An Update on Tumors of the Anal Canal
Jinru Shia, MD
NandContext.—The anal canal possesses complex anatomy
histology and gives rise to a variety of tumor types.
Challenging issues remain with regard to both the
pathologic diagnosis and the clinical management of these
tumors.
Objectives.—To provide an updated overview of the
histogenesis, clinical and pathologic characteristics, diag-
nostic terminology, and relevant clinical management of
the various types of anal canal tumors.
A s stated in Pathology and Genetics of Tumours of the
Digestive System: The World Health Organization
1
(WHO) Classification of Tumours, ‘‘despite its short length,
the anal canal produces a variety of tumor types reflecting
its complex anatomic and histological structure.’’ These
tumor types can be categorized as follows: (1) squamous
cell tumors including condyloma acuminatum, flat squa-
mous dysplasia, and invasive squamous cell carcinoma
and its variants; (2) adenocarcinoma including rectal type
adenocarcinoma, the so-called anal gland adenocarcinoma,
fistula-related mucinous adenocarcinoma, and intraepithe-
lial adenocarcinoma (ie, Paget disease); (3) neuroendocrine
neoplasms including carcinoid tumor, small cell carci-
noma, and non–small cell high-grade neuroendocrine
carcinoma; (4) malignant melanoma; (5) mesenchymal
tumors; and (6) malignant lymphoma. Although these anal
canal tumors as a group account for only about 1.5% of all
gastrointestinal tract neoplasms, their pathologic diagnosis
and clinical management can be challenging to both
pathologists and clinicians. This review aims at providing
an updated overview of the histogenesis, clinical and
pathologic characteristics, diagnostic terminology, and
clinical management of anal canal tumors. Attention will
also be directed to the differential diagnosis of poorly
differentiated anal canal tumors and the utility of im-
munohistochemical stains in difficult diagnostic scenarios.
NORMAL ANATOMY AND HISTOLOGY
The anal canal can be defined either surgically or
histologically. As illustrated in Figure 1, the ‘‘surgical anal
Accepted for publication December 9, 2009.
From the Department of Pathology, Memorial Sloan-Kettering Cancer
Center, New York, New York.
Presented in part at the Surgical Pathology of Neoplastic Diseases
course, Memorial Sloan-Kettering Cancer Center, New York, New York,
May 18–22, 2009.
The author has no relevant financial interest in the products or
companies described in this article.
Reprints: Jinru Shia, MD, Department of Pathology, Memorial Sloan-
Kettering Cancer Center, 1275 York Ave, New York, NY 10065 (e-mail:
shiaj@mskcc.org).
Arch Pathol Lab Med—Vol 134, November 2010
Data Sources.—Recent literature on clinical and patho-
logic characteristics of anal canal tumors.
Conclusions.—Although most anal canal tumors are of
squamous lineage, a complex variety of other tumors also
occurs. Recognition of such diverse tumor entities will
allow accurate pathologic diagnosis and most optimal
clinical management.
(Arch Pathol Lab Med. 2010;134:1601–1611)
canal’’ extends cephalocaudally from the level of the
pelvic floor (the anorectal ring or the junction of the
puborectalis portion of the levator ani muscle with the
external anal sphincter) to the proximal margin of the anal
verge. Thus defined, the anal canal corresponds to the
segment that is invested by the internal anal sphincter,
and is about 4.0 cm in length. The dentate (pectinate) line
is located roughly at the midpoint.
Histologically, the anal canal is divided into 3 parts on
the basis of its lining epithelium: colorectal zone lined by
colorectal-type glandular mucosa proximally, anal transi-
tion zone (ATZ) often lined by an epithelium that has
varying appearances in the middle, and the squamous
mucosa–lined distal portion (Figure 1). The delineation of
the lower border of the anal canal, that is, the junction
between squamous mucosa and anal margin epidermis, is
often evident grossly by the different consistency and
appearance of the 2 lining epithelia and microscopically
by the lack of skin adnexal structures in squamous mucosa
and their presence in anal margin. The upper border,
however, is difficult to delineate because of the lack of any
discernible landmarks either grossly or microscopically.
Most pathologists therefore tend to ignore the colorectal
zone and rely only on the grossly visible dentate line as the
superior border of the anal canal (which indeed is the
definition of the ‘‘anatomic anal canal’’ 2). Attempting
standardization, the 6th edition of the American Joint
Committee on Cancer (AJCC) staging manual3 arbitrarily
defines the anal canal as beginning 1 to 2 cm above the
dentate line; and the committee recommends that tumors
with an epicenter that is located 2 cm or less above the
dentate line be staged as anal canal cancer.3,4
The lining epithelium of the ATZ is a subject of interest.
Embryologically, this epithelium represents the fusion
point of 2 epithelial derivatives from different germ layers,
one from the endodermal hindgut lining the superior part
of the primitive anal canal, and the other from the
ectodermal protoderm lining the inferior part. Histologi-
cally, it is often composed of multiple cell layers,
extending from the small basal cells at the bottom to the
Tumors of the Anal Canal—Shia 1601
Figure 1. Schematic drawing of the anal canal.
surface cells (which can be columnar, cuboidal or
polygonal, or flattened)5 (Figure 2). Here small areas can
show umbrella-shaped surface cells and more distinct cell
borders. In the literature, this epithelium has been
variously designated as transitional, intermediate, or
cloacogenic. Scattered foci of colorectal crypts or small
areas of mature squamous epithelium may also be seen in
the ATZ. In some individuals, this histologic transitional
zone interposed between colorectal type glandular mu-
cosa and squamous mucosa may be absent altogether.
The epithelium in the ATZ may also contain mucin,
endocrine cells,6 and melanocytes.7 In addition, the anal
ducts that drain anal glands open in the ATZ (Figure 3).
The anal ducts may penetrate the internal anal sphincter
and may extend into perianal fat. The duct epithelium
shares morphologic and immunohistochemical character-
istics with the overlying ATZ mucosa. Both tend to be
positive for CK7 and negative for CK20.8 It is believed that
the anal glands and ducts may take part in the formation
of anal fistulae. These various epithelial elements account
for the different types of neoplasms that may occur in the
anal canal.
SQUAMOUS NEOPLASIA
Overview
Anal squamous cell carcinoma originates from either
the squamous epithelium of the lower part of the anal
canal (most common) or the ATZ mucosa. The occurrence
of histologic variants of squamous cell carcinoma, such as
transitional, basaloid, and cloacogenic types, has been
attributed to the varied nature of the ATZ epithelium.
Squamous cell carcinoma and its variants account for
about 70% of all anal cancers in the United States.9 Various
etiologies have been implicated in cancer development,
1602 Arch Pathol Lab Med—Vol 134, November 2010
the most significant being human papillomavirus (HPV)
infection. An anal squamous cell carcinoma may or may
not be preceded by squamous dysplasia. Subtyping of
anal squamous cell carcinomas was once advocated by the
WHO but is now believed to be of no clinical significance.
