is known as AIN 3, whereas LSIL consists of AIN 1 and 2.13

The incidence of anal cancer among HIV-positive homosexual men is 75–80/100,000 (a rate of 0.8/100,000 in the general population), more than twice the incidence of cervical cancer in women (35/100,000) prior to the introduction of routine cervical Pap smear cytology evaluations.10,17 Because of the dramatic reduction in cervical cancer (8/100,000 currently) attributed to the detection of dysplasia, it is widely believed that the same result could be seen in high-risk anal cancer populations if similar detection and ablation methods are used.

Previously, wide local excision had been the treatment of choice for HSIL (Bowen's disease). It was assumed, based on anecdotal evidence that a percentage of patients with HSIL progresses to invasive cancer. This led to attempts to surgically clear patients of the disease.

A 1999 survey of the practice patterns of members of the American Society of Colon and Rectal Surgeons revealed that 86–95% of surgeons treated HSIL with wide local excision.28 A distinction was made between microscopic disease and other manifestations. Most HSIL found incidentally in hemorrhoidectomy specimens were considered microscopic asymptomatic disease and simply followed without re-excision (74%).28 This survey coincided with other investigations highlighting the multifocal nature of HSIL and the difficulty presented by wide local excision under these circumstances. In one review of 34 patients undergoing wide local excision for macroscopically evident HSIL, 19 had positive margins at the time of initial resection, and 12 of the 19 had recurrent HSIL within 1 year.29 Even with a microscopically complete initial resection, 2 of 15 patients eventually developed HSIL. Although none of these individuals subsequently developed anal cancer, five developed significant surgical complications of resection including anal stenosis and incontinence.29

More structured techniques using mapping biopsies, intraoperative frozen section and selective wide-local excision yielded excellent results and long-term control, however recurrent disease still occurred.30

The true incidence of HSIL and its resultant progression to invasive SCC are not clearly known, however the rapid increase in incidence of SIL and anal SCC in clearly high-risk populations, combined with the potential morbidity of a radical surgical approach have led many to adopt a policy of either very specific ablative therapy or close and frequent observation.

Anal screening (Pap smear) was first described in the 1990s as a direct corollary of the cervical Pap smear, and has since been promoted as a diagnostic and screening tool in high-risk populations.31 However, evidence demonstrating a resulting decrease in the incidence of anal cancer similar to that of cervical cancer has not been forthcoming. Still, only a very short time has passed since the institution of the technique. The use of anal cytology as a screening technique has not gained the recognition afforded cervical Pap smears. Lack of recognition by clinicians of the increased incidence of anal cancer, limitation of the problem to high-risk populations, lack of knowledge of techniques, cost, and a dearth of supporting outcomes data may all conspire to limit the use of the technique. Ongoing outcome studies may clarify the role of anal Pap smear for high-risk patients.

In the early 1990s high-resolution anoscopy (HRA) was developed at the University of California, San Francisco (UCSF). Like anal Pap cytologies, HRA is a direct application of the technology for cervical intraepithelial detection and ablation to anal dysplasia. The technique can be done in either the office, or for more extensive disease, in the operating room.

After obtaining a Pap smear, a digital rectal examination is performed followed by placement of a cotton swab covered in gauze soaked in 3% acetic acid. The swab is held in place for 1 minute after which an anoscope is inserted, permitting examination of the anal canal by a colposcope providing 6- to 25-times magnification. Special attention is directed to the area surrounding the ATZ. Applying acetic acid causes these often unapparent lesions to become opaque or “acetowhite.” Lugol's iodine solution is then placed in the anal canal to further highlight these areas. HSILs fail to take up Lugol's, rendering the area yellow to tan, whereas normal tissue or LSILs stain dark brown or black.31 This approach, combined with the magnification, allows visualization of vascular changes such as punctate appearance, mosaicism, and atypical vessels characteristic of dysplastic change.31 Suspicious lesions are then destroyed by electrocautery.

Pineda et al published a retrospective review of 246 patients being treated for HSIL with HRA and targeted ablation over a 10-year period. In this group, 81% of patients had extensive, or circumferential lesions, and 79% were immunocomprised due to HIV. Recurrent HSIL occurred after ablation in 57% at an average of 19 months of follow-up, but only 25% of those patients required surgery. In this high-risk group, under careful surveillance, 1.2% of patients progressed to invasive cancer.32 This series represent the largest report of patients followed with HRA and targeted surgical ablation. It is especially notable for a much lower rate of progression to invasive cancers compared to other studies detailing expectant management without the use of HRA surveillance, where progression to invasive cancer ranges from 8.5 to 13%.33,34

HRA may provide objective evidence of the presence of disease that office examination alone does not. Whether ablative therapies should follow documentation of HSIL by any method remains unknown and controversial. There have been no randomized controlled trials to date which can clearly characterize the appropriate approach for this group. Slow progression from HSIL to invasive cancer, heterogeneity in follow-up, and lack of broad expertise in surveillance techniques make these studies impractical to perform. Nevertheless, it is clear that special attention needs to be paid to these highest-risk groups.

Human Papillomavirus Vaccines

Both prophylactic and therapeutic vaccines to HPV have completed Phase III trials targeting cervical cancer, testing both types of vaccine in high-risk populations.13,35. Prophylactic or preventive vaccines are typically made from structural viral proteins, while therapeutic vaccines are made from the early viral replication proteins E6 and E7. In 2006, the Food and Drug Administration (FDA) granted approval to the first vaccine designed to prevent cervical cancer. Gardisil (Merck & Co., Inc, Whitehouse Station, NJ) is a recombinant vaccine against HPV types 6, 11, 16, and 18. It is currently approved for use in females, 9–26 years of age and requires a series of three injections over a 6-month period. A total of 21,000 patients in four randomized trials demonstrated a dramatic, nearly 100% prevention rate in genital warts, and vulvar, vaginal, and cervical precancerous lesions caused by the serotypes against which the vaccine is directed. The vaccine is only effective, however, in patients not previously exposed to the viruses included in the vaccine, and it confers no protection against viruses not covered by the vaccine. Scattered reports of adverse side effects to Gardasil have been reported, including syncope, headache, and anaphylaxis raising some concern regarding its use. However, the significance of these events is yet to be determined.

