Clinical Features

Anorectal malformations occur in approximately 1 of 1500 to 5000 live births^12,13 and affect both sexes equally. They range from minor anal anomalies to complex cloacal malformations. The most common congenital anomaly of the anus is anal atresia, which accounts for as many as 75% of all anorectal malformations.¹⁴ Anorectal agenesis accounts for approximately 10% of all anal atresias. As much as two thirds of congenital anal anomalies are associated with anomalies in other organ systems, most often the genitourinary system but also the central nervous system, skeleton, cardiovascular system, and gastrointestinal (GI) tract.^14–16 Soon after birth, patients fail to pass meconium or have meconium that extrudes from a fistulous opening.¹⁷

Pathogenesis

Anorectal malformations have many causes, and genetic factors are considered an important component of their pathogenesis. Approximately 5% to 10% of malformations arise in children with a chromosomal abnormality, most commonly trisomy 21, and several monogenetic syndromes include anorectal malformations among their features.¹⁶ Anorectal malformations are more common in siblings of similarly affected patients.¹⁸ Ninety-five percent of patients with trisomy 21 and anorectal malformations have imperforate anus without a fistula, compared with only 5% of all patients with anorectal malformations. Congenital abnormalities, including anorectal malformations, have been associated with conception by using assisted reproduction methods.¹⁹

Pathology

Anorectal malformations are recognized by their gross anatomic features. The Wingspread classification of anorectal malformations is subdivided into high, intermediate, and low atresia based on the level of termination of the anorectum in relation to the levator ani muscle.^20,21 This classification has proved quite useful because it shows good correlation with the type of surgical approach needed for repair.²² Fistula formation is a common finding in anorectal malformations, particularly in high and intermediate forms of anal atresia, and it may be rectovesical, rectoprostatic, rectourethral, or anocutaneous in boys and rectovaginal or anoperineal in girls.

An updated classification of anorectal malformations, the Krickenbeck scheme (Table 32.1), emphasizes the type of fistula associated with the malformation. The type of fistula helps to determine the location of the blind pouch and to guide the surgeon’s expectations regarding the length of the atretic segment to be resected at the surgical pull-through procedure.²² In patients with a cloacal disorder, a single perineal orifice is found where the urinary tract, vagina, and rectum converge into a common channel.

Prognosis and Treatment

A posterior sagittal approach is considered the best method for defining and repairing anorectal anomalies.¹⁸ This approach has greatly improved outcomes in anorectal malformation repair over the past several decades.^22,23

The level of anal atresia (i.e., high, intermediate, or low) in relation to the levator ani muscle is directly related to functional prognosis and rates of fecal continence or constipation after a surgical pull-through procedure.²⁰ Patients with anorectal agenesis have a poor prognosis for fecal continence because they lack a functional puborectalis sling mechanism and internal and external anal sphincters. The prognosis for anal atresia located inferior to the levator ani muscle depends on the presence of functional sphincters and intact sensation. Atresias isolated to the anus, such as imperforate anal membrane, carry the best surgical prognosis.

For patients with cloacal disorders, prognostic factors include the quality of the sacrum, the quality of the muscles, and the length of the common channel. Surgical repair of a common channel less than 3 cm long is feasible for most pediatric surgeons, but for a common channel longer than 3 cm, the repair should be performed at a specialized center by an experienced surgeon.¹⁸

In the first 24 to 48 hours after birth of an infant with an anorectal malformation, the major issues are identification of life-threatening anomalies and deciding whether to repair the defect immediately or to perform a protective colostomy with repair at a later date. These decisions are based on the infant’s physical examination, the extent of the malformation, and any significant changes that occur in the first 24 hours of life. After this critical period passes and the malformation features are delineated, the surgeon can repair the defect.

Pathology

Hypertrophied anal papillae or fibroepithelial polyps may have the clinical appearance of hemorrhoids and are often submitted to the pathologist with this designation.²⁹ Unlike hemorrhoids, fibroepithelial polyps usually do not contain microscopic evidence of dilated or thick-walled vessels, recent or remote hemorrhage, or organizing thrombi. The mucosa covering the polyp is typically squamous, and the submucosal tissue is composed of the loose fibrovascular connective tissue characteristic of this region (Fig. 32.4). Some polyps may contain large, multinucleated, or stellate CD34-positive stromal cells or hyalinization of stromal vessels, likely caused by a reactive process of the stroma.^31,32

Some pathologists prefer to reserve the diagnosis of hypertrophied anal papillae for cases in which the stroma is composed of loose fibrovascular tissue, and they use the term anal fibroepithelial polyp when the stroma shows predominantly fibrous changes.²⁸ In essence, hypertrophied anal papillae and anal fibroepithelial polyps are identical to fibroepithelial polyps (i.e., acrochordons) of the cutaneous skin.

The differential diagnosis of a polypoid mass in the anal canal includes hemorrhoids, infection, abscess, and more serious disorders such as anal carcinoma or anal melanoma. Routine histologic examination is usually adequate to determine the underlying pathologic process. Occasionally, the squamous epithelium of a fibroepithelial polyp contains unsuspected intraepithelial neoplasia.

Prognosis and Therapy

Hypertrophic anal papillae tend to enlarge with time and may convert from an asymptomatic to a symptomatic mass associated with pruritus, anal discharge, and discomfort. Surgical resection in symptomatic cases often results in relief of symptoms. When anal papillae accompany an underlying chronic process, treatment often attempts to correct the primary cause and remove the hypertrophied papillae.²⁸

Pathology

ICPs are located in the anterior anal canal, may be single or multiple, and are typically sessile (see Chapter 22). The gross size varies, but most are between 1 and 2 cm in the greatest dimension. The histologic features include fibrosis of the lamina propria, thickening of the muscularis mucosae, hyperplasia of mucosal glands (often with a villous-like configuration) leading to a serrated contour of the epithelium, and telangiectasia of surface vasculature with or without fibrin thrombi (Fig. 32.5). The muscularis mucosae is typically thickened and irregular, with frequent extension of fibromuscular strands into the lamina propria that results in the formation of diamond-shaped crypts and deposition of mucosal elastin. The finding of elastin is distinctive because it is otherwise not seen in the normal rectum. The surface epithelium is characteristically composed of a mixture of colorectal, transitional, and squamous mucosa. Ischemic-type erosion of the surface epithelium is a common finding, and it may contribute to regenerative or hyperplastic (serrated) epithelial changes.

Differential Diagnosis

A variety of entities enter into the differential diagnosis of an ICP.³⁸ Because of their location in the anal canal and low-power appearance, ICPs may resemble tubulovillous adenomas or traditional serrated adenomas of the distal rectum. Recognition of the regenerative (non-neoplastic) epithelial cytology and eroded surface in ICPs and of the absence of cytologic dysplasia helps to distinguish an ICP from an adenoma. ICPs typically contain transitional and squamous epithelium. The fact that adenomas are somewhat uncommon in the age group in which ICPs are often diagnosed is also useful.