Staging of anal canal cancer is based on the tumor size and
the unique lymphatic drainage system in this region.
Cancer prevention and treatment have evolved over time.
Issues remain regarding optimizing diagnostic and
therapeutic approaches and determining the potential
for prophylactic HPV vaccines to prevent anal HPV
infection and anal cancer in at-risk groups.
Risk Factors
Epidemiologic studies have shown that most anal
cancers are associated with HPV infection, predominately
oncogenic types 16 (HPV-16) and 18 (HPV-18).9,10 A recent
meta-analysis reported an HPV prevalence of 91.5%,
93.9%, and 84.3% among 671 anal intraepithelial neoplasia
(AIN)1 cases, 609 AIN 2/3 cases, and 955 anal carcinomas,
respectively.11 Anal intercourse is among the presumed
mechanisms by which HPV is introduced into the anal
canal. Men with human immunodeficiency virus (HIV)
are also at increased risk for anal cancer.
Other risk factors include an increasing number of
sexual partners, a history of anogenital warts, previous
lower genital tract dysplasia or carcinoma, and a history of
smoking. In addition to HIV infection, immunosuppres-
sion in solid organ transplant and immune disorders has
also been shown to be a risk factor.10
Preinvasive Lesions
1) Anal Condyloma Acuminatum (Anal Wart).—Although
the natural history of progression to squamous dysplasia
Tumors of the Anal Canal—Shia
Figure 2. Anal transition zone mucosa composed of several layers of immature squamous-appearing cells. Some surface cells appear umbrella
shaped, while others contain cytoplastic mucin (hematoxylin-eosin, original magnification 3200).
Figure 3. Anal ducts present in deep submucosa and going down to the internal anal sphincter muscle. Note duct opening in overlying anal
transition zone mucosa. Inset shows the presence of mucinous cytoplasm in some surface cells (hematoxylin-eosin, original magnifications 340 and
3200 [inset]).
Figure 4. An anal canal squamous cell carcinoma with basaloid features showing a trabecular growth pattern with peripheral palisading. The cells
have high nuclear to cytoplasmic ratio. There are no distinct intercellular bridges (hematoxylin-eosin, original magnification 3200).
Figure 5. An anal canal squamous cell carcinoma showing cystic degeneration and focal presence of mucin-containing tumor cell (inset)
(hematoxylin-eosin, original magnifications 3400).

and squamous cell carcinoma is still to be defined, anal
condyloma acuminatum (anal wart) is in general regarded
as a lesion with the potential to develop cancer, the
reported incidence of squamous cell carcinoma of the anus
in patients with anal condyloma being 3% to 4% and likely
increasing.12,13
The histologic hallmark of a typical condyloma
acuminatum is the presence of koilocytic changes
(vacuolated cytoplasm and enlarged, hyperchromatic
nuclei with irregular nuclear membranes, with or with-
out binucleation) in a cauliflower-like growth that is
composed of papillary excrescences lined by hyperker-
atotic squamous epithelium. Koilocytic changes may be
absent, however, presumably because the viral infection
has subsided.14
While many anal condylomata, particularly those of
anal skin (perineum), harbor low-risk HPV (types 6 or 11)
and may therefore be regarded as having a minimal risk of
Arch Pathol Lab Med—Vol 134, November 2010
progression to cancer, those of anal canal may be
associated with high-risk types (such as HPV-16) or a
mixture of low- and high-risk types and may go on to
develop malignancy. This is particularly true in immuno-
compromised patients. Therefore, condylomata without
high-grade dysplasia do not necessarily always equate to
low-risk HPV infection, a misconception not uncommon
among both clinicians and pathologists.
For pathologic diagnosis, those lesions with typical
condylomatous features as described above may be
diagnosed as ‘‘condyloma acuminatum,’’ whereas those
that also have conventional histologic evidence of low-
grade dysplasia (ie, atypia and mitoses in the lower third
of the epithelium) as ‘‘condyloma acuminatum with mild
dysplasia (low-grade squamous intraepithelial lesion/
anal intraepithelial neoplasia 1 [LSIL/AIN I]).’’ Others
may classify all condylomata without high-grade dyspla-
sia under the term condyloma acuminatum (LSIL/AIN I), or
Tumors of the Anal Canal—Shia 1603
alternatively, condyloma acuminatum; there is no evidence of
high-grade dysplasia.15
2) Flat Dysplasia.—Compared to condyloma, flat squa-
mous lesions are more frequently associated with high-
risk HPV infection. Grading of dysplasia again follows
what has been established for cervical lesions. In the 2-
tiered system, mild dysplasia constitutes LSIL. High-
grade SIL (HSIL) then encompasses lesions that have
mitotic activity higher than the lower third of the mucosa
and that often have full thickness atypia and loss of
polarity (moderate and severe dysplasia).
Terminology for noncondylomatous squamous dyspla-
sia involving the anal canal mucosa (ATZ and the
squamous zone) versus that involving the hair-bearing
anal/perianal skin can be confusing. Conventionally,
lesions involving the former have been termed dysplasia,
in situ carcinoma, or anal intraepithelial neoplasia, whereas
lesions involving the skin, Bowen disease. Bowen disease
conventionally carries with it the implication of having a
strong tendency to recur after local treatment but with
only a small percentage progressing to invasive squamous
cell carcinoma.16,17 However, given that such squamous
lesions are not always restricted to 1 region, distinct
separation of terminology does not always seem logical.
Indeed, the more recent recommendation states that all
squamous lesions occurring both in the anal canal and the
perianal skin or skin of perineum be termed either low-
grade squamous intraepithelial lesion or high-grade squamous
intraepithelial lesion.18 Bowen disease would then fall into
the HSIL category. Such recommendations, however, have
yet to be fully adopted. At the current time, some
pathologists still prefer to use Bowen disease for anal
skin, while others are using dysplasia or squamous
intraepithelial lesion for all sites.