A recent double-blind placebo-controlled randomized clinical trial (RCT) involving over 4000 HPV genotype-naïve males showed that the quadrivalent vaccine was 86% effective in preventing persistent HPV infection, and 90% effective against genotype-specific condyloma.35 Although the potential usefulness of this vaccine can be extrapolated, further studies in high-risk population will be needed to determine efficacy and effects on SIL and cancer prevention.

Cervarix (GSK) is a bivalent vaccine targeted against HPV 16 and 18. It has been licensed for use in females as young as 10 years of age. No registration for treatment indications in males has been sought.35

Other studies have combined aerosolized delivery mechanisms with intramuscular injections to maximize antibody titers against the virus. Although the practicality of prevention of dysplasia and cancer by vaccines is unclear at this time, results from these studies may clarify the situation in the near future.

Stressgen Biotechnologies, Inc. has developed a therapeutic vaccine for anal HSIL that has completed Phase II clinical trials in HIV-negative patients.13 The vaccine is a recombinant fusion protein called HspE7. The immune response generated by the vaccine seems to be CD8 dependent alone—CD4 cells do not seem to be involved. Of those patients receiving 500 μg doses, 76% showed regression of their HSIL to LSIL, one of the primary endpoints of the study. Approximately 7 months elapsed before the first complete responses began to appear. Even so, results seem durable with 86% of this group remaining in remission at 15 months. Although the study samples are small and this approach has yet to be validated in HIV-positive immunocompromised patients, early evidence supports an optimistic outlook for the field of HPV therapeutic vaccines.13

Pathology, Diagnosis, and Staging of Anal Squamous Cell Cancer

Pathology

Nearly 80% of anal canal tumors are either SCCs or histologic variants of SCC. The great variation in terminology results from the histologically diverse microscopic anatomy and the fact that many tumors, especially in the anal transition zone, have a mixed histologic appearance, including squamous, basaloid, and rarely glandular elements. The WHO designates all squamous carcinoma variants in this location as “cloacogenic.”2 Tumors of the distal anal canal (anal), and particularly of the anal margin (perianal), are generally comprised predominantly of squamous cells, with fewer basaloid and no glandular characteristics.36 The more distal in the anal canal the squamous tumor arises, generally the more likely it is to contain keratinizing cells. Tumors of the proximal anal canal and ATZ are usually composed of nonkeratinizing cells.36 It is important to note that the difference in the cellular characteristics of these anal canal cancers does not result in a different mode of treatment. There are no data to suggest differences in outcome between squamous and basaloid histologic types in anal canal cancers. Perianal (anal margin) tumors, however, are typically treated like skin cancers, by local excision.

The treatment of anal cancer has undergone major changes within the past 30 years. Currently chemotherapy and radiation are usually the sole treatment for patients with this disease. Prior to 1974, the standard of care was either wide local excision if the tumor was judged to be superficial, or abdominoperineal resection (APR) for tumors invading the sphincter. Outcomes were poor, with overall survival rates after APR ranging from 30–70%, depending on tumor grade, stage, and size.8 The local recurrence rate after wide resection or APR was reported to be 25–35% with a 100% local recurrence rate for tumors invading through the submucosa in a series from Singh and associates at Roswell Park Memorial Institute.37 Perineal or pelvic recurrence occurs in 50–70% of patients undergoing APR, with less than 10% dying of distant disseminated disease.9 In 1974, Norman Nigro at Wayne State University used radiation and fluoropyrimidines in anal canal cancer as a way to reduce local recurrence following APR.38 He observed that often there was no residual cancer in the resected specimen. Thus began an exciting and revolutionary time in the treatment of this disease that resulted in a radical shift in treatment.

Diagnosis and Staging

Over 50% of patients present with a complaint of rectal bleeding. Delays in diagnosis are common because the tumor is often mistaken by both patients and physicians for benign conditions such as hemorrhoids or fissures (Fig. 42-2). Pain, tenesmus, and pruritus may be present. The initial physical examination should include a digital rectal examination, proctoscopy, and inspection of the inguinal lymph nodes. A biopsy of the anal mass is necessary to confirm the diagnosis. Inguinal masses should be aspirated with a fine needle for diagnosis and staging. Because the current nonoperative approach to anal cancer management is highly effective, excisional biopsy of suspected anal SCCs and inguinal node dissection for adenopathy should generally be avoided. The staging process includes CT of the chest, abdomen, and pelvis, and a transanal ultrasound to assess depth of invasion and aid in establishing the size of the tumor (Figs. 42-3A and B). In addition, the use of positron emission tomography (PET)/CT is becoming more standard is the pretreatment staging. This allows for anatomic and metabolic correlation of primary tumor as well as sensitive assessment for inguinal or distant nodal metastases. Importantly, post-therapy response to PET is predictive of long-term outcomes.39

The IUCC staging system for anal cancer was updated in 1997 and adopted by the AJCC7 (Table 42-1). In contrast to staging parameters for other gastrointestinal (GI) lesions, it is based on size rather than depth of invasion. Anal margin tumors are staged and treated the same as skin cancers (Fig. 42-4).

A number of reviews in the literature prior to and during the introduction of chemoradiotherapy for anal SCC document the strong correlation between tumor size, lymphatic spread, and prognosis.40 In a 1984 report from the M.D. Anderson Cancer Center, 132 patients treated by APR for anal SCC were studied. For patients with tumors 1–2 cm in size, survival was 78%; 3- to 5-cm tumors had survival of 55%; and patients with tumors greater than 6 cm experienced survival of only 40%.40 Other reviews suggest that survival for large tumors is considerably worse, at less than 20%, and that generally overall survival is diminished when tumor size is greater than 5 cm, whether the tumor is treated by excision or chemoradiotherapy.41–44 Recently, analysis of the completed Radiation Treatment Oncology Group (RTOG) 98-11, which represents the largest prospective trial database showed that pretreatment tumor size greater than 5 cm predicted colostomy requirement.45

The presence of regional nodal metastases is a poor prognostic indicator regardless of treatment modality. Although survival in the face of nodal metastases has improved significantly with the use of chemoradiation, patients who present with metastatic disease have a significant survival disadvantage.40,41 Prior to the routine use of chemoradiotherapy, a report in which surgery was done with and without preoperative radiation demonstrated a 5-year survival rate of 44% for node-positive patients compared to 74% for node-negative patients.40 Other studies confirm comparatively poor survival for patients with nodal metastases.41