In some ICPs, prominent misplacement of the reactive epithelium into the underlying fibrous stroma combined with hyperplastic musculature, also known as proctitis cystica profunda, may simulate an invasive mucinous carcinoma (see Chapter 22). Although prolapse changes may also be found in invasive carcinomas, they are more prominent in ICPs, and the misplaced epithelium is usually accompanied by lamina propria tissue. Non-neoplastic epithelium surrounding mucin pools may be found in cases of benign proctitis cystica profunda, whereas neoplastic epithelium is often found floating within mucin pools in cases of invasive mucinous adenocarcinoma. In some cases, features of prolapse may be the only finding in the mucosa overlying a malignancy, and an underlying malignancy should always be considered when biopsy findings of a mass lesion reveal prolapse changes but no neoplasm.³⁹

The finding of inflammation and a granulation tissue cap may cause diagnostic confusion with a juvenile or hamartomatous polyp or an inflammatory polyp (i.e., inflammatory bowel disease-related or sporadic).³⁸ This distinction may be particularly difficult because any type of colorectal polyp can undergo secondary prolapse. A low-power view of the polyp often helps the pathologist to differentiate prolapse from an injury resulting from a preexisting (often neoplastic) lesion. ICPs show muscularization in the base of the polyp, whereas juvenile or inflammatory polyps never have this feature. Inflammatory polyps in inflammatory bowel disease usually occur in the context of chronic colitis in adjacent mucosa.

Simple excision of ICPs is usually curative.³⁶ Recurrences are uncommon.

Pathology

Different pathogens infect different types of mucosa. HSV and T. pallidum infect the stratified squamous epithelium of the perianal area and anal verge. As in other sites, HSV infections are associated with small vesicles that ulcerate and are seen histologically as mucosal ulcerations with associated acute and chronic inflammation and granulation tissue. Virally infected multinucleated cells with smudged chromatin may be seen in routine sections stained with hematoxylin and eosin, although immunohistochemical stains are of great value in confirming the presence of virally infected cells (Fig. 32.7, A and B). In contrast, syphilis is commonly associated with a chronic inflammatory infiltrate rich in plasma cells and may show associated granulomatous inflammation (see Fig. 32.7, C to E).

Dark-field microscopy of exudates from anorectal ulcers may be inaccurate because of contamination from commensal spirochetes found in the normal flora of the colorectum; however, T. pallidum immunohistochemical stains have become a useful tool for surgical pathologists.⁵⁸ Demonstration of antibodies in the serum confirms the syphilis diagnosis. Serologic tests for syphilis include the use of nonspecific cardiolipin antigens (i.e., Venereal Disease Research Laboratory [VDRL] slide test and rapid plasma reagin [RPR] test). These tests can be used to diagnose primary and secondary syphilis and to monitor treatment response. The specific treponemal antigen tests (i.e., enzyme immunoassay, T. pallidum hemagglutination assay, T. pallidum particle agglutination assay, and fluorescent treponemal antibody absorption test) remain positive even after successful treatment and can detect latent syphilis in untreated adults.

Infections occurring between the anal verge and the dentate line tend to be extremely painful because of the abundance of sensory nerve endings in this area. Chlamydial infections and gonorrhea target the columnar epithelium of the rectum. The pathologist may see evidence of cryptitis, crypt abscesses, and reactive epithelial changes in affected individuals (see Fig. 32.7, F). Granulomas may occur. Polymerase chain reaction amplification for C. trachomatis DNA is the standard diagnostic test for anorectal chlamydial infection. Because the rectum has few sensory nerve endings, infections that spare the anus may be painless.⁵⁵

Differential Diagnosis

A solitary anal ulcer (e.g., in HSV infection) may be misdiagnosed clinically as a chronic anal fissure. Granulomatous inflammation can lead to the formation of a rectal mass in primary and secondary syphilis, and it must be differentiated from other causes of a rectal mass, such as a neoplasm. Condylomata lata occurring in the perianal region in syphilis patients appear as moist, wartlike lesions, and they may be confused with human papillomavirus (HPV) infection. The pain, tenesmus, bloody discharge, and constipation of untreated LGV may mimic Crohn’s disease. In the tertiary stage, anorectal fistulas and strictures may occur in LGV.

Clinical Features and Pathogenesis

In ulcerative colitis, involvement of the anus is typically nonspecific and indistinguishable from the appearance in patients without colitis. In contrast to ulcerative colitis, at least 50% of patients with Crohn’s disease have perianal or anal disease.⁵⁹ The anal canal is involved in approximately 25% of patients with small intestinal Crohn’s disease, more than 40% of patients with large bowel disease and rectal sparing, and more than 90% of patients with large bowel disease that includes the rectum.⁶⁰ In some patients, anal involvement is the first or only clinical manifestation of Crohn’s disease.⁶¹

The clinical features of Crohn’s disease involving the anus have been well characterized.^61,62 Symptoms include recurrent sepsis, local pain, discharge, ulceration, fecal incontinence, sleep disruption, psychological disturbance, and sexual dysfunction.⁶³ Signs most characteristic of Crohn’s disease affecting the anus are chronic inflammatory changes, induration of the anal skin, multiplicity of lesions, and skin discoloration. Anal canal lesions include fissures, ulcers, strictures, perianal fistulas that open into or near the anal canal, abscesses, rectovaginal fistulas in women, and cancer.⁵⁹ Anal skin lesions are usually perianal skin tags or hemorrhoids. Anal skin tags are the most common finding and tend to be larger than those from patients without Crohn’s disease. Fissures and fistulas are the next most common finding.⁵⁹ Fissures in Crohn’s disease tend to be large, deep, and located more atypically compared with traumatic anal fissures. Most fistulas in anal Crohn’s disease are classified as simple (see Prognosis and Treatment), although in some patients, they may contain multiple openings, sometimes occurring at a considerable distance from the anal canal (i.e., watering-can perineum).⁶¹ The cause of Crohn’s perianal fistulas is uncertain, but in some patients, it is thought to be a fistula-in-ano arising from inflamed or infected anal glands or penetration of fissures or ulcers in the rectum or anal canal.⁶¹

Differential Diagnosis

The differential diagnosis of granulomatous inflammation in the anal canal includes foreign body reaction to nonspecific fistulas, sarcoidosis, tuberculosis, granuloma inguinale caused by Calymmatobacterium granulomatis, and LGV caused by sexual transmission of C. trachomatis.^55,65–67 Tuberculous granulomas are typically caseating. Stains for acid-fast bacilli and other microbiologic studies are helpful. Patients with tuberculosis of the anal canal invariably have pulmonary disease. The finding of Donovan bodies with a Warthin-Starry stain supports a diagnosis of granuloma inguinale, whereas the finding of follicular lymphohistiocytic and plasma cell infiltrates in association with neural hyperplasia may suggest chlamydial infection.⁶⁸ Hidradenitis suppurativa may be confused with complex fistulizing perianal Crohn’s disease. A correct diagnosis can often be made by examination under anesthesia and by probing the cutaneous tracts and demonstrating that they traverse only to the subdermal space and do not communicate with the rectum or other parts of the digestive tract.⁶⁹

Prognosis

The prognosis for patients with anal Crohn’s disease is related to the extent of involvement, and for most patients, it is excellent.⁷⁰ Various classification schemes for perianal Crohn’s disease have been proposed, although most are not widely used because they have not shown a reproducible relation to meaningful clinical outcomes, such as fistula formation.⁶¹ Most clinicians classify perianal disease as simple or complex.