The utility of p16 and Ki-67 immunohistochemical
staining for diagnosing anal squamous dysplasia has
been studied less intensely than it has for gynecologic
cases. A recent study of anal intraepithelial neoplasia
indicated that, like in the uterine cervix, both markers
offer utility to distinguish low-grade SIL from high-grade
SIL. In general, high-grade SIL tends to have diffuse p16
staining and an increased number of Ki-67–positive cells
in the middle or upper third of the epithelium (positive
p16 staining refers to the presence of nuclear or
cytoplasmic reactivity and for Ki-67, the presence of
nuclear reactivity).19 It is to be noted, however, that p16
positive staining has also been observed in LSIL asso-
ciated with high-risk HPV types. Thus, these markers may
be more valuable in the distinction of reactive squamous
epithelium from dysplastic squamous epithelium than in
separating low-grade from high-grade dysplasia.15
Giant Condyloma of Buschke and Lowenstein/
Verrucous Carcinoma
First described by Buschke in 1886 and later by Buschke
and Lowenstein in 1925,20 the ‘‘giant condyloma of
Buschke and Lowenstein’’ refers to a slow-growing
neoplasm that has a tendency to recur and to form
abscesses and fistulae. Verrucous carcinoma was first
described by Ackerman21 in 1948 for a low-grade
carcinoma of oral mucosal viral warts. Subsequently, this
term has been expanded to lesions of other sites. It is
believed that the giant condyloma of Buschke and Low-
enstein represents the anogenital version of verrucous
1604 Arch Pathol Lab Med—Vol 134, November 2010
carcinoma and is an intermediate state between condy-
loma acuminatum and squamous cell carcinoma.1,15,22 It
appears that this tumor does not arise from malignant
transformation of a condyloma but represents a low-grade
form of squamous cell carcinoma. Like conventional
condylomata, these lesions are more likely to be associated
with low-risk HPV. It remains to be elucidated which viral
or host risk factors play a role in promoting the oncogenic
potential of HPV-6 and HPV-11 to result in a giant
condyloma of Buschke and Lowenstein that can be locally
invasive.23–26
The histologic appearance on a biopsy specimen may be
identical to that seen in common condyloma acuminatum
(surface hyperkeratosis, prominent acanthosis and papil-
lomatosis, with orderly arrangement of the epithelial
layers). However, excision specimens will often demon-
strate an endophytic component that is not present in
ordinary condylomata. The endophytic growth is com-
posed of an intact but often irregular base with blunt
downward projections and keratin-filled cysts and is
associated with destruction of the underlying tissue.
The biology of such lesions is typically that of local
invasion without metastasis. It is to be noted that the
presence of severe cytologic atypia or stromal invasion in
an infiltrative fashion (different from the pushing down-
ward growth as described above) should lead to the
diagnosis of squamous cell carcinoma, with appropriately
aggressive therapy (see below).
Squamous Cell Carcinoma
Anal squamous cell carcinoma occurs more frequently
in women than in men (1.5 versus 1.0 per 100 000).10 The
clinical presentation is nonspecific. Anal bleeding appears
to be the most common presenting symptom.
Histologic subtyping of anal squamous cell carcinomas
was once recommended by the WHO,27 and the categories
included large-cell keratinizing squamous carcinoma,
large-cell nonkeratinizing squamous carcinoma, and
basaloid carcinoma. Others28,29 have advocated dividing
these tumors into (1) squamous cell carcinoma, analogous
to its counterpart elsewhere in the skin, and (2) cloaco-
genic (transitional, basaloid) carcinoma, supposedly orig-
inating from the ATZ mucosa.
The morphologic patterns that are commonly associated
with basaloid or cloacogenic carcinoma include nested
and trabecular growth of small cells without the inter-
cellular bridges typical of conventional squamous cell
carcinoma (Figure 4), although foci of more conventional
squamous differentiation can usually be found. Peripheral
palisading similar to that of cutaneous basal cell carci-
noma may be present. Sometimes, these tumors can have
small cystic foci lined by mucin-producing cells, the
reason for the former designation of mucoepidermoid
carcinoma (Figure 5). Others may grow in lobules and
contain prominent eosinophilic, hyaline, paucicellular
basement membrane–like material around and within
tumor nests, resulting in an appearance simulating that of
an adenoid cystic carcinoma30 (Figure 6, A through D).
Central necrosis and mitotic figures may be prominent,
but cellular pleomorphism is not typical.
However, histologic subtyping of anal squamous cell
carcinomas is difficult in practice and has little clinical
value. Very commonly, a single tumor can show mixtures
of histologic subtypes, including foci of conventional
squamous cell carcinoma. Thus, the WHO now suggests
Tumors of the Anal Canal—Shia
Figure 6. An anal canal squamous cell carcinoma resembling adenoid cystic carcinoma with prominent eosinophilic, hyaline basement
membrane–like material around and within tumor nests seen at low (A) and higher (B) magnification. This tumor is positive for both 34BE12 (C) and
4A4/p63 (D) by immunohistochemistry and negative for chromogranin and synaptophysin (not shown) (hematoxylin-eosin, original magnifications
320 [A] and 3100 [B]; original magnifications 340 [C and D]).

that the generic term squamous cell carcinoma be used for all
these variants.1 However, the following histologic sub-
types seem to have a less favorable prognosis31: (1)
squamous cell carcinoma with mucinous microcysts
(areas with well-formed acinar or cystic spaces containing
mucin that are positive for Alcian blue and periodic acid–
Schiff with diastase), and (2) small cell (anaplastic)
carcinoma (not small cell neuroendocrine carcinoma).
Moreover, poor keratinization, prominent basaloid fea-
tures, and small tumor cell size are related to infection
with high-risk HPV.32 Thus, in the pathologic diagnosis, a
comment may be made about these histologic variations
that may affect prognosis or reflect different etiology.
Grading of squamous cell carcinoma should be based
on the degree of cellular differentiation, which is reflected
mainly by the extent of keratin production and cell
cohesiveness. Poor differentiation based on the degree of
dissociation of tumor cells33 has been related to poor
prognosis by univariate analysis. Grading based on biopsy
material may be misleading, as it may not be representa-
tive of the tumor as a whole.
As defined by the AJCC,3 pathologic tumor staging of
anal squamous cell carcinoma is based on tumor size, and
Arch Pathol Lab Med—Vol 134, November 2010
pathologic lymph node staging is based on the lymphatic
drainage system in this region. Tumors proximal to the
pectinate line drain into the pelvis along the middle rectal
vessels to the pelvic side walls and internal iliac chains
and superiorly via the superior rectal vessels to the
periaortic nodes. Tumors distal to the dentate line drain
along cutaneous pathways to the inguinal and the femoral
nodal chains.
Prevention and Management of Squamous Cell Lesions
Although many independent as well as interrelated
etiologies have been implicated in anal squamous cell
lesions, HPV remains the most significant. Thus, much
attention has been given to the role of HPV typing in
triaging patients with anal lesions, particularly in high-risk
patient populations. However, as yet, whether the presence
of specific HPV subtypes as detected by either HPV in situ
hybridization or polymerase chain reaction will allow
more accurate and cost-effective prediction of precancer-
ous lesions remains an unanswered question. Currently,
intense screening with high-resolution anoscopy and
cytology, and HPV testing with polymerase chain reaction
or fluorescence in situ hybridization are mainly limited to
Tumors of the Anal Canal—Shia 1605
investigations in high-risk subpopulations such as HIV-
positive homosexual men. The efficacy of HPV prophylac-
tic vaccines in preventing anal squamous neoplasia is also
still at an investigational stage, although its clinical use in
the near future may be anticipated.34
Management of precursor lesions remains a challenge.
Squamous intraepithelial lesion is often a multifocal
process; thus, complete local excision is often inadequate
or not possible because of the morbidity of the proce-
dure.35,36 Newer strategies include topical immunomodu-
lation, photodynamic therapy, and therapeutic vaccines.