Surgical Management

Operative therapy for anal SCC has largely been supplanted by chemoradiation and is now the exception rather than the rule. Historically, the failure rate for APR has depended rather predictably on the size of the primary tumor. This procedure was often accompanied by prophylactic inguinal node dissection, but the morbidity and lack of efficacy caused inguinal lymphadenectomy to be abandoned. Failure rates for APR range from 40 to 70%, with local failure rates of 40% and median survival time after recurrence of only 1 year.8

Although chemotherapy and radiation have been shown to result in higher disease-free survival rates, when chemoradiotherapy is refused or contraindicated, there may still be a role for local excision in some cases of anal canal carcinoma. A retrospective analysis of local excision at the University of Minnesota revealed a direct correlation between survival and tumor size. For tumors greater than 2.5 cm, 5-year survival rates were 60%.46 Although the sample size was small, the authors advocated local resection with curative intent for small (<1 cm) well-differentiated tumors confined to the submucosa.46 Corman and Haggitt reported a similar experience, with all tumors confined to the submucosa being cured by local excision or APR, and those invading more deeply suffering eventual local recurrence.47 Longo and colleagues recorded a 62% failure rate in stage I–III tumors undergoing solely local excision, in which all patients with stage II and III tumors recurred.48 Tumor accessibility, full-thickness excision, depth of invasion, and negative margins seem imperative technical considerations when considering local resection. Even so, very few candidates are suitable for this approach.

Chemoradiation

The treatment of anal (canal) carcinoma has changed radically since the late 1970s, with the advent of chemoradiation protocols. In 1974 Norman Nigro defined a treatment protocol involving the administration of 5-fluorouracil (5-FU), mitomycin-C, and preoperative radiation to shrink anal canal tumors.38 Fluoropyrimidines were known at the time to enhance the effect of radiation, and there was some evidence that mitomycin had an antineoplastic effect on squamous cell tumors. Nigro's protocol was neoadjuvant, and the radiation (30 Gy total) was given in 15 sessions over a 3-week period. The 5-FU was administered at a dose of 1000 mg/m2/d, for 4 days starting on the first day of radiation therapy, as a continuous infusion. It was then repeated on days 29 through 32. Mitomycin-C (15 mg/m2) was administered as a single dose on treatment day one.38 Of the three patients in the initial report, two underwent APR 6 weeks after treatment. The third refused surgery and remained disease-free. No evidence of tumor was found in the specimens of the two patients who underwent surgery.38

Following the dramatic results reported by Nigro's group, others followed suit, treating patients with both radiation alone and with multimodality therapy followed by surgical excision. In 1983, Michaelson and associates at Memorial Sloan-Kettering Cancer Center (MSKCC) reported that 52% of patients treated with both chemotherapy and radiation had a complete pathological response, and another 22% had only microscopic disease at operation.49 All of these patients had undergone APR or wide local excision following treatment. After Nigro's 1974 publication, a number of other investigators examined the effects of multimodality therapy. Most used 5-FU and mitomycin-C as the chemotherapeutic regimen, although several made dose and infusion modifications, and nearly all increased the radiation dose. Maximal doses were in the range of 50 Gy. Because of such variability among therapies, meta-analysis is difficult. However, direct comparison between studies is useful.38

Preliminary studies done by Nigro and others set the stage for prospective Phase II studies. Among these, Martenson and colleagues reported on an Eastern Cooperative Oncology Group (ECOG) study of 50 patients receiving 40 Gy of radiation with a 10–13 Gy boost to the tumor.50 Bolus 5-FU and mitomycin-C was given during radiation, and biopsy of the tumor or tumor site was performed 6–8 weeks later. APR was performed if the biopsy was positive. Of 46 patients completing treatment, 34 (74%) had a complete response and 11 had a partial response.50 Eighty percent had no locoregional recurrence and 58% were disease-free at 7 years.50

The RTOG and ECOG reported on an intergroup trial of 79 patients treated with combined radiation and chemotherapy in 1989. The radiation dose was 40.8 Gy and only 8 patients had evidence of disease requiring APR at the completion of therapy. At 3 years, overall survival and local control rates were 73% and 71%, respectively.51

Further series from MSKCC supported the ECOG and intergroup study. Forty-two patients were treated with a total dose of 30 Gy and the 5-FU/mitomycin-C combination.52 Eighteen patients had positive biopsy results after treatment but only half of these had local recurrence on follow-up. The 5 year disease-free survival rate was 82%.52

In all of these small Phase II trials, disease-free survival, colostomy-free survival, and local disease control compared very favorably to the standard surgical approach. However, the toxicities encountered were significant. In the ECOG study, 37% of patients suffered severe toxicities including severe neutropenia, moist desquamation of the perianal skin, and diarrhea.50 Treatment toxicities like these gave rise to questions regarding the necessity of chemotherapy in anal cancer in spite of its early promise. Concurrent studies examined the role of radiation alone, often in doses substantially higher than those used with chemotherapy. At the Institute Curie, 183 patients receiving a dose between 60 and 65 Gy showed a 59% 5-year survival rate with a local control rate of 69%.53 A similar 5-year survival rate of 61% was demonstrated in a review of 147 patients from the Hospital Tenon.43 Local control in this study was 71%. Complications of higher-dose radiation included anal ulceration and stenosis requiring surgery in 5–15% of cases.

In the late 1990s three Phase III trials reported direct comparisons between radiation alone and radiation with concurrent chemotherapy (Table 42-2). In 1996, the United Kingdom Coordinating Committee on Cancer Research (UKCCCR) published the largest prospective randomized study of chemotherapy and radiation versus radiation alone.54 The trial enrolled 585 patients, assigning them to either combined therapy or radiation, and then assessed them at 6 weeks. Poor responders were offered APR while those responding well received boost radiotherapy and reassessment. Those patients receiving only radiation had a local failure rate of 59%, whereas those with multimodality therapy recorded a 36% local failure rate with a mean follow-up time of 42 months. Although the early morbidity of combination therapy was higher than that with radiation alone (including two deaths from sepsis), the late morbidity rate was the same. Both the local failure rate as well as the number of patients requiring salvage surgery was halved compared to radiation alone. In all, 29/174 patients who had received combined therapy with a boost required salvage APR, compared to 72/188 who had received radiation alone. Although the local failure rate for radiation alone was higher, overall survival between the groups was not statistically significant (58% radiation vs 65% chemoradiation at 3 years).54