Proper classification involves thorough evaluation of the perianal area for involvement by skin tags, fistulas, fissures, abscesses, and stricturing disease and an endoscopic examination of the rectum to assess for inflammation. Fistulas may be classified as simple, indicating that the fistula tract arises below the external sphincter, has a single external opening, has no pain or fluctuation to suggest perianal abscess, is not a rectovaginal fistula, and has no evidence of anorectal stricture. Complex fistulas arise above the external anal sphincter and may contain one or more of the features described. Rectal Crohn’s disease may make a simple fistula more complicated to manage, but overall, patients with Crohn’s disease and simple fistulas have higher rates of healing than non-Crohn’s patients with complex fistulas.⁷¹

Treatment

Management of perianal Crohn’s disease includes medical and surgical treatment. It most often begins with conservative medical treatment that includes antibiotics, steroids, antimetabolites such as azathioprine or 6-mercaptopurine, anti–tumor necrosis factor-α agents, tacrolimus, and cyclosporine. These agents are used to reduce the number of draining fistulas and are most effective in simple perianal disease.^61,71

Combined with small bowel disease, perianal Crohn’s disease is an absolute contraindication to restorative proctocolectomy (i.e., ileal pouch anal anastomosis) in the otherwise well-motivated subset of Crohn’s patients seeking a continence-preserving procedure.⁷² Surgical management is reserved for patients with complex fistulizing disease. As many as 50% of patients require surgical management of their perianal disease, ranging from local procedures such as sepsis drainage and seton placement to segmental resections such as proctectomy with ostomy creation.^61,71

Pathology

Grossly, granular cell tumors appear as firm, solitary subcutaneous nodules that are white or yellow on cut section. Histologically, these lesions are composed of sheets of cells with small, uniform, hyperchromatic nuclei and with characteristic eosinophilic granular cytoplasm that is diastase resistant on periodic acid–Schiff (PAS) staining. The granularity of the cytoplasm corresponds to vacuoles that contain cellular debris, which is evident by electron microscopy (Fig. 32.10).

The distinctive histologic features of a granular cell tumor can distinguish it from other subcutaneous perianal masses. A positive immunostain for S100 protein may be used to support the diagnosis of a granular cell tumor. These tumors are also strongly positive for CD68 (KP1). Of particular significance is the occurrence of pseudoepitheliomatous hyperplasia of the epithelium overlying granular cell tumors in as many as 50% of cases, which may be mistaken for an infiltrating squamous carcinoma.⁸⁰ S100 protein–positive granular cells located immediately beneath the atypical epithelium should help to indicate its benign nature. Local excision is usually curative.

Risk Factors and Pathogenesis

Risk factors for squamous tumors and neoplasms of the anal canal include anal-receptive intercourse, heavy smoking, a history of sexually transmitted diseases, HIV-positive status, and immunosuppression.^82–86 In one study, 92% of HIV-positive MSM demonstrated evidence of anal HPV infection, compared with 66% of MSM who were HIV negative. In women, lower genital tract squamous neoplasia is a risk factor, and numerous similarities to the incidence and epidemiology of cervical and vulvar neoplasia have been observed.⁸⁷ Women with multifocal genital intraepithelial neoplasia have a 16-fold increase in the rate of anal canal intraepithelial neoplasia.⁸⁶ Solid organ transplant recipients have a 10- to 100-fold increased risk of anogenital neoplasia.

There are more than 100 types of HPV. Of the approximately 30 types that infect the anogenital tract, the most common are 6, 11, 16, and 18.⁸⁸ The finding of HPV in an anal carcinoma is related to the sensitivity of the technique used,⁸⁹ but some evidence suggests geographic or population differences in HPV genotypes associated with anal cancers. The prevalence of HPV-16–associated anal squamous cell carcinoma (SCC) was found to be significantly lower in India and South Africa compared with more developed countries.⁹⁰

Although HPV infection is identified in more than 80% of anal SCC,⁹¹ it is not sufficient to cause malignancy by itself. Evidence for this claim is largely related to two phenomena. First, HPV infection is highly prevalent among sexually active individuals, but cancer does not develop in most patients infected with high-risk genotypes (discussed later).^92,93 In most patients, cell-mediated immunity is responsible for clearance of the infection, and neutralizing antibodies prevent subsequent infections.⁹⁴ Second, in experimental studies, infection with high-risk (oncogenic) HPV genotypes is insufficient to induce transformation and tumor progression^95,96 This may reflect the inability of HPV to integrate into the host genome. For example, in benign lesions, the viral genome replicates as an extrachromosomal episome, whereas in malignant tumors, the viral genome becomes integrated into the host cell chromosomes.^97–100 Viral integration does not appear to occur randomly. In a systematic review of known HPV integration sites, integration at 8q24, the location of the MYC oncogene, and 3p14, the location of the FHIT tumor suppressor gene, were the most common sites reported.¹⁰¹

High-risk HPVs, most commonly genotypes 16 and 18, but also 31, 33, 35, 39, 45, 50, 51, 53, 56, 58, 59, 68, and others, encode for at least three oncoproteins, known as E5, E6, and E7, which are growth stimulating and have transforming properties. With integration of HPV DNA into the host genome, E6 and E7 expression is increased.¹⁰² The E6 protein binds to the TP53 tumor suppressor protein, which is encoded by the TP53 gene and is a critical regulator of cell growth and response to stress^103,104 E6 binding to TP53 leads to its rapid degradation.¹⁰⁴ Similarly, E7 protein binds to and inactivates the retinoblastoma-associated protein (RB1) tumor suppressor protein, which normally restricts cell proliferation to the basal layer.^105–107 In most human cancers with TP53 or RB1 mutations, the proteins are inactivated. However, wild-type TP53 and RB1 genes are common in anogenital cancers.^108–111 By inactivating TP53 and RB1 and deregulating cellular growth, the action of E6 and E7 on these tumor suppressor proteins appears functionally equivalent to genetic mutations and increases the risk of anogenital cancer. Consistent with this notion, HPV-negative cervical and anal cancers have been found to harbor inactivating mutations in TP53 and RB1.^111,112

HPV integration promotes general changes of the host genome, such as chromosomal instability. The degree of chromosomal instability is highly correlated with levels of E7 overexpression, although whether the instability precedes or follows viral integration is unclear.^113–115 Recurrent chromosomal alterations associated with chromosomal instability have been described in anal squamous carcinoma that include deletion of chromosome arms 11q, 3p, 4p, 13q, and 18q and nonrandom copy number increases of chromosomes 17 and 19.^116,117 In contrast, chromosomal instability is relatively uncommon in anal carcinomas of HIV-positive patients, suggesting that immunosuppression may promote carcinogenesis through an alternate pathway.^118,119

There is no known association between any HPV subtype and specific tumor morphology. However, HPV-6 and HPV-11 are most likely to be associated with condyloma acuminatum, and HPV-16 and HPV-18 are found in a significant minority of conventional condylomata and in lesions with high-grade dysplasia^89,120–122 Verrucous carcinoma is most commonly associated with infection by HPV subtypes 6 and 11, although occasionally HPV-16 and HPV-18 have been found in these lesions.^123–125 Similar to the documented progression of HPV-associated premalignant conditions of the uterine cervix, epidemiologic studies indicate that the development of anal canal intraepithelial neoplasia is associated with infection by HPV-16 and HPV-18 and less commonly by HPV-6 and HPV-11.^126,127