However, long-term follow-up to determine the efficacy of
these treatments is not available. Office-based or operating
room procedures directed by high-resolution anoscopy
and cytology are believed to result in clearance of HSIL in
up to 80% of patients, malignant progression in 1%, and
less morbidity than wide local excision.37
Typical giant condyloma of Buschke-Lowenstein/ver-
rucous carcinoma is often managed surgically with
complete excision being the preferred initial treatment.
Resection, chemotherapy, and focused radiation have all
been used; however, the response has been generally
unpredictable.25
Management of invasive anal squamous cell carcinoma
has shifted during the past 3 to 4 decades from surgery (ie,
abdominoperineal resection) to a combined therapy
incorporating pelvic radiation and chemotherapy (5-
fluorouracil and mitomycin C), with surgery reserved
for persistent or recurrent disease. This combined mod-
ality therapy has resulted in an estimated 5-year disease-
free survival of 60% and 5-year overall survival of 75%.38
Significant predictors for treatment failure include high T
stage and completion of radiotherapy.39 In the era of
highly active antiretroviral therapy, the outcome after
chemoradiotherapy for HIV-related cancers is comparable
to that for patients without HIV, although there may be
significant toxicity; earlier diagnosis and risk-adapted
therapy are key to success for such patients.40
ADENOCARCINOMA
Overview
Adenocarcinoma of the anal canal accounts for about
10% (5%–19%) of all anal canal cancers.41 Most such
tumors have a colorectal phenotype and represent
tumors originating from the colorectal zone in the upper
portion of the anal canal or from the glandular cells of
the ATZ mucosa. Distinction of these tumors from lower
rectal adenocarcinomas directly extending into the anal
canal can be very difficult or impossible. ‘‘Anal gland
adenocarcinoma’’ and ‘‘fistula associated adenocarci-
noma’’ constitute the 2 other major types of adenocarci-
noma; both are categorized by the WHO under the term
extramucosal (perianal) adenocarcinoma. Finally, Paget
disease represents a form of intraepithelial adenocarci-
noma and will be discussed in this section as well. As a
whole, adenocarcinoma of the anal canal represents a
unique challenge both for pathologic diagnosis and for
clinical management. Its rarity makes it difficult to
perform controlled clinical studies or large retrospective
reviews.
Risk Factors
Little is known about risk factors. High-risk HPV types
may play a role, as their presence has been documented in
1606 Arch Pathol Lab Med—Vol 134, November 2010
at least some cases of anal adenocarcinoma.42 Crohn
disease or other inflammatory conditions that result in
chronic anal fistula may predispose to development of
fistula-associated adenocarcinomas.43
Colorectal-Type Adenocarcinoma
The most significant clinical implication of distinguish-
ing the anal canal origin of these tumors relates to its
pattern of local spread, which reflects the dual lymphatic
drainage as mentioned under staging of squamous cell
carcinomas. Thus, a colorectal type adenocarcinoma
occurring within the anal canal carries a higher risk of
nodal disease along the inguinal and femoral nodal chains
than a rectum-based adenocarcinoma.
Histologic diagnosis rarely poses a significant chal-
lenge. However, it is worth mentioning that colorectal
type adenocarcinoma of this site (as is also true for the
distal rectum) may have unexpected CK7 expression,
although it is usually accompanied by coexpression of
CK20 (as opposed to anal gland carcinomas that usually
do not have CK20 positivity, see below).
Anal Gland Adenocarcinoma
Although the original WHO classification listed this
tumor under the category ‘‘adenocarcinoma of anal
glands,’’ these tumors may actually arise from or be
related to anal ducts that drain the anal glands, and may
therefore be more appropriately termed anal duct adeno-
carcinoma. In reality, the association or connection with
normal anal gland/ducts is often hard to demonstrate
histologically, and the reported cases under this category
vary widely in microscopic appearance.
Hobbs et al44 in 2001 studied 14 cases of anal canal
adenocarcinoma and identified 7 cases that fit a set of
criteria that the authors defined for carcinoma of anal
gland type: haphazardly dispersed, small glands with
scant mucin production that invade the wall of the
anorectal area without an intraluminal component; the
tumor glands are positive for CK7. Others reported
clinically and grossly typical anal gland–type cancers
with a histologic appearance inseparable from that of
conventional mucinous-type colorectal adenocarcinoma.
Moreover, carcinomas associated with anal fistulae are
often mucinous, and it remains to be determined if at least
some of such fistula-associated tumors are actually of anal
gland or anal duct origin.45
Thus, the histologic pattern of anal gland adenocarci-
noma remains to be sharply defined. In practice, it would
seem reasonable to render this diagnosis when the tumor
is primary to the anal canal, centered within the wall of the
anorectal area without a preexisting fistula and without
surface mucosa dysplasia, irrespective of the extent of
mucin production (Figure 7).
Adenocarcinoma Within Anorectal Fistula
Anorectal fistulae may be developmental or acquired.
The latter are often secondary to inflammatory conditions
such as Crohn disease. Adenocarcinomas arising in
anorectal fistulae have been documented and histologic
appearance of such tumors is often that of a well-
differentiated mucinous adenocarcinoma. The exact cell
of origin is difficult to determine. Some may possibly be
related to rectal-type glandular mucosa, while others are
related to anal glands or ducts. As mentioned above,
Tumors of the Anal Canal—Shia

Figure 7. This mucin-producing adenocarcinoma is centered in the
wall of the anorectal transition zone, not associated with surface
mucosa dysplasia or fistula disease. It is tempting to assume that this
tumor is of anal gland or anal duct origin (hematoxylin-eosin, original
magnification 340).
Figure 8. Anal Paget disease associated with nonneoplastic prolifera-
tion of background squamous mucosa (hematoxylin-eosin, original
magnification 3400).
distinction of anal gland adenocarcinoma from adenocar-
cinoma within anorectal fistula can be difficult, as some
fistulae may result from dilated anal glands or ducts.
Indeed, histochemical studies seem to suggest that at least
some fistula-associated adenocarcinomas have mucin
characteristics that are similar to those of anal glands.46
Paget Disease
Extramammary Paget disease was first described in
1889 in an article by Crocker47 entitled ‘‘Paget’s Disease
Affecting the Scrotum and Penis,’’ 15 years after James
Paget’s initial description of mammary Paget disease.48
Subsequently, the lesion was also seen in other anogenital
sites, including vulvar skin (most common), perineum,
anal margin skin, and anal canal.
Clinically, the disease often presents in elderly patients
as a pruritic, red or white, crusted patch in anal skin.
