Carcinoma of the anal canal accounts for about 1.9% of all malignant tumors of the digestive system in patients in the United States.¹ Despite the rarity of anal cancer, it is a model for successful oncology research, both in the laboratory and in the clinic. Epidemiologic observations about the increased incidence of anal cancer in some populations, along with advances in molecular biology that have allowed the identification of human papillomavirus (HPV) DNA in most anal tumors, have provided the initial clues to the mechanism of anal carcinogenesis. Retrospective studies have provided important information about the natural history and patterns of spread of anal cancer, as well as hypotheses to test in prospective trials. Prospective randomized trials have been completed successfully and have led to the adoption of a combination of radiotherapy and chemotherapy as the standard of care for patients with anal cancer. Questions remain, however, about the most effective and least toxic regimens of radiotherapy and chemotherapy.

This chapter focuses on the biologic characteristics of anal cancer, the rationale for patient evaluation, treatment recommendations, and results of treatment. Radiotherapy techniques for patients with anal cancer will also be discussed.

Etiology and Epidemiology

Anal cancer occurs much less frequently than other types of cancer of the digestive tract. It accounts for only 2% to 4% of all cancers of the anus or rectum.2,3,4,5 In the United States, the annual incidence is 0.47 per 100,000 white men and 0.69 per 100,000 white women.⁶ The annual incidence among African Americans is higher: 0.57 per 100,000 men and 0.78 per 100,000 women.⁶ Overall, 87% of patients diagnosed with anal cancer are non-Hispanic whites, 5% are African American, and 3% are Hispanic.⁷ The median age at diagnosis is 62 years.⁷ Thus, the typical American patient with anal cancer is a white woman in her seventh decade of life. In the United States, 5260 new cases of anal cancer (2000 men and 3260 women), were estimated for 2010, with an estimated 720 deaths (280 men and 440 women).¹

Epidemiologic studies from Europe and the United States have reported an increased incidence of anal cancer in the past 30 to 40 years. Since 1960, the incidence of anal cancer in Connecticut has doubled in both men and women.⁶ Between 1974 and 1985, the number of patients in Sweden diagnosed with anal cancer increased 4% per year, an increase similar to that reported for Swedish patients with malignant melanoma.⁵ Similarly, the number of new cases of anal cancer in Denmark has doubled in men and tripled in women.⁸

The incidence of anal cancer is higher in urban than in rural populations, and increases in the incidence of anal cancer have been greater in urban than in rural populations.^5,^6,⁸ In the United States, the increased incidence of anal cancer has been limited solely to densely populated regions. Young men account for a substantial proportion of this increase. In Denmark, the median age at diagnosis in men has decreased from 68 to 63 years, whereas in women, it has remained constant at 66 to 67 years.⁸

Association with HIV Infection

At least part of the increased incidence in young men can be explained by the observation that young homosexual men are at increased risk for the development of anal cancer, irrespective of their human immunodeficiency virus (HIV) status. An increased risk for anal cancer among never-married men, a surrogate marker for homosexuality,⁶ has been noted in several epidemiologic studies.⁹^–¹² In addition, men with anal cancer are more likely to never have married compared with control subjects who have stomach or colon cancer.^6,⁸ In most areas of the world, anal cancer is more common in women than in men of all age groups.^13,¹⁴ However, in areas with a relatively high proportion of homosexual men, anal cancer may be more common in men. In San Francisco, for example, the incidence of anal cancer in white men more than doubled from 0.53 per 100,000 in 1975 to 1.2 per 100,000 in 1989,⁶ and in Los Angeles, anal cancer has become more common in men than in women under age 35.⁹ A study of Danish homosexual men living in legally registered “marriage-like partnerships” identified an incidence of cancer double what was expected, which could be accounted for by HIV-related cancers, including Kaposi sarcoma, non-Hodgkin’s lymphoma, and anal cancer. The relative risk (RR) of anal cancer in this population was 31, and anal cancer appeared to be associated with a positive HIV status.¹⁵

The incidence of anal cancer is markedly increased in both men and women who have the acquired immunodeficiency syndrome (AIDS). The RR estimated from a linkage analysis of cancer registries and AIDS registries is 63, and it is higher for homosexual men (RR, 84) than for heterosexual men (RR, 38). The increased risk is also apparent during the 5-year period before an AIDS diagnosis. The absolute risk of anal cancer in AIDS patients is 1 per 1000.¹⁶