Evidence in support of the precursor potential of anal squamous intraepithelial neoplasia (ASIN) lies in its close similarity to dysplasia of the uterine cervix, including histologic similarities between the preinvasive and invasive lesions, its occurrence at a younger average patient age compared with its invasive counterpart, and its common occurrence adjacent to invasive cancers, particularly those associated with HPV infection in MSM.^84,86 Perianal intraepithelial neoplasia, including bowenoid papulosis and Bowen disease, also have an association with HPV infection; in particular, HPV-16 has been demonstrated in bowenoid papulosis and Bowen’s disease.^128,129

HPV vaccines have been developed to protect recipients against HPV types 16 and 18 (HPV2/Cervarix, GlaxoSmithKline Biologicals, Rixensart, Belgium) and HPV types 6, 11, 16, and 18 (HPV4/Gardasil, Merck, West Point, PA). These vaccines were first approved and recommended for administration to girls in 2009 and 2006, respectively; in late 2011, HPV4/Gardasil was recommended for administration to boys.¹³⁰ One analysis found that vaccinating boys and girls, rather than girls alone, could reduce the HPV-related cancer burden in boys by 65% compared with girls-only vaccination. The investigators further estimated that inclusion of boys in a HPV vaccination program would lead to an 86% reduction in the incidence of anal cancer compared with a 63% reduction with girls-only vaccination.¹³¹

Nomenclature of Preinvasive Anal Squamous Neoplasia

During the past several decades, various terms have been used to describe the same anal squamous pathology. Consensus groups such as the AJCC, the Bethesda System, and the World Health Organization (WHO) have used different nomenclature. This chapter uses the WHO nomenclature, but a brief discussion of the latest attempt at unifying squamous neoplasia nomenclature is warranted because it may become standard in the future (Table 32.2).

In 2012, the College of American Pathologists (CAP) and the American Society for Colposcopy and Cervical Pathology released the product of a joint Lower Anogenital Squamous Terminology (LAST) Project, in which the groups reported their recommendations for a unified terminology for all HPV-related squamous lesions in the lower anogenital tract.¹³² They contended that the clinical, epidemiologic, and biologic similarities between the various anogenital sites justified the use of a single nomenclature for HPV-associated lesions. When presented with a lesion, a pathologist cannot reliably predict from which site it was derived (e.g., cervix, vagina, anus) or whether it came from a male or female patient. There are no differences in the molecular or biomarker phenotypes in the various HPV-related lower anogenital sites.

The CAP/LAST investigators contend that HPV interacts with all the anogenital squamous epithelium by two basic mechanisms. Squamous cells transiently support the virus, which manifests as a low-grade squamous intraepithelial lesion (LSIL). Alternatively, precancerous lesions resulting from viral oncogene expression cause a clonal proliferation of relatively undifferentiated cells in a high-grade squamous intraepithelial lesion (HSIL). With time, persistent HPV infection results in a substantial risk of malignancy. In the LAST nomenclature, additional pathologic descriptors such as ASIN, condyloma acuminatum, or bowenoid papulosis may be added in parentheses as needed.

Some investigators and clinicians are skeptical of applying the cervical HPV screening and treatment paradigm to anal pathology. They cite reasons such as unclear natural history data for anal HPV infection. Many of the studies of anal squamous neoplasia focus on MSM, which may be a significantly different population from women with cervical neoplasia. For example, HPV infection and anal intraepithelial neoplasia tend to persist in MSM as the population ages, whereas cervical HPV infection and intraepithelial neoplasia dramatically decrease with patient age.⁹¹ Some researchers report high-grade anal squamous intraepithelial neoplasia (ASIN-H) with a relatively high risk of progression to invasive carcinoma; Scholefield and colleagues estimate 10% at 5 years.¹³³ However, one meta-analysis estimates the risk to be much lower: 0.15% per year for HIV-positive MSM and 0.02% per year for HIV-negative MSM.⁹¹ No randomized trials have investigated the benefits and risks of managing ASIN.

Pathology

The perianal skin is the most commonly affected area, although condylomata located solely within the anal canal may occur. Grossly, condylomata are soft, fleshy, and tan, gray, or pink papillomatous growths that often occur in groups or clusters.

Histologically, the papillomatous appearance is best appreciated at low magnification. The squamous epithelium shows marked acanthosis and a variable expansion of the stratum corneum layer. Surface parakeratosis is common. The rete pegs may appear elongated with broad bases. On higher magnification, the surface epithelium contains squamous cells with an enlarged, irregular, and hyperchromatic nucleus and an accompanying perinuclear halo. These cells, known as koilocytes, are the histologic hallmark of HPV infection (Fig. 32.11). Dyskeratotic cells are usually easily found, as well as multinucleate cells, which are also characteristic of HPV infection. Basal cell hyperplasia is typically associated with orderly and progressive maturation of the epithelium. The squamous epithelium is well delineated from the underlying stroma, which often contains chronic inflammation, edema, and vascular dilation.

Differential Diagnosis

The differential diagnosis of condyloma includes verrucous carcinoma on one end of the spectrum and benign anal or perianal lesions, such as seborrheic keratosis, fibroepithelial polyp, and prolapse-type polyp, on the other end. Condyloma acuminatum may be impossible to distinguish from verrucous carcinoma, particularly in biopsy samples or when the clinical impression is not provided. Verrucous carcinomas are usually larger than condylomata acuminata, and verrucous carcinomas display exophytic and endophytic growth patterns, whereas condylomata usually are only exophytic. Separating condyloma from another type of benign lesion rests on identifying an HPV cytopathic effect in the condyloma or the characteristic histologic features in the other benign lesions. Immunohistochemistry or in situ hybridization studies for HPV can be helpful in equivocal cases.

Prognosis

There is controversy regarding the malignant potential of condylomata. Although they usually harbor low-risk HPV genotypes, as many as 35% demonstrate high-risk HPV, most commonly genotypes 16 and 18.¹²² High-grade intraepithelial neoplasia has been described in as many as 20% of HIV-infected patients with condylomata, and cases of invasive squamous carcinoma arising in an anal condyloma have been reported.¹²¹

Treatment

The most common treatments for anal condylomata are ablative or cytodestructive.¹³⁵ Methods of ablative treatment include cryotherapy, local excision, electrocautery, and laser therapy. Cytodestructive therapies include podophyllotoxin and trichloroacetic acid. Both forms of treatment are associated with a high recurrence rate, which is attributed to latent HPV in clinically unremarkable adjacent epithelium. Immunosuppression is also associated with an increased rate of recurrence after surgical removal.¹³⁶ The immunomodulatory agent imiquimod has potent antiviral and antitumor activity without the common side effects of the other forms of treatment.¹³⁷

Pathology

ASIN is characterized by cells with enlarged nuclei, irregular nuclear membranes, and increased nucleus-to-cytoplasm ratios, with a lack of cytoplasmic maturation toward the luminal surface.¹³² Mitotic figures may be encountered beyond the normal regenerative basal zone of the squamous epithelium. Individual cell keratinization and dyskeratosis are common and may represent markers of HPV infection. These changes typically occur in the absence of inflammation.