Conventionally, anal Paget disease, similar to its counter-
part in other extramammary sites, has been divided into
‘‘primary’’ and ‘‘secondary’’ diseases. Primary Paget
disease encompasses those cases that originate from the
Arch Pathol Lab Med—Vol 134, November 2010
epidermis or squamous epithelium and may or may not
evolve into an invasive lesion during its course. The
histogenesis of primary Paget disease remains elusive,
with the proposed cells of origin including pluripotent
epidermal stem cells,49,50 adnexal stem cells,51 intraepider-
mal Toker cells,52,53 or apocrine glands.54 ‘‘Secondary’’
Paget disease in the anal region is often associated with an
underlying visceral malignancy, most commonly a colo-
rectal type adenocarcinoma. The visceral malignancy may
be synchronous or metachronous. Given that the literature
reports are largely limited to case reports or small series, it
is difficult to accurately determine the prevalence of
primary versus secondary Paget disease. Some55 estimate
that in patients who have perianal Paget disease, the
incidence of a synchronous visceral carcinoma is more than
50%. Others56,57 suggest that perianal Paget disease may be
associated with tubo-ovarian adenocarcinoma in 7% to
24% of cases and gastrointestinal carcinoma in 12% to 14%.
Sometimes, a distal rectal or anal canal adenocarcinoma
may be associated with very limited pagetoid extension of
cancer cells in the overlying or immediately adjacent
squamous epithelium that does not result in a gross
appearance of Paget disease. Strictly speaking, this is a
form of secondary Paget disease as well, although its
clinical significance in this setting seems trivial.
Diagnostic issues arise when a primary anal Paget
disease is associated with an invasive component, or when
a secondary Paget disease is associated with a metachro-
nous internal malignancy that is not apparent at the time
the Paget disease manifests clinically. Consequently, there
have been research efforts that aim at distinguishing
primary from secondary conditions by immunohisto-
chemical phenotyping. The rarity of the disease, however,
has been a significant limiting factor to such efforts. Thus,
although promising patterns have emerged, issues re-
main. Goldblum and Hart58 reported that 4 of 4 cases of
anal Paget disease that were associated with rectal
adenocarcinoma showed a CK7+/CK20+/GCDFP152
phenotype in both the Paget disease and the rectal
adenocarcinoma, whereas 4 of 6 cases of anal Paget
disease without a documented rectal adenocarcinoma had
a CK7+/CK202/GCDFP15+ phenotype. Similar results
also have been reported by others in both anal and vulvar
Paget disease. Thus, CK7 appears to be a universal marker
of Paget cells, regardless of their site or the presence of an
associated internal malignancy. CK20 positivity seems to
be a uniform finding in cases that are associated with
rectal adenocarcinoma. However, CK20 expression has
been observed in occasional cases of primary anal and
vulvar Paget disease; thus, its specificity is less than
perfect. In such cases, GCDFP15 expression may be
helpful.
It is intriguing to note that a more recent study8 reported
a CK7+/CK20+ phenotype in only 13% (4 of 30) of rectal
adenocarcinomas unassociated with Paget disease, a rate
comparable to that reported for nonrectal large bowel
adenocarcinomas. The question then is what promotes
certain CK7-positive rectal adenocarcinomas to develop
secondary Paget disease.
Interestingly, anal Paget disease is typically associated
with hyperplastic changes in the involved squamous
mucosa (Figure 8). Awareness of such changes may help
point to the diagnosis of Paget disease (particularly at the
time of frozen section procedure) or avoid the misdiag-
nosis of squamous dysplasia.
Tumors of the Anal Canal—Shia 1607
Figure 9. Virtually all malignant tumors occurring in the anal canal can become poorly differentiated and difficult to recognize. Shown here are a
large cell–type, high-grade neuroendocrine carcinoma (A), a poorly differentiated squamous cell carcinoma (B), a poorly differentiated
adenocarcinoma (C), and a malignant melanoma (D) (hematoxylin-eosin, original magnifications 3200 [A, B, and C] and 3400 [D]).

Prevention and Management
Little is known about cancer prevention that is specific
for anal adenocarcinoma. Treatment options vary widely
from primary surgical resection to definitive chemoradia-
tion or neoadjuvant chemoradiation and surgical resec-
tion.41,59,60 The management of Paget disease is primarily
surgical. Local recurrence is common because of the
common presence of multifocal disease, and the difficulty
in delineating the lesion grossly. Survival outcome in
these patients is often dictated by the presence or absence
and stage of an invasive component.
NEUROENDOCRINE NEOPLASMS
Neuroendocrine neoplasms may occasionally occur in
the anal canal. Most such tumors probably originate from
neuroendocrine cells residing in colorectal type mucosa,
although neuroendocrine cells are known to exist in ATZ
mucosa as well.
Studies dedicated solely to anal canal neuroendocrine
neoplasms are essentially limited to case reports. As a
group, well-differentiated neuroendocrine tumors, that is,
carcinoid tumors, of the anorectum are believed to have an
indolent clinical course. In a recent study of 70 carcinoid
tumors of the rectum (including those of anal canal),61 we
1608 Arch Pathol Lab Med—Vol 134, November 2010
observed the following features as indicators for poor
prognosis: large size, deep invasion, lymphovascular
invasion, and an elevated mitotic rate (.2 per 50 high-
power fields [HPFs]).
High-grade neuroendocrine carcinomas, small cell or
non–small cell type, may occur in the anal canal as well,
constituting 5% (3 of 65) of all high-grade neuroendocrine
carcinomas of the entire tubular gastrointestinal tract in
our series.62
From a diagnostic point of view, anorectal carcinoid
tumors with a tubular pattern (a pattern not uncom-
monly seen in this location) may be confused with
conventional adenocarcinoma; and high-grade neuroen-
docrine carcinomas may be confused with poorly
differentiated adenocarcinoma, basaloid squamous cell
carcinoma, or melanoma. This may be particularly
problematic in biopsy samples. In this scenario, immu-
nohistochemical stains are helpful. Expression of neu-
roendocrine markers is common but not universal, and
labeling may be focal. Small cell carcinomas of the anus
may show immunoreactivity for thyroid transcription
factor–1 (TTF-1), a phenomenon reported in a variety of
other extrapulmonary sites as well. Stains for squamous
differentiation (34bE12 and p63/4A4) yield negative
Tumors of the Anal Canal—Shia
 Table 1. Morphologic and Immunophenotypic Features of Anal Canal Malignanciesa
Immunohistochemical Staining
Lym-
HMW NE Melanoma phoma
Tumor Type Morphology Keratin 4A4/P63 Ki-67 Markers Markers Markers
Squamous Squamous differentiation ++ ++ 2 2 2
cell carcinoma
Adenocarcinoma Mucin, gland formation 2/+ 2 2 2 2
NE tumor
Carcinoid Trabecular, organoid, or 2 2 Usually ,10% ++ 2 2
nested. No or low mitotic
activity.
Small cell Can have foci of squamous 2 2 Usually .50% + 2 2
carcinoma differentiation (Squamous (Squamous
foci +) foci +)
High-grade NE May simulate carcinoid, but 2 2 Usually + 2 2
carcinoma with high mitotic activity .30%–50%
and often foci of necrosis
Melanoma In situ component, melanin 2 2 2 ++ 2
pigment
Sarcoma Lacks cohesiveness 2 2 2 2 2
Lymphoma Lacks cohesiveness 2 2 2 2 ++
Abbreviations: HMW, high molecular weight; NE, neuroendocrine.
a
No entry indicates data not relevant.