Risk Factors

The pathogenesis of anal cancer is multifactorial. A diagnosis of anal cancer represents the result of an interplay of multiple environmental and host factors. Patients with epidermoid anal cancer are more likely than control patients to have had a previous diagnosis of malignant disease, including cancers of the vulva, vagina, or cervix and lymphoma or leukemia. Patients with anal cancer who are diagnosed before age 60 years are at higher risk of subsequent development of malignant diseases of the respiratory system, bladder, vulva, vagina, and breast, but they are not at increased risk for subsequent development of colon or rectal cancers.¹⁷ Overexpression of the c-myc oncogene has been implicated in the pathogenesis of both anal squamous cell neoplasia and breast cancer.^18,¹⁹ This finding, along with an observed pattern of second malignant diseases and prior malignant diseases in patients with anal cancer, suggests a multifactorial pathogenesis and common oncogenetic risk factors. These factors may include sexually transmitted viruses, environmental carcinogens such as cigarette smoke, immunosuppression, and genetic susceptibility.¹⁷

Anal cancer and cancers of the female genital tract share a common pathogenesis.²⁰ In the embryo, the anal and cervical canals are both derived from closely related anlagen.²¹ Women patients with anal cancer are more likely than women with colon or stomach cancer to have had a prior diagnosis of cervical intraepithelial neoplasia.²⁰ An association between anal cancer and the number of lifetime sexual partners has been reported in women.²² Case-control studies have also found an association between anal cancer and some sexually transmitted diseases.^13,²²^–²⁴ This association, along with the increased incidence in young homosexual men, strongly suggests an etiologic factor for anal cancer associated with increased sexual activity.⁶

Among sexually transmittable infectious agents, HPV has been the most thoroughly studied potential causative agent. Although there are more than 60 types of HPV, those most commonly associated with benign genital condylomata acuminata are HPVs 6 and 11, whereas HPVs 16, 18, 31, 33, and 35 are associated with malignancy or high-grade dysplasia.²⁵ Several investigators have also reported an association between genital warts and anal cancer in men and women.²³^–²⁷ The development of anal warts in both sexes has been linked with the practice of anal intercourse.²⁴

The likelihood of finding HPV DNA in specimens of anal cancer varies according to patient demographics and laboratory technique. Patients with HPV-associated anal cancer are 10 years younger on average than patients with reportedly HPV-negative cancers.²⁸ In addition, HPV-associated anal cancer has been reported more frequently in Europe and South America than in South Africa and India.²⁹ When in situ hybridization techniques are used to analyze biopsy specimens of anal cancer, the rate of HPV DNA detection ranges from 17% to 73%.^25,²⁸ HPV DNA is more likely to be detected with polymerase chain reaction analysis. Studies comparing both techniques have found that polymerase chain reaction techniques detect HPV DNA in 78% to 85% of patients, whereas in situ hybridization techniques find HPV DNA in only 17% to 50% of the same patients.^25,³⁰

Exposure to HPV may be a risk factor even for patients in whom HPV DNA is not detected in the carcinoma specimen. Serum IgA antibodies to a peptide antigen from the E2 region of HPV 16 have been found in 89% of patients with anal cancer compared with only 24% of controls.²⁸ In another study, antibodies to HPV 16 capsids were elevated in 55% of patients with anal cancer compared with 3% of controls. Antibodies to HPV-capsid antigen were detected in an equal number of cancer specimens from patients whose HPV DNA was negative or positive (as assessed by in situ hybridization), which suggests that some presumed HPV-negative patients had been exposed to HPV.³¹

HPV infection and anal cytologic abnormalities are common in patients with HIV infection.³² HIV-positive patients with HPV DNA found in anal biopsy specimens seem to have a high rate of cytologic abnormalities. In one study of 12 HIV-positive patients without AIDS, 11 patients (92%) with normal cytologic findings but with HPV DNA found in anal mucosa later had cytologic abnormalities at 17-month follow-up.³³ Many AIDS patients most likely die of opportunistic infections before anal cancer is manifested. Although the impact of effective antiretroviral therapy on the incidence of anal cancer is not yet clear,³⁴ further increases in the incidence of anal cancer may be observed in AIDS patients as survival time is prolonged.³²

