Squamous cell carcinoma of the anal margin or perianal skin is relatively uncommon, and most physicians, even those practicing at large referral centers, encounter very few patients with this entity. The goal of treatment is to cure the patient while preserving anal function, thus avoiding a permanent colostomy. Traditionally, treatment has consisted of either local excision or, in advanced cases, abdominoperineal resection (APR). In recent years, a few centers have reported promising results with radiotherapy alone or combined with concomitant chemotherapy. The purpose of this paper is to review the epidemiology, diagnosis, staging, natural history, and results of treatment for this disease.
Squamous cell carcinoma of the anal margin is defined as a lesion originating between the dentate line and the outer limit of the perianal skin, defined to be 5 cm from the anal verge in any direction.[1-5] These lesions represent only one-fourth to one-third of all squamous cell carcinomas of the anus and should be distinguished from squamous cell carcinoma of the anal canal, which has a different natural history and a less favorable prognosis.[2,4-7]
Most patients with anal margin carcinoma are 60 to 70 years old, but the age range of affected individuals is wide (approximately 25 to 90 years). Although some authors have observed a slight female preponderance, others have reported that the disease is more likely to occur in men.[3,5,6,8]
Jensen et al observed the following symptoms in 76 patients with squamous cell carcinoma of the anal margin treated in Denmark: palpable mass (100%), bleeding (78%), pain (70%), change in bowel habits (29%), discharge (20%), and pruritus ani (20%). The median duration of symptoms was 6 months (range, 2 to 60 months). Associated condylomata and chronic fistulae are observed in approximately 15% of patients. Jensen et al observed that an erroneous diagnosis was made at the first physician visit in 29% of patients with squamous cell carcinoma of the anal margin, as compared with 55% of 125 patients with squamous cell carcinoma of the anal canal.
The majority of anal margin tumors tend to be well- or moderately differentiated keratinizing squamous cell carcinomas; less than 10% of lesions are poorly differentiated or cloacogenic carcinomas.[2,9]
Squamous cell carcinoma of the anal margin is staged according to either the American Joint Committee on Cancer (AJCC) or Union Internationale Contre le Cancer (UICC) staging system.[2,12,13] The AJCC staging system for anal margin carcinoma is the same one used for other skin cancers (Table 1) and differs from the staging system used for the anal canal. The AJCC and UICC systems for anal margin carcinoma are virtually identical.[10,11]
The primary tumor usually starts as a slow-growing nodule that remains localized to the perianal skin until late in the course of the disease, when it may invade the anal canal. The lesion is usually ulcerated and may have a significant palpable subcutaneous component. The sphincter muscle is rarely invaded. The distribution of primary tumor size varies, depending on referral patterns. Pinna Pintor et al found the following UICC T-stage distribution in 83 patients treated at St. Mark's Hospital for Diseases of the Colon and Rectum in London: stage T1, 14%; stage T2, 50%; stage T3, 32%; and stage T4, 4%.
The medial inguinal nodes are the first-echelon lymph node drainage for the anal margin, whereas the perirectal nodes are the first-echelon drainage for the anal canal. The iliac nodes are also occasionally involved. The incidence of inguinal lymph node involvement is approximately 15% to 25%, and is related to the size and histologic differentiation of the primary tumor.[2,3,14,15]
Cummings et al reported on the relationship between primary tumor diameter and the risk of inguinal lymph node invasion at diagnosis in a series of 29 patients treated at the Princess Margaret Hospital, Toronto. They found inguinal node invasion in 0 of 13 (0%) of patients with tumors less than 5 cm in diameter, as compared with 4 of 16 (25%) of those with tumors 5 cm or more. In 57 patients, Papillon and Chassard documented the following rates of inguinal lymph node involvement, according to primary tumor size: less than 2 cm, 0 of 10 (0%); 2 to 5 cm, 9 of 38 (24%); and 5 cm or more, 6 of 9 (67%).
Distant metastases are rare at presentation.
The pretreatment evaluation of the patient should take into account the spread patterns of the disease and should include a chest roentgenogram and CT scan of the abdomen and pelvis. Computed tomography is obtained to evaluate the presence and extent of lymph node metastases, exclude the unlikely possibility of liver metastases, and complement the physical examination of the primary lesion.
Early lesions of the anal margin may be successfully treated with local excision; a skin graft may be necessary if the surgical defect cannot be closed primarily or healed by secondary intention. An APR is necessary for resection of more advanced lesions. The inguinal lymph nodes are not dissected unless they are deemed to harbor metastatic disease.
Greenall et al  reported on 31 patients treated with local excision alone at Memorial Sloan-Kettering Cancer Center between 1950 and 1978. Local recurrence alone developed in nine patients (29%), one patient had recurrences at both the primary site and the inguinal lymph nodes, and isolated inguinal node metastases developed in three patients. Of the nine patients who experienced a local recurrence alone, eight underwent a second local excision and one required an APR. The 5-year absolute and cause-specific survival rates were 68% and 88%, respectively.
Greenall et al  described an additional 11 patients who underwent a primary APR; one patient died postoperatively and two patients died secondary to recurrence. Seven patients (64%) were alive and disease-free 5 years or more after surgery.
At the Cleveland Clinic, 10 patients were treated with local excision between 1951 and 1971, as reported by Al-Jurf et al. Local recurrence developed in 3 patients (30%), 2 of whom were salvaged by a second local excision. Seven patients were alive and disease-free at 5 years or more, one patient died of intercurrent disease at 15 months, one patient was alive with disease at 6 years, and one patient died of disease at 8 years.
Schraut et al  reported on 16 patients treated surgically at the University of Chicago. The disease was controlled in 9 of 11 patients after a local excision and in 4 of 5 after an APR.
Beahrs and Wilson described 27 patients treated with local excision for in situ or superficial squamous cell carcinoma of the anal margin at the Mayo Clinic between 1950 and 1970. The local control rate was not stated; all patients apparently survived 5 years.
In a series of 49 patients who underwent local excision alone or combined with radiotherapy at St. Mark's Hospital for Diseases of the Colon and Rectum, London, the 5-year absolute and cause-specific survival rates were 65% and 68%, respectively. An additional 16 patients underwent an APR alone or combined with radiotherapy; 5-year absolute and cause-specific survival rates were 38% and 40%, respectively. Local control rates after surgery were not presented.