results in anal small cell carcinomas. However, similar to
its counterpart in other sites, anal small cell carcinoma
may contain scattered nests of cells with squamous
differentiation, usually constituting less than 5% of the
entire tumor volume.62 Awareness of such a phenomenon
will help avoid misdiagnosing such tumors as true
squamous cell carcinomas.
MALIGNANT MELANOMA
Anal melanomas account for about 4% of anal canal
tumors and less than 1% of all melanomas.63
The histologic features of anal melanoma resemble
those of cutaneous melanomas. Most show a junctional
component adjacent to the invasive tumor, and this
finding is evidence that the lesion is primary. The
desmoplastic variant may occasionally occur at this site
as well, and diagnosis may be challenging and requires
immunohistochemical support.
Management of anal melanoma remains a major
challenge. Despite the fact that most patients present with
localized and apparently curable primary tumors, the
mean survival is only 2 years. The extent of surgical
resection (abdominoperineal resection versus local exci-
sion) does not seem to significantly impact outcome, as
patients often die of distant metastases. In a study of 46
patients, 64 perineural invasion was an important prog-
nostic factor.
Recent studies65 have shed light on the pathogenetic
molecular pathways in melanoma of various sites. In a
study of 20 anal melanomas, Antonescu et al66 identified 3
KIT mutation-carrying tumors, and by in vitro drug
testing, showed that the KIT (L576P) mutant was
responsive to specific kinase inhibitors.
MESENCHYMAL TUMORS
Although a variety of mesenchymal tumors may occur
in the anal canal, such as smooth muscle tumors,
gastrointestinal stromal tumors (GISTs), schwannomas,
sarcomas related to endometriosis or the müllerian
system, and lymphagiomas,67 the most commonly seen
are smooth muscle tumors and GISTs. An accurate
Arch Pathol Lab Med—Vol 134, November 2010
estimation of their relative frequency is difficult because
many tumors previously designated as smooth muscle or
neural tumors may now be classified as GISTs.68
Leiomyomas often originate from muscularis mucosae,
appear as intraluminal polyps,69 and are cured by
complete removal. Leiomyosarcomas, on the other hand,
are histologically high-grade sarcomas with at least focal
pleomorphism and high mitotic activity. Most leiomyo-
sarcomas also present as intraluminal polypoid tumors,
although origin specifically from the muscularis mucosae
could not be documented in any of the 8 leiomyosarcomas
of the rectum and anus studied by Miettinen et al.70
Anorectal leiomyosarcomas may have a better prognosis
when compared to GISTs with comparable mitotic
counts.70
Gastrointestinal stromal tumors only infrequently arise
in the large bowel. Within the large bowel, however, the
rectum is the most common site.70 Gastrointestinal stromal
tumors of the anal canal are on record as well,71,72
estimated to account for about 2% to 8% of anorectal
GISTs. In the anorectum, GISTs may be of spindle cell
(65%), epithelioid, or mixed types. Immunohistochemi-
cally, 84% of c-kit–positive GISTs stain positively for
CD34, 29% for smooth muscle actin, and 4% for S100
protein.73 As with GISTs of other sites, tumor size and
mitotic count are significant prognostic factors. Miettinen
et al70 have shown that in the rectum and anus, GISTs
larger than 5 cm or with more than 5 mitoses per 50 HPFs
are likely to behave in a malignant fashion, whereas those
smaller than 2 cm and with less than 5 mitoses per 50
HPFs only rarely recur or cause death, and those with a
size between 2 cm and 5 cm and a mitotic count of less
than 5 mitoses per 50 HPFs have a low frequency of
malignant behavior, intermediate between the above 2
categories.
MALIGNANT LYMPHOMA
Primary lymphoma of the anal canal is rare. However,
cases of both Hodgkin disease and non-Hodgkin lym-
phomas are on record.74–76 Immunocompromised patients
infected with HIV are particularly at risk.77 In this
Tumors of the Anal Canal—Shia 1609

Table 2. Potential Pitfalls in the Interpretation of
Immunohistochemical Staining in Anal
Canal Neoplasms
Melanoma versus GIST A103 (Melan-A) may stain some
epithelioid GISTs78
c-kit stains not only GIST but also
melanoma
Poorly differentiated CK18 may be positive in GISTs and
carcinoma versus smooth muscle tumors
sarcoma
Neuroendocrine tumors Neuroendocrine tumors can be
versus others positive for CK7 and/or CK20
Small cell carcinoma primary to the
anal canal may be positive for
TTF-1
Abbreviations: GIST, gastrointestinal stromal tumor; TTF-1, thyroid
transcription factor 1.
population, the lymphomas are mainly B-cell type and
high grade. In contrast, in the non-HIV–infected popula-
tion, anal lymphomas tend to occur in older patients, are
also B-cell type, but are often lower grade.
TUMORLIKE CONDITIONS
A variety of lesions may mimic anal neoplasms. For
example, ectopic prostatic tissue, endometriosis, inflam-
matory cloacogenic polyp, and prolapse-related lesions
can all form a mass lesion.
DIFFERENTIAL DIAGNOSIS OF POORLY
DIFFERENTIATED ANAL CANAL NEOPLASMS
Essentially all malignant tumor entities occurring in the
anal canal may be poorly differentiated, such that the
distinction among the various histologic types becomes a
challenge. Commonly, these tumors assume the appear-
ance of a cellular neoplasm with a solid growth pattern
(Figure 9, A through D). A typical challenging scenario is
distinguishing a poorly differentiated squamous cell
carcinoma with basaloid features from a small cell
(neuroendocrine) carcinoma. Also typically challenging
is the distinction of a malignant melanoma from literally
every other cellular tumor with a solid growth pattern,
including GIST. Table 1 provides some key morphologic
and immunohistochemical features that are useful in
resolving these diagnostic dilemmas. Morphologically, a
number of points deserve special attention. First, small cell
carcinoma may have foci of abrupt squamous differentia-
tion, as described above, and as such, can be mistaken for
squamous cell carcinoma.62 Second, carcinoid tumors and
small cell carcinomas (or non-small cell–type high-grade
neuroendocrine carcinomas) can be superficially similar,
particularly in a small and crushed biopsy specimen. In
this scenario, the presence of increased mitotic activity
and high Ki-67 labeling index favor a diagnosis of high-
grade neuroendocrine carcinoma or small cell carcinoma.
Lastly, in situ growth and pigment production can be very
helpful in diagnosing melanoma. From an immunohisto-
chemical point of view, additional worthwhile points are
listed in Table 2. Ultimately, however, the key to achiev-
ing a correct diagnosis is the integration of morphologic
and immunohistochemical findings.
References
1. Hamilton SR, Aaltonen LA, eds. Pathology and Genetics of Tumours of the
Digestive System. Lyon, France: IARC Press; 2000. World Health Organization
Classification of Tumours.