Although the true influence of HPV in anal carcinogenesis is uncertain, some laboratory findings point to the involvement of the HPV E6 and E7 proteins. These findings include the observation that the E6 and E7 oncogenes are consistently expressed in tumor cells but that normal E6 and E7 regulatory mechanisms are absent.³⁵ In addition, the E7 protein of HPV 16, in cooperation with the E6 protein, is able to transform mammalian cells in vitro while blocking E6 and E7 gene function, which results in reversal of the malignant phenotype.³⁵^–³⁷ Interactions of HPV oncoproteins with known tumor suppressor gene products have been reported. The E6 protein of HPV 16 and HPV 18 forms stable complexes with the p53 protein product of the P53 tumor suppressor gene.³⁸^–⁴⁰ The cellular p53-E6 protein complex results in a lack of appropriate G1 arrest and subsequent genomic instability. The E7 protein also forms complexes with the retinoblastoma gene product and with p107, p130, p33, cdk2, and cyclin A. E7 also activates the cyclin-A promoter and overrides two inhibitory functions that restrict the expression of cyclin A and cyclin E.³⁶

Some evidence suggests that infectious agents other than HPV may contribute to anal carcinogenesis. Associations have been found between the development of anal cancer and a history of syphilis or gonorrhea in men ^12,²³ and between anal cancer and Chlamydia and herpes simplex virus type 2 in both sexes.^22,²⁴ In a study of patients in and around San Francisco, herpes simplex virus DNA was detected in 3 of 15 patients with invasive anal cancer and in 3 of 4 patients with high-grade intraepithelial neoplasia.³⁰ No Epstein-Barr virus or cytomegalovirus DNA was found in the tumor specimens of the 15 patients with anal cancer. Further studies are needed to determine whether these and other sexually transmitted agents are involved in anal oncogenesis or are merely surrogate markers for probable HPV infection.

Chronic immunosuppression is associated with an increased risk for anal malignant disease. Renal transplant recipients have an increased incidence (as high as 100-fold) of carcinoma of the vulva or anus.^41,⁴² These cancers occur at an earlier age than they do in the general population.⁴¹ Renal transplant patients also have an increased incidence of cutaneous neoplasia and viral warts. Those with a high susceptibility to cutaneous malignant disease (≥4 skin cancers) are more likely to have HPV DNA-associated skin cancer and are more likely to have an anogenital malignant tumor.⁴³ The increased incidence of anal malignant disease in HIV-positive patients is most likely caused in part by chronic immunosuppression.

A number of authors have reported an association between anal cancer and cigarette smoking.^22,^24,^26,⁴⁴ When compared with a control group of patients with colon cancer, both men and women smokers were found to have an increased risk of anal cancer.²⁴ In a population-based case-control study of patients with anogenital cancers in the Pacific Northwest, 60% of patients with newly diagnosed anal cancer were current smokers compared with 25% of the controls. The risk of anal cancer was positively correlated with both the number of cigarettes smoked per day and the number of years the patient had been a smoker.⁴⁴

An association between anal cancer and benign anal conditions (e.g., hemorrhoids, anal fissure, or fistula) has been reported frequently,^14,⁴⁵ and chronic irritation or inflammation of the anal tissue has been assumed to play a role in anal carcinogenesis.²⁶ In a Danish population-based study, patients with anal fissure, fistula, perianal abscess, or hemorrhoids were found to be at increased risk for anal cancer. However, the RR for invasive anal cancer was highest in the year immediately after a diagnosis of benign anal pathology, and it declined from a high of 12 the first year to 1.8 after 5 or more years, suggesting that a so-called benign anal condition is often a symptom rather than a cause of anal cancer.⁴⁶ Supporting evidence for this view is found in a study of patients treated for benign anal conditions at U.S. Veterans Affairs hospitals. The elevated RR for anal cancer in these patients was most pronounced in the first year after the benign pathologic condition was diagnosed, and it decreased rapidly thereafter until there was no increased risk of anal cancer between year 5 and year 22.⁴⁷

Prevention and Early Detection

Prevention of anal cancer should include educational efforts about the causal link of sexually transmitted HPV infection with malignant diseases of the anogenital tract. Recommendations of the 1996 National Institutes of Health Consensus Panel on cervical cancer prevention are also applicable to anal cancer and include encouragement to delay onset of sexual intercourse.⁴⁸ Barrier methods, such as condoms, do not prevent HPV transmission.⁴⁹ The panel also recommended development of an effective vaccine to prevent transmission of HPV, and a subsequent phase III clinical trial provided encouraging results in this regard.⁵⁰ Additional preventive efforts should be focused on the treatment of HPV infection, including the development of antiviral agents, targeting of E6 and E7 to block transforming activities, and vaccines to prevent progression of HPV infection.^48,51 Education is also required about the causal role of cigarette smoking in anal and other cancers.