The results of the European Organization for Research and Treatment of Cancer (EORTC) supported those of the UKCCCR trial.55 Patients with locally advanced (T3–T4) cancers were randomized to radiation alone (45 Gy plus a boost of 15–20 Gy) versus combination therapy with 5-FU/mitomycin-C. With the addition of chemotherapy, the local failure rate dropped from 69% to 42% and colostomy-free survival increased from 22% to 41%. Early and late complication rates were similar except for anal ulcers, which were slightly increased in the combined group. As in the UKCCCR trial, although local control and colostomy-free survival rates were much improved over that of radiation alone, the rate of distant spread was unchanged. Overall survival between the two groups in this study was 56% at 5 years.55

In 1996, Flam and colleagues explored further the role of mitomycin-C as a radiation sensitizing agent in a phase III RTOG/ECOG trial.56 Three hundred ten patients were enrolled and randomized to receive either radiation/5-FU or radiation/5-FU and mitomycin-C. They concluded that although the addition of mitomycin-C produced slightly greater toxicity, at 4 years the disease-free survival was higher (73% vs 51%; p = .0003) and the colostomy rate was lower (9% vs 22%; p = .002). While the 5-FU/mitomycin-C/radiation group had a good overall survival of 76%, this was not statistically different from that of the comparison group at 67% (p = .18). However, the role of mitomycin-C was validated.56

The RTOG/ECOG study also examined the ability to salvage patients with residual cancer in their post-treatment biopsy with additional chemotherapy and radiation. Of the 24 patients on the trial eligible to undergo salvage, 12 were rendered free of disease with a 9-Gy boost, 5-FU, and cisplatin (100 mg/m2).56 It has been suggested that the patients who underwent salvage chemoradiotherapy in this trial may have been free of disease secondary to radiation-induced apoptosis if the period prior to biopsy had been extended. It is unclear whether cisplatin was actually responsible for the results, but interest in the agent was sparked, given the treatment-limiting toxicities of mitomycin-C. Cisplatin is well known as a radiation sensitizer and effective agent in the treatment of SCC in other areas such as cervix, head and neck, and esophagus. There have been two Phase II trials of high-dose radiation and 5-FU in combination with cisplatin for anal canal cancer. These studies showed complete response rates of 70–95% with reduced toxicity compared to mitomycin-C.50,57

RTOG 98-11 was a Phase III study designed to directly compare the efficacy of cisplatin-based therapy to the standard mitomycin-based regimen, in the setting of large SCC primaries (T2 or larger), a group known to have decreased response to chemoradiotherapy. The authors hypothesized that induction chemotherapy with fluorouracil and cisplatin would shrink the primary tumor, and render subsequent concurrent chemoradiotherapy more effective. Primary outcome measure was 5-year disease-free survival, while secondary end points were overall survival, time to relapse, and colostomy rate. After a median follow-up of 2.51 years, as compared to standard (mitomycin) therapy, the cisplatin group showed decreased 5-year disease-free survival (54 vs 60%), worse 5-year overall-survival (70 vs 75%), higher 5-year local recurrence and distant metastasis rates (33 and 19% vs 25 and 15%), respectively. Colostomy rates were significantly better in the mitomycin group (10 vs 19%, p =.02).58

Based on these data, the authors conclude that cisplatin-based therapy failed to achieve improved disease-free survival compared to a standard mitomycin-based regimen, and in fact, worsened colostomy rates.58 Currently, at the University of Pennsylvania, a mitomycin-based regimen, without induction, is the standard treatment.

Treatment for anal cancer does not differ in the HIV-positive population. Combined chemotherapy and radiation is the best approach to this disease in the setting of HIV/AIDS. Studies have consistently documented responses to standard therapy that equal those in the HIV-negative population.10,17

Thus, for any stage of invasive SCC of the anal canal, the primary mode of treatment should be chemoradiotherapy with a 5-FU/mitomycin-based regimen and external beam radiotherapy. Surgery is reserved for in situ (wide local excision), residual or recurrent disease (APR).

Newer Modalities for Radiotherapy

The above series of randomized trials firmly established concurrent chemoradiotherapy with 5-FU and mitomycin-C as standard of care for anal cancer. This approach has resulted in very effective disease control (5-year overall survival 50–61% and 5-year colostomy-free survival from 76 to 78%).54–56,58 The tradeoff, however is that this sphincter-sparing approach has significant toxicity. The EORTC trial and UKCCCR trial reported significant acute dermatologic toxicity on 49–76% of patients, and acute GI toxicity in 33–45%.54,55

The technique of intensity-modulated radiation therapy (IMRT) is a new way to plan and deliver conformal radiation. Using advanced imaging and computer-guided techniques, radiation dosage can be delivered with higher accuracy to target tissue while sparing normal nearby tissues. A multicenter US group conducted a prospective cohort study using concurrent 5-FU/mitomycin C and IMRT techniques on 53 patients with SCC of the anus.59 Outcomes showed a favorable toxicity profile when compared with classic controls, as exemplified by the outcomes of the RTOG 98-11 (Table 42-3).58 There was decreased dermatologic toxicity(grade 3, 38 vs 78%, and no grade 4), and fewer patients needed breaks in treatment; in addition 57% of patients requiring breaks left for less than or equal to 4 days of treatment. GI toxicity was similarly decreased with only 15.1% experiencing grade 3 toxicity compared to 34% in the RTOG trial.58

Cancer response rates were similar in this study compared to controls; there was a 92.5% complete response (CR), with failures associated with advanced stage, although not statistically so. Survival rates were also comparable, with 18-month colostomy-free survival, and overall survival at 83.7 and 93.4%, respectively.

Further randomized trials need to be conducted to validate standard concurrent chemoradiotherapy with IMRT techniques; however at the University of Pennsylvania we use conformal radiation techniques as standard therapy.