1610 Arch Pathol Lab Med—Vol 134, November 2010
2. Morgan CN, Thompson HR. Surgical anatomy of the anal canal with special
reference to the surgical importance of the internal sphincter and conjoint
longitudinal muscle. Ann R Coll Surg Engl. 1956;19(2):88–114.
3. Greene Fl, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual. 6th ed.
New York, NY: Springer-Verlag; 2002.
4. Compton CC. Colorectal carcinoma: diagnostic, prognostic, and molecular
features. Mod Pathol. 2003;16(4):376–388.
5. Fenger C. The anal transitional zone. Acta Pathol Microbiol Immunol Scand
Suppl. 1987;289:1–42.
6. Fetissof F, Dubois MP, Assan R, et al. Endocrine cells in the anal canal.
Virchows Arch A Pathol Anat Histopathol. 1984;404(1):39–47.
7. Clemmensen OJ, Fenger C. Melanocytes in the anal canal epithelium.
Histopathology. 1991;18(3):237–241.
8. Ramalingam P, Hart WR, Goldblum JR. Cytokeratin subset immunostaining
in rectal adenocarcinoma and normal anal glands. Arch Pathol Lab Med. 2001;
125(8):1074–1077.
9. Frisch M, Melbye M. Cancer Epidemiology and Prevention. 3rd ed. New
York, NY: Oxford University Press; 2006.
10. Joseph DA, Miller JW, Wu X, et al. Understanding the burden of human
papillomavirus-associated anal cancers in the US. Cancer. 2008;113(suppl 10):
2892–2900.
11. De Vuyst H, Clifford GM, Nascimento MC, Madeleine MM, Franceschi S.
Prevalence and type distribution of human papillomavirus in carcinoma and
intraepithelial neoplasia of the vulva, vagina and anus: a meta-analysis. Int J
Cancer. 2009;124(7):1626–1636.
12. Metcalf AM, Dean T. Risk of dysplasia in anal condyloma. Surgery. 1995;
118(4):724–726.
13. Byars RW, Poole GV, Barber WH. Anal carcinoma arising from
condyloma acuminata. Am Surg. 2001;67(5):469–472.
14. Rock B, Shah KV, Farmer ER. A morphologic, pathologic, and virologic
study of anogenital warts in men. Arch Dermatol. 1992;128(4):495–500.
15. Longacre TA, Kong CS, Welton ML. Diagnostic problems in anal
pathology. Adv Anat Pathol. 2008;15(5):263–278.
16. Marchesa P, Fazio VW, Oliart S, Goldblum JR, Lavery IC. Perianal
Bowen’s disease: a clinicopathologic study of 47 patients. Dis Colon Rectum.
1997;40(11):1286–1293.
17. Sarmiento JM, Wolff BG, Burgart LJ, Frizelle FA, Ilstrup DM. Perianal
Bowen’s disease: associated tumors, human papillomavirus, surgery, and other
controversies. Dis Colon Rectum. 1997;40(8):912–918.
18. Welton ML, Sharkey FE, Kahlenberg MS. The etiology and epidemiology of
anal cancer. Surg Oncol Clin N Am. 2004;13(2):263–275.
19. Bean SM, Eltoum I, Horton DK, Whitlow L, Chhieng DC. Immunohisto-
chemical expression of p16 and Ki-67 correlates with degree of anal
intraepithelial neoplasia. Am J Surg Pathol. 2007;31(4):555–561.
20. Buschke A, Lowenstein L. Uber carcinomahnliche condylomata acumi-
nata des penis. Klin Wochenschr. 1925;4:1726–1728.
21. Ackerman LV. Verrucous carcinoma of the oral cavity. Surgery. 1948;
23(4):670–678.
22. Bertram P, Treutner KH, Rubben A, Hauptmann S, Schumpelick V.
Invasive squamous-cell carcinoma in giant anorectal condyloma (Buschke-
Lowenstein tumor). Langenbecks Arch Chir. 1995;380(2):115–118.
23. Ergun SS, Kural YB, Buyukbabani N, Verim L, Akbulut H, Gurkan L. Giant
condyloma acuminatum. Dermatol Surg. 2003;29(3):300–303.
24. Geusau A, Heinz-Peer G, Volc-Platzer B, Stingl G, Kirnbauer R.
Regression of deeply infiltrating giant condyloma (Buschke-Lowenstein tumor)
following long-term intralesional interferon alfa therapy. Arch Dermatol. 2000;
136(6):707–710.
25. Trombetta LJ, Place RJ. Giant condyloma acuminatum of the anorectum—
trends in epidemiology and management: report of a case and review of the
literature. Dis Colon Rectum. 2001;44(12):1878–1886.
26. Chu QD, Vezeridis MP, Libbey NP, Wanebo HJ. Giant condyloma
acuminatum (Buschke-Lowenstein tumor) of the anorectal and perianal regions:
analysis of 42 cases. Dis Colon Rectum. 1994;37(9):950–957.
27. Jass J, Sobin LH. Histological Typing of Intestinal Tumours. 2nd ed. New
York, NY: Springer; 1989.
28. Klotz RG Jr, Pamukcoglu T, Souilliard DH. Transitional cloacogenic
carcinoma of the anal canal: clinicopathologic study of three hundred seventy-
three cases. Cancer. 1967;20(10):1727–1745.
29. Gillespie JJ, MacKay B. Histogenesis of cloacogenic carcinoma: fine
structure of anal transitional epithelium and cloacogenic carcinoma. Hum
Pathol. 1978;9(5):579–587.
30. Chetty R, Serra S, Hsieh E. Basaloid squamous carcinoma of the anal canal
with an adenoid cystic pattern: histologic and immunohistochemical reappraisal
of an unusual variant. Am J Surg Pathol. 2005;29(12):1668–1672.
31. Shepherd NA, Scholefield JH, Love SB, England J, Northover JM.
Prognostic factors in anal squamous carcinoma: a multivariate analysis of
clinical, pathological and flow cytometric parameters in 235 cases. Histopathol-
ogy. 1990;16(6):545–555.
32. Frisch M, Fenger C, van den Brule AJ, et al. Variants of squamous cell
carcinoma of the anal canal and perianal skin and their relation to human
papillomaviruses. Cancer Res. 1999;59(3):753–757.
33. Goldman S, Auer G, Erhardt K, Seligson U. Prognostic significance of
clinical stage, histologic grade, and nuclear DNA content in squamous-cell
carcinoma of the anus. Dis Colon Rectum. 1987;30(6):444–448.
Tumors of the Anal Canal—Shia
 34. Gillison ML, Chaturvedi AK, Lowy DR. HPV prophylactic vaccines and the
potential prevention of noncervical cancers in both men and women. Cancer.
2008;113(suppl 10):3036–3046.
35. Brown SR, Skinner P, Tidy J, Smith JH, Sharp F, Hosie KB. Outcome after
surgical resection for high-grade anal intraepithelial neoplasia (Bowen’s disease).
Br J Surg. 1999;86(8):1063–1066.