Because of the rarity of anal cancer, efforts at early detection through widespread screening are not feasible. However, screening may be feasible in certain high-risk subsets. Anal intraepithelial neoplasia is a common finding in HPV-infected persons.^32,³³ Analogous to the situation with cervical cancer, where cervical squamous intraepithelial lesions may progress to cervical carcinoma, high-grade squamous intraepithelial lesions (HSIL) may be a precursor to invasive anal cancer.⁵²^–⁵⁴ Anal cytologic smears have been used to diagnose HSIL in high-risk patients. Because the specificity of anal cytology for the detection of HSIL is low, anoscopy with biopsy is required to differentiate anal condylomata acuminata from HSIL.⁵² However, the colposcopic criteria to distinguish low-grade squamous intraepithelial lesions (LSIL) from cervical HSIL have been used to distinguish LSIL from HSIL of the anus in homosexual and bisexual men.⁵⁵ This information may allow for a targeted biopsy of suspected HSIL, resulting in increased sensitivity for detection. One study of anal cytology in homosexual and bisexual men reported the sensitivity and specificity for detection of HSIL to be 69% and 59%, respectively, in HIV-positive men and 47% and 92%, respectively, in HIV-negative men.⁵⁴

Biologic Characteristics

Unlike most gastrointestinal tract malignant diseases, anal cancer is predominantly a locoregional disease, and distant metastasis is relatively rare. Most relapses after curative therapy are located in the pelvis, perineum, or inguinal regions.⁵⁶ Only 5% to 10% of patients will have cancer that has spread beyond the pelvis at diagnosis,^45,57,58 and 10% to 20% of patients will have disease relapse at distant sites after curative local therapy.^56,59,60 Although chemotherapy is considered a component of standard therapy for anal cancer, its addition has not decreased the number of patients with distant metastases.^58,59,60 As the number of metastatically involved regional nodes increases, so does the risk of distant metastasis.⁶¹ The most common site of distant metastasis is the liver, followed, in variable order, by the lungs, extrapelvic lymph nodes*, skin, or bones.

Characteristics of anal cancer that have consistently been correlated with local control and survival are the size and extent of the primary tumor and the status of the inguinal and pelvic lymph nodes. The 5-year survival for patients with tumors 2 cm or more in diameter that are treated surgically is about 80% but declines to 55% to 65% when the tumor is 2 cm to 5 cm or to 40% to 55% when the tumor is more than 5 cm.61,65 The depth of invasion and the size and extent of the primary tumor are prognostic for response to treatment, local control, and survival when patients are treated nonsurgically with radiotherapy and chemotherapy.*

The survival for patients with regional lymph node metastases who are treated primarily with surgical procedures with or without adjuvant therapy is about half that observed in similar patients without nodal metastases.^65,⁶⁹ Similarly, 5-year survival after radiotherapy alone in patients with inguinal lymph node metastases range from 0% to 36% compared with 5-year survival of more than 50% in patients without lymph node involvement.⁶⁹^–⁷¹ There is relatively little information in the medical literature about the prognostic significance of regional nodal metastases in patients who receive combined radiotherapy and chemotherapy. In a retrospective, recursive, partitioning analysis of patients treated at Princess Margaret Hospital, node-positive patients at 5-year follow-up were found to have a trend toward lower cause-specific survival (CSS) (57% vs. 81%; p = .07).⁵⁹ In the European Organization for Research and Treatment of Cancer (EORTC) randomized trial of radiotherapy alone versus radiotherapy and chemotherapy, regional nodal metastases were associated with significantly worse local control (p = .004) and survival (p <.001).⁷² However, in node-positive patients, the number of involved nodes (≥1), their size (<2 cm or >2 cm), and the extent of involvement (stage N1 or N2 and N3) added no further prognostic information.⁷²