Treatment of Inguinal Nodal Metastases

Palpable inguinal lymph nodes (LN) should be biopsied or evaluated by fine-needle aspiration (FNA) at the onset of treatment for staging. Several reviews have confirmed the poor prognostic outlook conferred by inguinal LN metastases. In 1970, Stearns and Quan reviewed the MSKCC experience with anal canal cancer and noted that only 14% of patients with synchronous nodal metastases survived for 5 years.60 Similarly, O'Brien and colleagues reported in 1982 that none of the 52% of patients presenting with synchronous LN involvement survived more than 3 years after diagnosis.61 Both Stearns and O'Brien observed independently that patients presenting with metachronous LN metastases had better survival following therapeutic inguinal lymph node dissection. In the MSKCC review, 75% of patients survived longer than 5 years after groin dissection.

The use of radiation on the inguinal lymph nodes, both prophylactically and for treatment, was explored by Papillon.66 In 1974, he reported on 19 patients with synchronous inguinal nodal involvement who underwent groin irradiation for disease control. Eleven of the 19 had no evidence of disease at 3 years. Cummings and associates treated nodal disease in a similar fashion and showed that 87% of patients had good disease control or cure without groin dissection.67

With the use of radiation fields expanded to include inguinal, internal, and external iliac nodes, the current treatment paradigm is to treat inguinal nodal metastases with chemotherapy and radiation concurrently with the primary tumor. Metachronous lymph node involvement is treated with salvage chemotherapy and radiation if dose limits have not been exceeded, as well as groin dissection if warranted.

Recurrent Disease and Salvage Therapy

The goal of early detection of local post-treatment recurrence is to prevent lymphatic spread of disease and maximize salvage. Most clinicians advocate a thorough physical examination including a digital rectal examination and anoscopy every 3–4 months for at least 2 years. An additional strategy involves the use of endoanal ultrasound (EAUS) inspection. At the Hospital of the University of Pennsylvania, the current protocol is EAUS every 4 months for 3 years, followed by every 6 months for 2 years. Suspicious tissue or lymph nodes are biopsied with the aid of ultrasound guidance.

There is some evidence that local regression of disease following radiation therapy can occur up to 6–9 months following chemoradiation. Routine biopsy of the anal canal following treatment is no longer recommended within this time period. Rousseau and associates advise allowing the anal canal to heal completely, reserving biopsy for nonhealing ulcers and recurrent or enlarging anal canal masses after a period of at least 6 months following therapy.68 After this point, any disease detected is residual and salvage therapy is warranted.

In spite of success with nonoperative anal (canal) cancer management, depending on the stage of disease, 10–30% of patients will recur, most locally. The treatment of recurrent or persistent disease is APR with negative margins. In a retrospective analysis of salvage therapy for recurrent disease following chemotherapy with radiation, Allal and colleagues found that APR results in a 53% actuarial 5-year survival rate versus 28% in those who did not receive additional treatment.69 Pocard and colleagues' data from St. Antoine University Hospital examined salvage APR in 21 patients who had either residual disease after sphincter conservation or recurrence. The group found an actuarial 5-year survival benefit of 30%.70 Factors resulting in failure were lymphadenopathy, positive margins, and distant disease. Longo and associates compared salvage with chemoradiation versus APR and found that only 27% of patients treated with additional combined therapy survived long term, whereas 57% of those in the APR group did71 (Table 42-4).44,72,73 Similarly, a recent retrospective review from the University of Toronto looked at a cohort of 40 patients who underwent surgical salvage after failure of chemoradiotherapy and found an overall survival of 41 months, with 5-year overall- and disease-free survival 39 and 30%, respectively.74

Patients with recurrence die of locoregional complications including ureteral obstruction, perineal sepsis and necrosis, bowel obstruction, and venous thrombosis. Contraindications for salvage surgery include medical debilitation, known distant metastases, invasion of the pelvic sidewalls, and obvious inguinal lymphadenopathy. The preoperative assessment should include a chest x-ray and an MRI or CT scan of the abdomen and pelvis. A multidisciplinary approach is appropriate for local invasion of resectable structures such as the urinary bladder, cervix, vagina, or the sacrum. A team including urologists, neurosurgeons, orthopedic surgeons, and plastic surgeons may be required. Recurrences close to the pelvic sidewall may be indistinguishable intraoperatively from fibrosis and scarring from prior radiation or surgery. An intraoperative frozen section may be useful if one is considering placing afterloading catheters or delivering intraoperative brachytherapy to these areas. The role and long-term outcomes of brachytherapy as a treatment adjunct for salvage surgery has not yet been validated.

The complications of salvage pelvic surgery may be severe and debilitating and include perineal wound dehiscence and necrosis. Tissue coverage in previously irradiated fields improves wound healing and many consider it essential for postexenteration reconstruction. Pedicle and rotational flaps may be fashioned from the gluteus, gracilis, or rectus abdominis muscles.

There is little published regarding long-term follow-up in patients salvaged with radiation or chemoradiation following local excision of anal cancer. Patients who undergo primary excision for anal canal carcinoma do so for a number of reasons, usually inadvertently, including polypectomy, hemorrhoidectomy, or excisional biopsy, as well as local excision with intent to cure. Although it is unclear at this point whether further treatment for completely excised, early-stage lesions is appropriate, patients with positive margins, or those with tumors harboring vascular or lymphatic invasion with poorly differentiated characteristics are candidates for further therapy. A retrospective analysis from MSKCC in 1999 reviewed 14 patients who received postoperative chemoradiation (either 30 or 45–50 Gy) after local excision.75 Actuarial 5-year local control rates were 93% with no difference between outcomes in the higher- and lower-dose groups. Longo and associates published the largest single retrospective analysis of outcomes in 1994, reviewing chemoradiation following local excision.48 The overall local control rate at 5 years was 79% in 109 patients receiving a median dose of 42 Gy. Stratification of the data by stage revealed a 90% local control rate with stage I, 54% with stage II, and 100% with stage III (6/6 patients).48 There have been no prospective studies comparing local excision alone versus chemoradiation for T1 favorable-histology tumors.

More recently, a multicenter group from France looked at their experience with adjuvant treatment of very early anal (canal) tumors. Of 62 patients with either Tis or T1 anal SCC, 26 had undergone primary excision followed by adjuvant radiation as compared to 43 patients treated with definitive radiation alone. Local recurrence rates were higher for the former group (3/23 vs 3/43). Local control was obtained in all six local failures via APR. Long-term survival was no different.76 There are no data to clearly show superiority of local excision of anal SCC over definitive chemoradiation. However, current studies suggest that tumors that are incompletely excised, those with poor histologic characteristics, and those that are stage II and above are candidates for chemoradiation following excision.68,71,75 As with primary therapy, giving chemotherapy (principally infusional 5-FU with mitomycin-C or cisplatin) seems to promote effective local control at lower radiation doses.