ASIN is divided into two grades, primarily based on the level at which orderly cytoplasmic maturation begins and the level at which mitotic activity stops. In low-grade anal squamous intraepithelial neoplasia (ASIN-L), mitotic figures and cellular pleomorphism are predominantly located in the basal third of the epithelium, with evidence of surface maturation in the upper two thirds or evidence of an HPV cytopathic effect (i.e., koilocytic atypia) is present. In contrast, ASIN-H is characterized by cellular pleomorphism, mitotic figures, and lack of cytoplasmic maturation extending into the middle and superficial thirds of the epithelium (Fig. 32.12). These preinvasive changes may colonize the underlying anal ducts and glands, which may not be detected or excised during treatment. Intraepithelial neoplasia that extends into the ducts can be graded using the same principles as surface dysplasia; however, the pathologist must be cautious with the use of tangential sectioning (see Fig. 32.12, C).

Differential Diagnosis

The differential diagnoses of ASIN include other neoplasms within the squamous epithelium such as melanoma and primary or secondary Paget disease. Some ASIN-H lesions may demonstrate a pattern of pagetoid spread that can only be differentiated from Paget disease by immunohistochemistry (see Paget Disease of the Anus).¹⁴⁴

Differentiating Reactive Changes or Immature Squamous Metaplasia from High-Grade Squamous Intraepithelial Lesions

Distinguishing reactive changes from ASIN-L in biopsy material is fraught with high observer variability, and usually rests on deciding whether or not unequivocal HPV cytopathic effect is present (Table 32.3).¹⁴⁵ Unfortunately, there are no reliable adjunctive pathologic methods to separate these two processes.

Table 32.3

Differential Diagnosis of Anal Intraepithelial Neoplasia and Separation from Reactive Lesions

Feature	Reactive Lesions	ASIN-L	ASIN-H
N : C ratio	Low	Low	High
Koilocytes	No	Yes	Often
Mitotic activity	Basal layers only	Lower one third of mucosa	Upper two thirds of mucosa (may have abnormal mitotic figures)
Maturation	Orderly	Dysmaturation in lower one third of mucosa only	Dysmaturation in upper layers to full thickness
Cell polarity	Normal polarity	Relatively well-maintained polarity	Loss of polarity
Proliferation	Basal layers	Lower one third of mucosa	Throughout mucosa
Staining for p16	Negative or patchy positivity	Varies	Blocklike positivity (nuclear and cytoplasmic)

ASIN-H, High-grade anal squamous intraepithelial neoplasia; ASIN-L, low-grade anal squamous intraepithelial neoplasia; N : C, nucleus-to-cytoplasm ratio; p16, p16 protein.

Separating reactive changes from ASIN-H, a precancerous lesion, is clinically important and more feasible (Fig. 32.13). Although reactive changes or immature squamous metaplasia and ASIN-H may show immature cells located beyond the basal layer, ASIN-H demonstrates a more disorganized or jumbled arrangement of the epithelium, and reactive processes tend to retain an orderly configuration of the squamous epithelial cells. Identification of inflammation or regular nuclei with open chromatin and prominent nucleoli favor a reactive process, whereas ASIN-H usually shows nuclear hyperchromasia and irregular nuclear contours.

When this differential is encountered, p16 immunohistochemistry often becomes necessary to separate reactive processes from ASIN-H. Weak and patchy immunoreactivity favors a non-neoplastic process, whereas diffuse, strong, and block positivity argues in favor of ASIN-H (Table 32.4).¹³² The latest recommendations argue against the use of multiple adjunctive immunohistochemical tests (e.g., p16 and Ki67) in the workup of a possible ASIN-H case.¹³² Evidence supports p16 as the best marker for ASIN-H, but Ki67^146–148 and ProEx C (Becton Dickinson, Franklin Lakes, NJ)^149–151 may also be useful when difficulties with p16 arise.

Table 32.4

Use of P16 Immunohistochemistry in the Diagnosis of Anal Squamous Intraepithelial Neoplasia

Diagnostic Problem	P16 Recommended?	Interpretation or Reasoning
Reactive versus ASIN-H	Yes	Block positive staining supports ASIN-H diagnosis
Diagnosis of ASIN-L versus ASIN-H (equivocal for ASIN-H)	Yes	Block positive staining supports ASIN-H diagnosis
Professional disagreement about histologic diagnosis (provided ASIN-H is in differential diagnosis)	Yes	Block positive staining supports ASIN-H diagnosis
Diagnosis of negative or reactive on histology	No	Staining may be confusing
Diagnosis of ASIN-L on histology	No	Staining may be confusing
Diagnosis of ASIN-H on histology	No	Staining may be confusing

ASIN-H, High-grade anal squamous intraepithelial neoplasia; ASIN-L, low-grade anal squamous intraepithelial neoplasia; P16, p16 protein.

From Darragh TM, Colgan TJ, Cox JT, et al. The lower anogenital squamous terminology standardization project for HPV-associated lesions: background and consensus recommendations from the College of American pathologists and the American Society for Colposcopy and Cervical Pathology. Arch Pathol Lab Med. 2012;136:1266-1297.

Distinguishing Low-Grade and High-Grade Squamous Intraepithelial Lesions

In immunocompetent hosts, low-grade squamous lesions usually are transient and not precancerous, although this is often not the case in immunocompromised patients. Aside from the traditional morphologic features that separate ASIN-L from ASIN-H, the CAP and Lower Anogenital Squamous Terminology Project investigators identify at least two special circumstances for “bumping up” a lesion with the cytoarchitectural features of ASIN-L to ASIN-H: atypical mitotic figures or significant nuclear atypia beyond that expected in typical ASIN-L and thin ASINs, which are immature intraepithelial lesions less than 10 cells thick. As long as there are mitotic figures or cytologic features of ASIN above the basal layer, thin ASIN lesions should be considered ASIN-H. In both circumstances, a diffuse, strong, and blocklike positivity for p16 immunohistochemistry helps to confirm ASIN-H (Fig. 32.14).¹³²

Differentiating High-Grade Squamous Intraepithelial Lesions from Superficially Invasive Squamous Cell Carcinoma

Early invasion by SCC can be difficult to diagnose with certainty, particularly in the setting of a condyloma, because condylomata often have an undulating interface with the underlying stroma (Fig. 32.15). ASIN may colonize the anal ducts and glands, mimicking stromal invasion. It is best to maintain strict criteria for stromal invasion by a squamous carcinoma. Invasive carcinomas tend to arise in mucosa extensively involved by high-grade intraepithelial neoplasia. Several features support the diagnosis of invasion:

• Irregular and angulated downward extension of the neoplastic squamous cells with an accompanying reactive or desmoplastic stromal response

• Small nests or individual cells that often acquire more abundant eosinophilic cytoplasm (“paradoxical maturation”) with prominent nucleoli that are clearly separate from the main lesion

• Blurring of the epithelial-stromal interface with loss of the regular arrangement of the squamous basal cells¹⁵²

The pathologist must be careful not to misinterpret artifacts or reactive changes such as those accompanying prior biopsy, intense stromal inflammation, or pseudoepitheliomatous hyperplasia as evidence of invasive carcinoma.