Patient-related variables that have been evaluated as potential prognostic factors for local control and survival in patients with anal cancer include age, gender, race, and performance status. Age probably does not have independent prognostic significance for any endpoint. A retrospective Canadian study found that older patients were treated with less aggressive chemotherapy and were less likely to be offered salvage surgery for recurrence; therefore patients older than 65 years of age were found to be at higher risk for death from anal cancer.⁷³ Conflicting data exist about the impact of gender on prognosis. Several studies have reported no difference in outcome by gender in patients treated with radiotherapy or combined radiotherapy and surgery.^59,71,73,74 However, a trend toward higher survival in women has been reported in some surgical series,^61,65 and in the EORTC randomized trial, female sex was associated with better survival (p = .12) and local control (p = .05).⁷² Lower survival has been reported in patients with a lower performance status and in nonwhite patients with anal cancer who received combined-modality therapy.⁷⁵

Several histopathologic variables have been evaluated as potential prognostic factors in patients with anal cancer. The histologic subtype (squamous cell cancer vs. cloacogenic subtype), tumor cell morphology, extent of differentiation or keratinization, cell size, architecture, and pleomorphism are of no prognostic significance.59,71,73,74,76 The extent of differentiation may be associated with tumor stage because patients with advanced-stage cancer tend to have tumors that are less differentiated.⁶¹ The depth of invasion of the primary tumor has been reported to have prognostic significance in patients treated primarily with surgical resection.⁷⁶ DNA ploidy has also been associated with prognosis in surgically treated patients with DNA aneuploid tumors predictive of an inferior outcome.⁷⁶ In a Mayo Clinic study of surgically treated patients, those with aneuploid tumors had inferior survival compared with patients with DNA diploid or tetraploid tumors on univariate analysis. On multivariate analysis, however, DNA ploidy was not a significant predictive variable.⁷⁷

There are currently no clinically useful tumor markers. Despite negative liver imaging, patients with elevated alkaline phosphatase or lactate dehydrogenase are at increased risk for subsequent liver metastases.⁷⁸ In one study, reduced expression of p21^waf1, a cyclin-dependent kinase inhibitor, was associated with shorter OS.⁷⁹ Preliminary information suggests that serum antibodies to HPV proteins may eventually be useful as prognostic markers. Elevated serum IgA antibodies to HPV 16 E2 : 9 peptide have been associated with lower survival rates independent of tumor size.²⁸ In another study, patients who died of anal cancer were found to have higher levels of IgG against HPV E7 : 5 peptide than did patients with anal cancer in complete remission or patients who died of other causes.³¹

Pathology and Pathways Of Spread

The World Health Organization (WHO) classification of malignant epithelial tumors of the anal canal includes squamous cell carcinoma, adenocarcinoma, small cell carcinoma, and undifferentiated carcinoma. Most cases of adenocarcinoma of the anus are actually distal rectal adenocarcinomas with extension into the anal canal. True primary anal canal adenocarcinoma is rare, as is small cell carcinoma of presumed neuroendocrine origin.14,61 Subsets of squamous cell carcinoma in the WHO system include large-cell keratinizing carcinoma, large-cell nonkeratinizing or transitional carcinoma, and basaloid carcinoma.⁸⁰ Although basaloid carcinomas have historically been considered to be nonkeratinizing tumors, most basaloid carcinomas exhibit keratinization and should be considered squamous cell carcinomas.⁵³

Although most anal canal malignant tumors are squamous cell carcinomas or squamous cell variants, marked morphologic heterogeneity is characteristic.⁵³ Classification of epidermoid anal cancers on the basis of morphologic appearance has led to the use of several potentially confusing terms. These include transitional cell carcinoma, basaloid carcinoma, and mucoepidermoid carcinoma. These tumors all arise from the anal transition zone and are often grouped together as cloacogenic carcinoma. Mucoepidermoid carcinomas contain mucous microcysts and are histologically dissimilar to mucoepidermoid carcinomas of the salivary glands. The natural history is identical to that of squamous cell carcinoma of the anus without microcysts. Although sometimes considered to be distinct lesions in the medical literature, cloacogenic carcinoma, transitional cell carcinoma, basaloid carcinoma, and mucoepidermoid carcinoma are all subtypes of squamous cell carcinoma, because patients with these various tumor subtypes have similar clinical characteristics and the tumor subgroups do not differ in natural history or response to therapy.^14,^45,^56,⁸¹