Anal (canal) carcinoma metastasizes in 10–20% of patients late in the course of disease and prognosis is exceedingly poor.9 Liver and lung metastases predominate and cisplatin-based chemotherapy is the only strategy shown to be somewhat effective.68

Squamous cell carcinoma of the perianal area is at least five times less common than anal (canal) carcinoma, and for the most part, is treated by primary surgical excision similarly to skin cancers. These tumors arise on the perianal skin beyond the anal verge. They are usually well- or moderately differentiated keratinized SCCs and generally have a favorable prognosis.9 Metastases are late and rare, and recurrences are typically locoregional. Symptoms include pain, bleeding, itching, and palpable mass. In a study from Denmark, Jensen and associates noted a 6-month median duration of symptoms prior to diagnosis, with an erroneous initial diagnosis made in 29% of cases.6 Because these tumors are fairly slow growing and uncommon, they are frequently mistaken for hemorrhoids or other benign conditions at initial presentation.

Diagnosis is often suspected by the experienced clinician on inspection, but biopsy prior to definitive treatment is imperative. If the lesion is small, excisional biopsy can be accomplished with adequate margins (1 cm). If the tumor is larger, a small incisional biopsy allows accurate classification of the tumor and appropriate preoperative counseling.

Metastases to the inguinal lymph nodes occur in 15–25% of patients. The rate of nodal metastases is directly proportional to the size of the tumor. Papillon and Chassard reported that for tumors less than 2 cm in size, the rate of nodal metastasis was 0%, for those 2–5 cm 24%, and for those greater than 5 cm 67%.77 Cummings and colleagues found that those with tumors less than 5 cm in size had 0% nodal metastases, whereas metastases occurred in 25% of those with tumors 5 cm or larger.67

Surgery

Although surgical excision (either local excision or APR, depending on location) is considered the standard of care for perianal (anal margin) tumors, outcome data for this rare neoplasm are primarily retrospective. In most studies, overall and disease-specific survival is considered for all stages together and subgroup analysis for large numbers of patients is not available. Unfortunately, evaluation of local recurrence data is similarly limited by the small numbers of patients affected; however, in general a trend toward increased recurrence in larger tumors is apparent.78

Surgical treatment of the primary perianal tumor is accomplished by wide local excision with 1-cm margins. At MSKCC, Greenall and associates reported a series of 51 patients with perianal squamous cell carcinoma (anal margin).42 Five-year survival was 88%, although local recurrence was 46%. Local recurrences were amenable to re-excision. Inguinal nodal dissection was employed for metachronous inguinal nodal metastases. Thirteen patients in this series underwent APR as initial treatment. The local recurrence rates for these patients were identical to those of the local excision group. Tumor size was the most important factor for local control and survival (Fig. 42-5). In 1979, Cleveland clinic reviewed their experience with surgery for anal margin tumors over a 20-year period.79 Eight patients were identified for whom follow-up was available. A disease-specific survival rate of 70% was noted after 8 years, with a local recurrence rate of 30%. At the University of Chicago, a 19% local recurrence rate was noted in 16 patients undergoing surgical therapy alone.80 Two of 11 patients recurred following local excision, and 1 of 3 recurred after APR. Of 27 patients with either stage I perianal (anal margin) cancer or carcinoma in situ treated at the Mayo Clinic between 1950 and 1970, 5-year survival rates were 100%, although local recurrence rates were unavailable.81

After surgery alone (local excision or APR), the overall survival rate for all stages is 60–90% with a local recurrence rate of approximately 30%. Survival rates after surgery for recurrence are unknown.9,78

Radiotherapy

The optimal treatment of perianal (anal margin) tumors is dependent on location. Significant challenges and functional problems may result when the anal sphincters are present within the boundaries of optimal surgery. If adequate excision compromises the sphincters, APR is an option. However, many surgeons and oncologists would advocate a more conservative approach and use radiotherapy. Cummings and associates demonstrated local control rates of 100% for anal margin tumors less than 5 cm in size with a dose of 50 Gy over 4 weeks.67 Local control rates were inversely proportional to the size of the tumor. For those tumors 5–10 cm, 70% local control was achieved, but for tumors greater than 10 cm, only 40% sustained a durable response. Similar results were reported by Papillon and Chassard at Centre Leon Berard in France.77 In this review, a 78% overall survival rate was achieved using external beam radiation (40 Gy cobalt 60 source) with a perineal field. Again, those with tumors greater than 5 cm in size fared considerably worse, with overall survival rates less than 50%.77

There have been numerous retrospective reviews of the response of perianal (anal margin) tumors to radiation in the past 40 years documenting stage-specific local recurrence rates, disease-specific survival rates, and overall survival rates. Overall, local control rates of 52–87% are typical, with 5-year overall survival rates ranging from 52 to 90%.78 T1 and T2 tumors have better local control rates with overall and disease-specific survival rates ranging from 82 to 100%.78

It is difficult to evaluate the sphincter preservation rate from these reviews. Small numbers and retrospective design limits direct comparison of this technique to surgery alone. There are no prospective studies comparing surgery alone to radiotherapy. Although the addition of chemotherapy (5-FU and mitomycin-C or cisplatin) seems logical, there are few data to support that approach. The rationale for these agents is extrapolated from the prospective trials of chemoradiation in the setting of anal (canal) carcinoma. Even so, it is reasonable to believe that primary radiotherapy with or without chemotherapy for perianal (anal margin) tumors in close proximity to the anal sphincters, where adequate excision may compromise function, will result in both sphincter preservation and good local control. It is also reasonable to expect that surgical salvage for recurrence after primary radiotherapy is a possibility, with rates of local control of approximately 50%. Long-term disease-specific survival following this scenario is unknown.