Natural History and Treatment

Progression of ASIN largely depends on the immune status of the patient and extent of disease.¹⁵³ Small series of immunosuppressed patients have found that as many as two thirds of cases progress from ASIN-L to ASIN-H within 2 years.¹⁴¹ After ASIN has developed, the rate of progression to carcinoma is uncertain, but some estimate that as many as 5% to 10% of patients will ultimately develop invasive disease.¹⁴³ In one long-term surveillance study of 35 patients with ASIN-H, invasive carcinoma developed in 3 of 6 immunosuppressed patients (all with multifocal disease) in the follow-up period, but cancer did not develop in any of the other 3 patients.¹⁵³ However, compared with some small observational studies, a meta-analysis found much lower progression rates; as few as 0.02% of cases progressed to invasive cancer per year.⁹¹

There are no well-established guidelines regarding appropriate screening for ASIN, even in at-risk populations, and groups such as the American Society of Colon and Rectal Surgeons¹⁵⁴ and the Association of Coloproctology of Great Britain and Ireland¹³³ have opposite opinions regarding the benefit of screening for ASIN. Management guidelines range from watchful waiting to topical immunomodulatory or ablative therapies to surgical excision. Management of ASIN should probably be individually tailored, taking into account the patient’s symptoms and immune status and the size and focality of disease.¹³³

When treatment is determined, the goal is to eradicate ASIN, prevent the development of anal carcinoma, and maintain anal function. The success of surgical resection is largely related to the extent of involvement and the ability to achieve a disease-free margin, although HIV status also seems to be an important factor.¹⁴³ Unfortunately, many of these treatment modalities have been associated with high rates of recurrence, particularly in HIV-positive patients and those with extensive or multifocal disease.¹⁵⁴

Pathology

The WHO recognizes several types of malignancies of the anal canal¹⁶⁶ (Table 32.7). Each carcinoma is associated with several histologic patterns, which is probably a reflection of the histologic complexity of the anal canal. The locations of the most common tumors of the anal canal are shown in Figure 32.18.

Most malignant neoplasms of the anal canal are SCCs, which are often histologically heterogeneous.¹⁶⁷ The second edition of the WHO classification¹⁶⁸ subclassified anal canal SCC as large cell keratinizing or nonkeratinizing and basaloid types; however, the prior and current WHO editions eliminated this classification because anal canal invasive SCCs are most often treated with combination chemotherapy and irradiation after diagnostic biopsy rather than primary resection. Because anal canal SCCs are often heterogenous, biopsies cannot sufficiently subclassify them. The WHO 1989 classification system did not yield consistent prognostic groups based on histologic type. The WHO 2010 system¹⁶⁶ encourages adding additional descriptors, such as degree of differentiation, predominant cell size, basaloid features, degree of keratinization, or adjacent ASIN, to the generic diagnosis of invasive SCC. Terms such as basaloid, transitional cell, cloacogenic, or mucoepidermoid carcinoma are discouraged.

SCCs located above the dentate line are thought to arise from the anal transitional epithelium. Most tumors are nonkeratinizing SCCs that are similar in appearance to nonkeratinizing squamous cell cancers in other anatomic locations. Basaloid features may predominate (Fig. 32.19). Microscopically, basaloid zones of anal canal SCC frequently show an irregular, angulated, nested, or trabecular pattern composed of relatively small cells without intercellular bridges. Central necrosis is a common histologic feature, and mitotic figures are frequently observed. Some basaloid zones also contain small, cystic foci lined by mucin-producing cells, formerly called mucoepidermoid carcinoma or squamous cell carcinoma with mucinous microcysts. Some pathologists require prominent peripheral palisading within the tumor cell nests, similar to cutaneous basal cell carcinomas, to consider a tumor to have basaloid features. Not uncommonly, a single tumor shows a mixture of histologic patterns, including foci of squamous differentiation.

Tumors in this region may extend proximally or distally, which often obscures their precise site of origin. Examination of the adjacent mucosa may identify ASIN. SCCs with or without basaloid features may occasionally arise from anal duct epithelium, which reflects the common embryologic origin with anal transitional epithelium. In some instances, ASIN may involve the surface mucosa and anal duct epithelium.¹⁶⁹ Carcinomas that arise below the dentate line are typically SCCs and tend to be better differentiated and show more keratinizing features than those that arise above the dentate line.¹⁷⁰ It is often difficult to grossly or endoscopically ascertain the precise site of origin of any neoplasm, because carcinomas often destroy the normal anatomic landmarks (see Anatomy). Preoperative irradiation and chemotherapy of anal canal carcinomas may also affect the ability to determine the correct site of origin of the neoplasm.¹²⁶

Differential Diagnosis

SCCs of the anal canal should be differentiated from other primary and metastatic malignancies, including low rectal adenocarcinoma, anal gland carcinoma, melanoma, and small cell neuroendocrine carcinoma. The diagnosis is greatly aided by judicious use of immunohistochemical stains, particularly for poorly differentiated malignant neoplasms.

Prognosis

Staging of anal carcinomas should be performed in accordance with the AJCC criteria (Tables 32.8 and 32.9).⁴ The most important prognostic indicators of anal carcinoma are depth of invasion and extent of tumor spread. Carcinomas located above the dentate line commonly spread to the lower rectum and involve perirectal and inguinal lymph nodes. Aside from the rare small cell anaplastic carcinoma and carcinoma with microcyst formation, the specific histologic type of squamous carcinoma from this region is not a significant predictor of survival, although stage for stage, nonsquamous carcinomas tend to have a 10% to 20% worse 5-year survival rate compared with anal canal SCC. The degree of differentiation correlates with the presence of lymph node metastases.¹⁷¹ Worse survival has been observed among patients with node-positive compared with node-negative tumors.^172,173

Carcinomas arising in the perianal skin beyond the junction of anal mucosa and hair-bearing skin are staged by using skin cancer criteria.^4,174,175 Carcinomas of the perianal region often grow more slowly than proximally located tumors, and because they are more frequently diagnosed at an early stage, the prognosis tends to be considerably better. Small perianal carcinomas are usually treated with surgical excision and wide margins (≥1 cm) rather than chemoradiotherapy, but larger tumors may require chemoradiation therapy. The outcome of perianal skin SCC also depends on the tumor size and locoregional spread. Metastases to inguinal lymph nodes may occur, but visceral metastases are uncommon.

As reviewed by Eng,¹⁷⁶ features that have an impact on local control of anal canal carcinoma include advanced tumor stage, older age at diagnosis, lower radiation dose, and reduced overall treatment time. In HIV-positive patients, CD4+ T-cell counts of less than 200/mm³ can be a negative prognostic indicator of local control, although possible differences in the molecular biology of HIV-positive anal cancers may affect outcome.^118,177,178 Highly active antiretroviral therapy (HAART) has revolutionized HIV therapy, but its benefit in anal HPV infection and intraepithelial neoplasia is still debated. Although some larger studies^179–181 have found a beneficial effect for HAART on the prevalence of HPV infection or ASIN, or both, others have not.^182,183 Studies suggest that HAART does not impact the high incidence of anal carcinoma among HIV-positive patients.^184,185 Five-year survival rates are 70% for stage I disease (71.4% for SCC only), 62% for stage II (63.5% for SCC only), 46% for stage IIIA (48% for SCC only), 40% for stage IIIB (43% for SCC only), and 15% for stage IV (21% for SCC only).⁴

Treatment

Before the 1970s, patients with anal cancer were treated primarily by radical surgery, resulting in an abdominoperineal resection and a permanent colostomy and producing a 5-year survival rate of 40% to 70%. Current protocols for patients with anal cancer include combined chemoradiation therapy. The National Comprehensive Cancer Network (NCCN) guidelines for anal canal cancer recommend combination mitomycin C and 5-flourouracil chemotherapy and multifield radiotherapy as primary therapy for invasive anal canal SCC.¹⁷⁴ With this approach, 70% to 90% of patients have a complete clinical response, with no tumor found on subsequent biopsies after 6 weeks.¹⁸⁶ One trial found that 29% of patients who did not demonstrate an early complete response (by 11 weeks) had a complete response by week 26, arguing for conservative management within 6 months of completing chemoradiation.¹⁸⁷ This remarkable improvement in local control with primary chemoradiotherapy allows preservation of the anal sphincter and an improved quality of life.