Primary anal melanoma is a rare tumor that accounts for only 1% of all anal cancers. Anal melanoma is similar to melanoma of the skin in that it rarely affects African Americans and is characterized by the distant spread of disease.^3,⁸² Outcome is poor after wide local excision or abdominoperineal resection, with just a 10% survival in most series at 5-year follow-up.⁸³

Pathways of Tumor Spread

Anal cancer tumors spread by direct extension to surrounding tissues, lymphatic dissemination to pelvic and inguinal lymph nodes, or hematogenous spread to distant viscera.⁸⁴ At diagnosis, about half of all anal cancers have been found to invade the anal sphincter or surrounding soft tissue. Although Denonvilliers fascia is usually an effective barrier to prostatic invasion in men, direct extension to the rectovaginal septum is a common occurrence in women.³

The anal canal has several potential lymphatic drainage pathways. The superficial inguinal nodes are the primary drainage basin for that part of the anal canal distal to the dentate line.^85,⁸⁶ Lymphatic drainage around the dentate line occurs to lymphatic plexuses of the rectal mucosa and along the pathway of the inferior and middle hemorrhoidal vessels to obturator and hypogastric lymph nodes. Lymphatic connections also join the anus to presacral, external iliac, and deep inguinal nodes.⁸⁷

Metastatic involvement of pelvic lymph nodes has been reported in 25% to 35% of patients treated primarily with abdominoperineal resection.45,61 Inguinal node metastases are found in about 10% of patients at diagnosis; the risk depends on the size and extent of the primary tumor.^56,57 The incidence of inguinal node metastases may be as high as 20% for tumors more than 4 cm in diameter and as high as 60% when there is direct invasion of adjacent pelvic organs.⁵⁷ Recurrence in undissected inguinal nodes has been reported in 13% of surgically managed patients with clinically negative inguinal lymph nodes.⁶¹

Clinical Manifestations, Patient Evaluation, and Staging

Rectal bleeding is the most common symptom of anal malignant disease. Perineal pain, mass sensation at the anus, and a change in bowel habits are also frequently reported.^14,^45,⁸⁸^–⁹¹ Many patients with symptoms of early anal cancer are diagnosed initially with a benign anal condition, such as hemorrhoids, anal fissures, or fistulae.^14,⁴⁵

Most patients with anal cancer are diagnosed at an early stage. A national survey of such patients found that 73% were diagnosed with stage 0 to II cancer, whereas only 8% were diagnosed with stage IV cancer.⁷ The interval from the onset of symptoms to diagnosis may be quite prolonged, however, exceeding 1 month in 80% of patients and 6 months in 33%. A study of Norwegian patients with anal cancer found that nearly one third of them had a delay of more than 6 months from onset of symptoms to diagnosis.⁹¹ This finding emphasizes the importance of a thorough digital rectal examination in patients with anal symptoms.^46,^47,⁹¹

Patient Evaluation

Evaluation of the patient with known or suspected anal cancer should begin with a thorough history and physical examination. The patient should be questioned about anal sphincter function and any history of risk factors (sexual or drug abuse) for HIV infection.

In addition to a complete general physical examination, a detailed examination should be conducted of the abdomen, inguinal region, anus, and rectum. The extent of circumferential involvement of the anal canal should be noted, and documentation should be made of the size, extent, and location of the primary tumor. The size, location, and mobility of palpable inguinal lymph nodes should be noted. Pararectal lymph nodes may be involved metastatically, but these are rarely palpable by digital rectal examination.⁷¹

Laboratory studies should include a complete blood cell count, measurement of serum creatinine levels, and liver function studies of bilirubin, alkaline phosphatase, lactate dehydrogenase, and glutamic-oxaloacetic transaminase. For patients with HIV risk factors, a determination of HIV status should be made before the initiation of therapy. Although concentrations of serum carcinoembryonic antigen (CEA) are elevated in 20% of patients with anal cancer, post-treatment CEA values have not been found to correlate with clinical outcomes and have not proven useful in patient management.⁹² There are no other clinically useful tumor markers for anal cancer.