Melanoma of the anus and rectum is a rare malignancy accounting for less than 1% of all colorectal and anal neoplasms.82 The majority of GI melanomas are metastases; autopsy studies have shown GI metastases in up to 30% of patients with extraintestinal primaries.83 Primary GI melanoma predominantly affects the anorectum and esophagus. Although there is a female predominance, with an almost 1:2 ratio, there is evidence that the median age of affected males is significantly less (57 vs 71 years).79 Cagir and colleagues examined the epidemiology and demographics of anorectal melanoma using the Surveillance, Epidemiology and End Results (SEER) database. These investigators note a recent emergence of a bimodal age distribution of anorectal melanoma for all patients, with males occupying the younger aspect of the curve. Survival rates were slightly better in this group (63 vs 51% at 1 year and 41 vs 27% at 2 years; p < .01).84 The most common symptoms include bleeding, itching, the presence of a mass, pain, tenesmus, or changes in bowel habits. Like anal SCC, misidentification of the tumor as a hemorrhoid is a common mistake. Diagnosis is frequently made following hemorrhoidectomy or local excision of the perianal mass. The tumor can appear small and polypoid, or large and ulcerating. About 30% of these tumors are amelanotic and unpigmented making immediate recognition of the problem difficult.82 On pathology, 70% of lesions show some evidence of melanin production either grossly or microscopically.82 Commonly, anal melanoma arises at the mucocutaneous junction. Occasionally, the lesion arises more proximally, within the rectal mucosa. Although the origin of these tumors is speculative, they are believed to arise in areas of heterotopic anal canal epithelium in the rectum, or to start from proximal microscopic mucosal spread from a small lesion located more distally.82

Staging and Prognosis

Like melanoma of the skin, anorectal melanoma is staged by depth or thickness of the lesion. Lymphatic metastases can occur in the inguinal, mesorectal, and internal iliac nodal distribution. Mesorectal lymph node metastases are found in 40–60% of patients at initial presentation and inguinal adenopathy is present in at least 20% of patients.85,86 Distant spread occurs to the bone, lung, and liver.

Regardless of stage, 5-year survival rates for patients diagnosed with anorectal melanoma are very poor, averaging about 6%. The median survival time following diagnosis is 12–18 months.82

Surgery

In recent years, local excision has replaced APR for the treatment of anal melanoma. A 64-year review, concluding in 1995, of the anorectal melanoma experience at MSKCC demonstrated a trend toward increased survival for patients with localized, node-negative disease, and for those that were resectable, the authors advocated radical surgery.86 A more recent report of the last 20 years of the MSKCC experience showed no disease-specific or overall survival for any surgical approach.87 Other outcomes data comparing local recurrence rates and survival are similar, and do not demonstrate a survival difference between the two approaches; therefore the preservation of fecal continence is a priority when possible. A number of retrospective series published from 1990 to 2003 reviewing institutional experience with local excision and APR found that 5-year survival rates range from 0 to 29% for those undergoing wide local excision, and from 0 to 26% for those undergoing APR. Most authors now recommend wide local excision with negative margins for those patients without anal sphincter involvement.82

Even though survival differences are minimal between local and radical approaches, local recurrence rates may be higher after local excision. A study from the M.D. Anderson Cancer Center found that recurrence after local excision was significantly higher than recurrence after APR (58 vs 29%), and that median survival times were the same (~19 months for both groups).88 Patients with local recurrence in this study developed synchronous regional and distant disease. Roumen in the Netherlands also reported an increased rate of local recurrence with local excision, but no overall survival disadvantage.89

Inguinal lymph node dissection in anorectal melanoma is usually reserved for those with clinically positive nodes and is a palliative intervention. Prophylactic nodal dissection does not seem to provide a survival benefit and there currently is no clear indication for it. The role of sentinel lymph node mapping in this disease is not clear. The benefits of the technique are now well established in cutaneous melanoma, but it has not been investigated in anorectal melanoma and is not currently routinely performed.

Adjuvant Therapy

High-dose interferon-α (IFN-α) is currently used in the treatment of cutaneous melanoma. It confers a survival benefit in this group, improving disease-free survival rates.82 However, there are no data demonstrating its efficacy in anorectal melanoma, and current reports of adjuvant chemotherapy in this setting are anecdotal. In fact, two large meta-analyses show no consistent benefit from the use of IFN-α.90,91

Aggressive multidrug chemotherapeutic regimens show some potential efficacy, but toxicities with these approaches are considerable.92,93 Temozolomide, whose metabolite is dacarbazine has shown efficacy against cutaneous melanoma,94 and is being investigated for use in anal melanoma. Innovative new approaches, guided by molecular targets are also underinvestigation,95 but currently, there is no data to guide effective adjuvant chemotherapy.

External beam irradiation for symptomatic pelvic and local recurrences and metachronous inguinal nodal disease has been incorporated into the palliative treatment of anorectal melanoma, but again, no data are available to assess overall efficacy. It seems reasonable, however, to extrapolate treatment paradigms from cutaneous melanoma to anorectal melanoma in stage IV disease.

The surgical treatment of anorectal melanoma has changed over time, evolving from radical to local excision. No survival benefit is conferred by APR in most studies, and in most reviews survival is quite poor in spite of surgical excision, with median survival less than 20 months from the time of diagnosis. Although adjuvant chemotherapy is shown to be effective in cutaneous melanoma, lack of data hinders acceptance of this therapy in anorectal melanoma.

Anal adenocarcinomas are uncommon, comprising 10% of all anal (canal) carcinomas.82 Symptoms of bleeding, pain, and change in bowel habits are nonspecific and similar to other anal canal and distal rectal neoplasms. Anal adenocarcinomas may occasionally arise from chronic anal fistulas.

Although outcomes data are few, anal adenocarcinoma has a poor prognosis when compared to rectal cancers or anal SCC. In small series, 5-year survival rates range from 64% to less than 5%.96,97 These neoplasms have a high rate of both local and distant failure.97

Treatment is similar to therapy for adenocarcinoma of the rectum. Neoadjuvant chemoradiation followed by surgical excision is recommended. Postoperative adjuvant chemotherapy may be prudent, as it is in rectal adenocarcinoma, to reduce the risk of distant spread.