The risk of local failure after treatment is greatest within the first 3 years after treatment.^{172,188–190} Residual disease is defined as persistent disease within 6 months of completion of chemoradiation therapy, and recurrent disease is defined as carcinoma diagnosed 6 or more months after a complete clinical response. Because additional medical therapy is unlikely to provide a cure in this setting, the preferred method of treatment for locally residual or recurrent disease is an abdominoperineal resection.^188–190 As many as 70% of patients are cured after salvage surgery. Positive surgical margins are the best predictor of a bad outcome.¹⁹⁰

Verrucous Carcinoma and Giant Condyloma Acuminatum of Buschke-Löwenstein

Once thought of as separate entities, verrucous carcinoma and giant condyloma acuminatum (Buschke-Löwenstein tumor) are now considered synonymous. These lesions are characterized by bulky exophytic growth and tumor size that often exceeds 10 cm in the greatest dimension (Fig. 32.20). Despite their inability to metastasize, giant condylomata are locally invasive and therefore considered to be malignant tumors.^191,192 They are more common in men, and the incidence is even greater for men younger than 50 years. Rarely, they may occur in children.¹⁹³ The clinical presentation includes a perianal mass and pain, abscess, or fistula. Because of their large size, these tumors have a propensity to ulcerate, and bleeding is a common complaint.

Their histologic appearance is paradoxically benign. These tumors appear as very-well-differentiated squa­moproliferative lesions on microscopic examination (Fig. 32.21; see Fig. 32.20). They contain minimal cytologic atypia but show a broad invasive front. They are typically locally invasive into adjacent chronically inflamed stromal tissue, and examination reveals a pushing rather than an infiltrating tumor margin. It is this downward, destructive growth pattern that best separates verrucous carcinoma from a typical condyloma. The epithelium often shows evidence of surface maturation and extensive crater-like keratinization.

The various treatment strategies include topical chemotherapy, wide local excision, abdominoperineal resection, and adjuvant and neoadjuvant systemic chemotherapy and radiation therapy.^194,195 Recurrences are common after surgical resection. As many as 20% of patients die of their disease.¹⁹¹ Because of their slow growth, local excision is the preferred method of management, although for large and bulky tumors that are locally destructive, chemoradiation therapy, cryotherapy, and more extensive surgical management may be required.¹⁹⁶

Clinical Features and Pathogenesis

Primary adenocarcinomas of the anal canal are uncommon.¹⁹⁷ Most adenocarcinomas in this region result from secondary involvement by a distal rectal tumor. Primary adenocarcinomas are commonly diagnosed in the seventh decade, and men may be more frequently affected than women.^8,164 The clinical presentation typically is a painful mass of the buttock accompanied by a mucin-like discharge. Anorectal bleeding or obstruction is not common.

Primary adenocarcinomas of the anal canal (not the rectum) are categorized as those that arise in preexisting anorectal sinuses or fistulas and those that presumably arise from anal glands (without fistulas). They are often slow-growing tumors. Patients with Crohn’s disease are at higher risk for fistula-associated tumors.¹⁹⁸ The causal relationship between perianal fistulas and abscesses and the subsequent development of carcinoma is a subject of debate. Some authorities have proposed that the inflammatory process precedes the neoplasm, whereas others think that these rather slow-growing neoplasms undergo secondary fistulization with time. A possible association of anal adenocarcinoma with HPV infection has been described, similar to that observed for endocervical adenocarcinomas.¹⁶⁹

Pathology

Primary adenocarcinoma of the anal canal characteristically arises in the deep perianal tissues, without evidence of surface mucosal involvement. On cut section, the tumor appears as an infiltrative mass with an epicenter located deep within the soft tissues and with involvement of the adjacent structures of the anal canal. Most primary anal gland adenocarcinomas are well differentiated, comprising haphazardly dispersed, small glands with scant mucin production (Fig. 32.22). These carcinomas typically show an anal gland immunophenotype; they are CK7 positive and CK20 negative.⁸

Carcinomas arising within fistulas are usually associated with abundant mucin production (see Fig. 32.22), and they are sometimes referred to as perianal mucinous (colloid) adenocarcinoma.^8,199 Perianal mucinous adenocarcinomas are more commonly associated with fistulas and may more often be associated with a rectal phenotype (i.e., CK20 and CDX2 positive).²⁰⁰ Multiple, deep biopsies are often required to establish a diagnosis of colloid carcinoma, because the abundance of mucin may make identification of tumor cells difficult.^201,202

Differential Diagnosis

The main differential diagnosis of a primary anal adenocarcinoma is a distal rectal cancer that extends into the anal canal (Table 32.10). Demonstration of a precursor adenomatous component in rectal mucosa helps to demonstrate its origin in rectal columnar epithelium. Immunohistochemical labeling for CK7 and CK20 may also be useful, because primary anal gland adenocarcinomas are universally CK7 positive, whereas most rectal tumors are CK7 negative and CK20 positive. Because 10% to 20% of rectal carcinomas are CK7 positive,^5,203,204 CK7 positivity does not exclude the possibility of a rectal primary. CDX2, a marker for colorectal adenocarcinoma, is extremely useful for differentiating an intestinal phenotype from an anal gland phenotype,²⁰⁵ but fistula-associated adenocarcinoma more commonly demonstrates a rectal phenotype than an anal gland phenotype.²⁰⁰ Use of MUC2 or MUC5AC immunohistochemical stains may increase diagnostic specificity.^200,206,207

Prognosis and Treatment

Primary adenocarcinoma of the anal canal is considered a more aggressive tumor than SCC in this location. Patients are at higher risk for local and distant recurrences.²⁰⁸ Tumor stage, lymph node status, and differentiation status are independently associated with outcome.

Treatment of primary adenocarcinomas of the anal canal is similar to that of rectal cancer. It usually includes surgical resection and aggressive chemoradiation. The overall 5-year survival rate is approximately 30%.^208,209

Clinical Features

Primary (extramammary) Paget disease of the anal canal is rare. Among patients with primary anal Paget disease, men and women are equally affected, and most are diagnosed in the fifth to seventh decades of life.²¹⁰

The clinical appearance of extramammary Paget disease is well described.^211,212 Typical lesions consist of erythematous patches or plaques that may be scaly, eroded, or ulcerated, and they may be located anywhere between the dentate line and the perianal skin.²¹⁰ Pruritus is a common complaint. Lesion size varies greatly, and regions of involvement by neoplastic cells may extend well beyond the impression of involvement.