Radiographic evaluation should include chest radiographs and computed tomography (CT) scanning of the abdomen and pelvis. The CT scan is generally inferior to the physical examination for the characterization of primary tumors, but it is useful for the evaluation of the liver and the perirectal, inguinal, pelvic, and para-aortic lymph nodes. Magnetic resonance imaging (MRI) has not yet been proven to be more clinically useful than CT scanning. The depth of invasion of the primary tumor may be evaluated with ultrasonography.⁹³

PET imaging is useful in further evaluating the extent of the primary tumor and the presence of regional lymph node metastases, and distant metastases, as well as in evaluating the response to therapy.⁹⁴^–⁹⁶ A retrospective evaluation of the sensitivity of PET/CT imaging compared with physical examination and CT scanning alone among 41 patients with anal cancer showed that PET/CT imaging detected the primary tumor in 91% of patients compared with only 59% by CT scanning alone.⁹⁷ In addition, PET/CT identified positive inguinal nodes in 17% of patients who were found to be clinically negative by CT and physical examination. Another series reports that use of PET during initial staging led to a change in stage in 23% of patients and radiotherapy field changes in 13% of patients.⁹⁸ Furthermore, a post-therapy PET scan showing resolution of metabolic activity was reported to be highly associated with improved progression-free survival (PFS) (95% at 2 years vs. 22%; p <.0001).⁹⁹

Staging

Anal cancer should be staged according to the TNM staging system of the American Joint Committee on Cancer (AJCC)¹⁰⁰ (Table 50-1). Tumors are classified according to their maximum diameter and their invasion of adjacent structures, as determined by the physical examination and any imaging studies.

TABLE 50-1 2010 TNM Staging System of the American Joint Committee on Cancer

In earlier editions of the staging system of the Union Internationale Contre le Cancer (International Union Against Cancer [UICC]),⁷¹ primary tumors were classified not by size but rather according to the length and extent of their circumferential involvement with the anal canal. Tumors involving less than one third of the length and circumference of the anal canal were classified as T1 category of disease, whereas those that involved more than one third of the length or circumference or that invaded the external sphincter were classified as T2. As in the current system, T3 tumors involved the rectum or perianal skin, and T4 tumors invaded adjacent structures.

The AJCC staging system for anal canal cancer is applicable to all carcinomas that arise from the anal canal. Cancers of the anal margin (distal to the anal verge) are staged as skin cancers, but melanoma of the anal canal is excluded. For staging purposes, the regional lymph nodes in the AJCC system are the anorectal, perirectal, lateral sacral, internal iliac (hypogastric), and superficial and deep inguinal lymph nodes.¹⁰⁰

Primary Therapy

Surgery Alone

Before the establishment in the 1980s of sphincter-sparing therapy as the standard of care for epidermoid anal cancer, most patients with anal cancer in North America were treated surgically with abdominoperineal resection. Reported 5-year OS after abdominoperineal resection for anal cancer ranged from 25% to 70% (average, 50%).61,65,101 Locoregional recurrence developed in 25% to 35% of patients and distant metastasis in 10%.^61,65,101 Patients at highest risk for local recurrence (range, 36% to 48%) after abdominoperineal resection are those with extension of the primary tumor beyond the anal sphincter or metastases to inguinal or pelvic lymph nodes.⁶¹ A component of locoregional disease is present in as many as 84% of patients with relapse after abdominoperineal resection.⁶¹ When the inguinal lymph nodes are metastatically involved, 5-year OS after primary surgical therapy is only 10% to 20%.¹⁰¹ Although abdominoperineal resection is now rarely used initially, it is still useful for treatment of patients with local recurrence after conservative therapy and for management of complications after conservative therapy.¹⁰¹

Irradiation Alone or Plus Chemotherapy

High-dose radiotherapy without surgical resection or other adjuvant therapy is an effective treatment for small stage T1 tumors without inguinal adenopathy. Table 50-2 summarizes the local control and survival results from retrospective series of radiotherapy alone for small anal cancers. For patients with anal cancers of 2 cm or less in diameter, 100% local control 5 years after radiotherapy without chemotherapy has been reported in several small series of less than 10 patients each.^59,60,102 Local control rates at 5 years are lower (57% to 76%) in patients with 2- to 5-cm tumors.^59,60,103

TABLE 50-2 5-Year Disease Control and Survival after Radiotherapy Alone for Early-Stage Anal Cancer