Paget's disease was first described in 1874 by Sir James Paget, who reported 15 cases involving the nipple.98 To date, less than 200 cases of perianal Paget's have been documented in the literature since Darier and Couillaud reported the first case in 1893.98

Paget's has a female predominance (1.5:1) with a median presentation age of 65 years.99 The disease is usually present for an extended period of time prior to diagnosis because the symptoms are nonspecific and often mistaken for a benign dermatitis. Paget's occurs in apocrine, hair-bearing areas. Erythematous, pruritic, scaling plaques with well-defined serpiginous borders are a typical feature of the disease. These lesions may also appear ulcerated and crusty with a serous discharge. The disease can be found in both the anal canal and margin.82,98 Histologically, Paget's disease is defined by the presence of large intraepidermal anaplastic tumor cells lying separately or in small clusters. Perianal Paget's cells are foamy and vacuolar in appearance and stain light blue with hematoxylin and eosin. They are positive for periodic acid-Schiff, mucicarmine, Alcian blue, and cytokeratin 7.82

The pathogenesis of Paget's disease is still somewhat unclear. Because it can be associated with the presence of rectal adenocarcinoma, it is speculated that Paget's represents a downward extension of the tumor or that a “neoplastic milieu” may create an environment hospitable to the presence of multiple GI primary tumors. Another hypothesis holds that it is a primary tumor of the apocrine glandular elements of the distal anal canal and margin. Others have suggested that Paget's may arise from a neoplastic pluripotent epidermis basal cell.82

Perianal Paget's is associated with an underlying visceral malignancy in 20–86% of cases.82,98,99 Colorectal adenocarcinoma is the most common synchronous tumor, but urogenital, breast, and bile duct carcinomas have also been reported. Screening for other malignancies is imperative. A colonoscopy and prostate examination are basic preventive and diagnostic tests that can be helpful. Some authors recommend computed tomography of the abdomen and pelvis as well.98

Complete excision is the treatment for Paget's disease. The extent of the disease is usually determined by taking circumferential biopsies of the anal canal and margin. After the disease is mapped, wide local excision is performed. Often, the procedure creates large defects that may require skin grafts or flaps (rotational, island, or myocutaneous). Because excision to negative margins is critical to cure, techniques to ensure this may be required. Surgeons may obtain frozen sections of the margins of the specimen in the operating room prior to reconstruction. Others prefer to cover the wound with saline-soaked gauze, admit the patient to the hospital, and await permanent pathology results for up to 2–3 days prior to reconstruction. If a large flap reconstruction is placed in the anal canal, some recommend diversion with a colostomy or ileostomy at the time of the perineal excision.

Recurrence rates as high as 61% have been reported following excision of perianal Paget's disease.100 Re-excision is the usual recommendation, although in cases in whom underlying rectal or anal adenocarcinoma exist, radiation followed by APR is advisable. Although recurrence rates are high, the prognosis of Paget's limited to the perianal area with no concomitant neoplasm is very good.82 Because of the association with additional visceral neoplasms, continued surveillance is required for patients with perianal Paget's disease. Physical examination, including a prostate and pelvic examination, and periodic colonoscopy are probably prudent. Biopsies of new lesions at the edges of the flap or graft may reveal residual disease. Local excision of these recurrences and continued surveillance is required.

Buschke-Löwenstein tumors are also referred to as giant condylomas and were first described in 1925 by Buschke and Löwenstein as “carcinoma-like condylomata acuminata.”101 They are rare entities belonging to a wider group of lesions called verrucous carcinomas, which include oral and cutaneous fungating condylomata. The key feature of giant condyloma that differentiates it from benign anal condyloma is the presence of local invasion.

Although the natural history of these lesions is poorly understood, the etiology is assumed to be similar to that of condyloma.101 Human papillomavirus (HPV) has been isolated from the tumors. Histologically, the lesions are benign in appearance and do not invade the basement membrane as carcinomas do. Instead, they destroy surrounding tissue by expansion rather than direct invasion (Fig. 42-6). The tumor does not metastasize. Deaths from untreated Buschke-Löwenstein tumors have occurred following deep invasion into unresectable pelvic structures followed by superinfection and recurrent sepsis. The overall mortality rate from this rare entity is 20%.102

The literature consists almost entirely of case reports,103 thus, there are no consistent treatment guidelines. Primary treatment consists of surgical resection to clear margins.102 However, adequate surgery may be impossible when the tumor deeply invades the pelvis. There have been several case reports demonstrating the efficacy of intralesional injection of interferon-α 2b.101 At least three reported cases of deeply infiltrating giant condylomata have completely responded to long-term therapy, including one patient who would have required hemipelvectomy with limb amputation to achieve negative margins. Interferon-α 2b may be a good alternative or supplement to radical resection in select cases. Long-term outcomes are not available.

Gastrointestinal stromal tumors (GISTs) are of mesynchymal origin, believed to be derived from the interstitial cells of Cajal. Immunohistochemical analysis shows the majority to be positive for CD34 and CD117 (c-kit). The vast majority of GISTs affect the stomach and small bowel; only 5% of GISTs affect the anorectum, and purely anal GIST comprise only 2–8% of these.104 Because there are so few, most reports consider anal and rectal GISTs together for analysis.

Current assessment of prognosis for GIST is based on certain clinicopathologic features; tumor size greater than 5 cm, and mitoses greater than 5/high-power field (HPF) predict malignant behavior and decreased survival. Other factors such as pleomorphism, tumor necrosis, and infiltration of muscularis propria and symptoms are indicative of more aggressive behavior, but effects on survival are less clear.104,105

Due to the exceeding rarity of this tumor, few data-driven recommendations can be given but extrapolation from management of rectal GISTs demonstrates that surgical excision is recommended for all tumors. Smaller tumors (<2 cm) may be excised locally, while larger, or more locally invasive tumor may require APR for an R0 resection. Local recurrence is common after limited resection, up to 60% in one series,104 while radical resection yielded 100% recurrence-free survival. Anecdotally, salvage APR for local recurrence after limited excision may yield affective long-term disease control. However, there is no data available to show that extent of surgery can affect overall survival.

While adjuvant therapy with molecularly targeted tyrosine kinase inhibitors like imatinib (Gleevec) and sunitinib (Sutent) for advanced or metastatic disease is becoming standard of care for most GISTs, however there is no data yet to support or extrapolate its use in anal GISTs.

Although Kaposi's sarcoma (KS) is the most common malignancy in AIDS, the incidence of anorectal KS is still quite small in this population. Current HAART treatment makes this number continue to fall.106 Anal KS usually presents as pedunculated or round purplish-red anal lesions that can easily be mistaken for prolapsing hemorrhoidal tissue. Treatment is primarily medical, with chemotherapy directed at the tumor, as well as appropriate antiretroviral therapy. Use of external beam radiation has also been described for cutaneous anal lesions.107