Secondary anal or perianal Paget disease may occur in association with a distal adenocarcinoma of the rectum, which may be diagnosed before the anal Paget disease, at the same time, or later.^211–214 Approximately one half of cases of Paget disease of the anus are associated with a visceral malignancy, such as rectal adenocarcinoma. Perianal Paget disease may also develop as an extension of perineal Paget disease, particularly in postmenopausal white women.²¹⁵ In this instance, the finding of Paget cells likely represents migration of malignant cells into the anal epidermis from an associated rectal adenocarcinoma or extramammary Paget disease of the perineum.²¹⁶ These varied patterns of occurrence of anal Paget disease suggest that it may have many causes.²¹³

Pathogenesis

In cases of primary extramammary anal Paget disease, there is general agreement that the neoplastic Paget cell is a secretory (glandular) epithelial cell. Although some authorities favor an eccrine sweat gland origin, most support an apocrine gland origin of Paget cells.^217,218 The immunohistochemical demonstration of gross cystic disease fluid protein (GCDFP) expression, which is a marker of apocrine epithelium, in many cases of extramammary Paget disease supports this viewpoint. Paget cells may also be positive for epithelial membrane antigen, carcinoembryonic antigen, androgen receptors, and low-molecular-weight keratins.

Pathology

Histologically, Paget disease is characterized by large, cytologically malignant cells with pale, granular, or vacuolated cytoplasm scattered throughout the epidermis (Fig. 32.23). The cells contain intracytoplasmic mucin, which is usually PAS and Alcian blue positive. The cutaneous appendages may also be involved. Paget cells tend to be more numerous in the basal half of the epidermis. In some cases, intraepithelial glands with intraluminal necrosis may be found.²⁰³ The involved epidermis may also show a variety of reactive changes, such as squamous hyperplasia, papillomatous hyperplasia, hyperkeratosis, parakeratosis, or acanthosis.²¹⁹

Differential Diagnosis

The differential diagnosis of anal Paget disease is pagetoid spread by a primary rectal adenocarcinoma, malignant melanoma, or squamous intraepithelial neoplasia (Table 32.11). Fortunately, a limited panel of immunohistochemical studies can help to establish a correct diagnosis in most cases (Table 32.12). A CK20-positive, CDX2-positive, and GCDFP-negative immunophenotype is typical of Paget disease associated with a coexistent rectal cancer, whereas a CK7-positive, CK20-negative, and GCDFP-positive immunophenotype is typical of primary anogenital Paget disease.²⁰³ Paget disease caused by rectal adenocarcinoma usually has a prominent signet ring cell morphology and more abundant cytoplasmic mucin. The cells derived from a rectal adenocarcinoma tend to invade the epidermis more discretely and with less predilection for the basal portions of the epithelium.

CDX2, Caudal type homeobox 2; CK, cytokeratin; GCDFP-15, 15-kDa gross cystic disease fluid protein; HMB-45, human melanoma black 45; HPV, human papillomavirus; MUC, mucin; −, negative; +, positive.

Clinical Features and Pathogenesis

Melanomas of the anus account for 0.05% of all colorectal malignancies diagnosed each year.²²⁷ Patients usually are diagnosed in the seventh decade of life. Women are affected almost twice as often as men.^227,228 There are no known risk factors for anal melanoma.

Anal melanomas are thought to arise from malignant transformation of melanocytes that are normally present in the anal mucosa and transition zone. The clinical presentation is not unlike that seen for a variety of anal lesions, even those as innocuous as hemorrhoids or anal tags, and is characterized by bleeding, pain, or a mass lesion. These often mild and nonspecific features contribute to the typical late-stage diagnosis of anal melanomas.^229,230

Molecular diagnostics are playing an increasingly important role in the characterization of all melanomas, including anorectal melanoma. There is a particular focus on treatment-determining somatic mutations. Most clinically significant tests can be performed on formalin-fixed, paraffin-embedded tissue. The two most treatment-specific mutations tested in melanoma are BRAF and KIT, and the results of testing may determine therapy with BRAF inhibitors (e.g., vemurafenib) or tyrosine kinase inhibitors (e.g., imatinib or sunitinib), respectively. Activating codon 600 BRAF (often V600E) mutations are uncommon events in mucosal melanomas, occurring in as many as 11% of lesions.^231,232 Other studies have found no BRAF V600E mutations in mucosal melanomas, including some anorectal-specific series ^233–236

KIT mutations occur in approximately 26% of mucosal melanomas²³⁷ and approximately 15% of anorectal melanomas.^236,238 Exons 9, 11, 13, and 17 are the most commonly analyzed in KIT mutation testing, and it seems that mutations in exons 11 and 13 may have the best possibility of treatment response to tyrosine kinase inhibitors.²³⁹

Pathology

Anal melanomas usually arise as a polypoid mass adjacent to or at the dentate line.²²⁷ Gross pigmentation is uncommon, but microscopic foci of pigmentation may be found in up to 80% of cases. Pigmentation may be obscured by hemorrhage in the lesion. Anal melanomas demonstrate the same immunohistochemical and ultrastructural features as their cutaneous counterparts.^220,240 However, similar to melanomas that arise in other mucous membranes, they are typically of the acrolentiginous type. Invasive melanomas in this region are commonly epithelioid, although sarcomatous and desmoplastic melanomas have also been described (Fig. 32.24).²²⁰

Differential Diagnosis

Differentiation of anal melanomas from primary anal Paget disease is discussed in the section on Paget disease. Similar to melanomas that occur elsewhere, the histologic features that support a diagnosis of anal melanoma are a nested growth pattern and asymmetric junctional changes in association with the appropriate immunohistochemical labeling pattern (i.e., positive for S100 protein, HMB-45, or melan-A, and other melanocytic markers and negative for cytokeratins).^220,240 Junctional changes, although useful when found, may be obscured by ulceration. Because melanomas may also extend proximally into the rectum, they may infiltrate colorectal mucosa in a manner analogous to lymphoma at that site. Negative immunohistochemical labeling for lymphoid markers (i.e., CD20, CD3, and CD45) should be helpful in ruling out this possibility.

Prognosis and Treatment

The prognosis for anal melanoma is poor. Reported 5-year survival rates range from 15% to 35%.^{227,228,230,241,242} Survival rates are better for younger patients (25 to 44 years old).²²⁷ The specific histologic type of anal melanoma or lymph node status does not correlate with survival, but the thickness of the tumor (measured from the top of the overlying intact mucosa or ulcerated tumor) and perineural invasion are related to outcome.^240,242,243 Anal melanomas 2 mm thick or smaller have a much better prognosis than those more than 2 mm thick.

Because there is no relationship between the type of surgical management (i.e., abdominoperineal resection versus wide local excision) and local recurrence or outcome, current practice is to perform wide local excision if technically feasible.^241,242 Given the lack of benefit of radical surgery in the absence of large or bulky tumors, disease-free survival appears to be a function of distant spread rather than local control.²⁴⁴ Although systemic therapy for melanoma is evolving based on the molecular characteristics of the tumor, the specific effects on anal melanoma are not well established.