36. Hu K, Minsky BD, Cohen AM, et al. 30 Gy may be an adequate dose in
patients with anal cancer treated with excisional biopsy followed by combined-
modality therapy. J Surg Oncol. 1999;70(2):71–77.
37. Pineda CE, Welton ML. Controversies in the management of anal high-
grade squamous intraepithelial lesions. Minerva Chir. 2008;63(5):389–399.
38. Ajani JA, Winter KA, Gunderson LL, et al. Fluorouracil, mitomycin, and
radiotherapy vs fluorouracil, cisplatin, and radiotherapy for carcinoma of the anal
canal: a randomized controlled trial. JAMA. 2008;299(16):1914–1921.
39. Roohipour R, Patil S, Goodman KA, et al. Squamous-cell carcinoma of the
anal canal: predictors of treatment outcome. Dis Colon Rectum. 2008;51(2):147–
153.
40. Wexler A, Berson AM, Goldstone SE, et al. Invasive anal squamous-cell
carcinoma in the HIV-positive patient: outcome in the era of highly active
antiretroviral therapy. Dis Colon Rectum. 2008;51(1):73–81.
41. Beal KP, Wong D, Guillem JG, et al. Primary adenocarcinoma of the anus
treated with combined modality therapy. Dis Colon Rectum. 2003;46(10):1320–
1324.
42. Koulos J, Symmans F, Chumas J, Nuovo G. Human papillomavirus
detection in adenocarcinoma of the anus. Mod Pathol. 1991;4(1):58–61.
43. Belkacemi Y, Berger C, Poortmans P, et al. Management of primary anal
canal adenocarcinoma: a large retrospective study from the Rare Cancer
Network. Int J Radiat Oncol Biol Phys. 2003;56(5):1274–1283.
44. Hobbs CM, Lowry MA, Owen D, Sobin LH. Anal gland carcinoma.
Cancer. 2001;92(8):2045–2049.
45. Wong NA, Shirazi T, Hamer-Hodges DW, Corfield AP, Lessells AM.
Adenocarcinoma arising within a Crohn’s-related anorectal fistula: a form of anal
gland carcinoma? Histopathology. 2002;40(3):302–304.
46. Fenger C, Filipe MI. Mucin histochemistry of the anal canal epithelium:
studies of normal anal mucosa and mucosa adjacent to carcinoma. Histochem J.
1981;13(6):921–930.
47. Crocker HR. Paget’s disease affecting the scrotum and penis. Trans Pathol
Soc London. 1889;40:187.
48. Paget J. On disease of the mammary areola preceding cancer of the
mammary gland. St Bartholomew Hosp Rev. 1874;10:87.
49. Jones RE Jr, Austin C, Ackerman AB. Extramammary Paget’s disease: a
critical reexamination. Am J Dermatopathol. 1979;1(2):101–132.
50. Kurzl RG. Paget’s disease. Semin Dermatol. 1996;15(1):60–66.
51. Regauer S. Extramammary Paget’s disease—a proliferation of adnexal
origin? Histopathology. 2006;48(6):723–729.
52. Marucci G, Betts CM, Golouh R, Peterse JL, Foschini MP, Eusebi V. Toker
cells are probably precursors of Paget cell carcinoma: a morphological and
ultrastructural description. Virchows Arch. 2002;441(2):117–123.
53. Smith KJ, Tuur S, Corvette D, Lupton GP, Skelton HG. Cytokeratin 7
staining in mammary and extramammary Paget’s disease. Mod Pathol. 1997;
10(11):1069–1074.
54. Mazoujian G, Pinkus GS, Haagensen DE Jr. Extramammary Paget’s
disease—evidence for an apocrine origin: an immunoperoxidase study of gross
cystic disease fluid protein-15, carcinoembryonic antigen, and keratin proteins.
Am J Surg Pathol. 1984;8(1):43–50.
55. Billingsley KG, Stern LE, Lowy AM, Kahlenberg MS, Thomas CR Jr.
Uncommon anal neoplasms. Surg Oncol Clin N Am. 2004;13(2):375–388.
56. Chanda JJ. Extramammary Paget’s disease: prognosis and relationship to
internal malignancy. J Am Acad Dermatol. 1985;13(6):1009–1014.
57. Marchesa P, Fazio VW, Oliart S, Goldblum JR, Lavery IC, Milsom JW.
Long-term outcome of patients with perianal Paget’s disease. Ann Surg Oncol.
1997;4(6):475–480.
Arch Pathol Lab Med—Vol 134, November 2010
58. Goldblum JR, Hart WR. Perianal Paget’s disease: a histologic and
immunohistochemical study of 11 cases with and without associated rectal
adenocarcinoma. Am J Surg Pathol. 1998;22(2):170–179.
59. Chang GJ, Gonzalez RJ, Skibber JM, Eng C, Das P, Rodriguez-Bigas MA. A
twenty-year experience with adenocarcinoma of the anal canal. Dis Colon
Rectum. 2009;52(8):1375–1380.
60. Klas JV, Rothenberger DA, Wong WD, Madoff RD. Malignant tumors of
the anal canal: the spectrum of disease, treatment, and outcomes. Cancer. 1999;
85(8):1686–1693.
61. Fahy BN, Tang LH, Klimstra D, et al. Carcinoid of the rectum risk
stratification (CaRRS): a strategy for preoperative outcome assessment. Ann Surg
Oncol. 2007;14(2):396–404.
62. Shia J, Tang LH, Weiser MR, et al. Is nonsmall cell type high-grade
neuroendocrine carcinoma of the tubular gastrointestinal tract a distinct disease
entity? Am J Surg Pathol. 2008;32(5):719–731.
63. Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report
on cutaneous and noncutaneous melanoma: a summary of 84,836 cases from the
past decade: the American College of Surgeons Commission on Cancer and the
American Cancer Society. Cancer. 1998;83(8):1664–1678.
64. Yeh JJ, Shia J, Hwu WJ, et al. The role of abdominoperineal resection as
surgical therapy for anorectal melanoma. Ann Surg. 2006;244(6):1012–1017.
65. Curtin JA, Fridlyand J, Kageshita T, et al. Distinct sets of genetic alterations
in melanoma. N Engl J Med. 2005;353(20):2135–2147.
66. Antonescu CR, Busam KJ, Francone TD, et al. L576P KIT mutation in anal
melanomas correlates with KIT protein expression and is sensitive to specific
kinase inhibition. Int J Cancer. 2007;121(2):257–264.
67. Val-Bernal JF, Mayorga M, Diego C, Gonzalez-Vela MC. Pedunculated
polypoid lymphangioma of the anal canal. Pathol Int. 2008;58(7):442–444.
68. Kitamura Y. Gastrointestinal stromal tumors: past, present, and future. J
Gastroenterol. 2008;43(7):499–508.
69. Agaimy A, Wunsch PH. True smooth muscle neoplasms of the gastro-
intestinal tract: morphological spectrum and classification in a series of 85 cases
from a single institute. Langenbecks Arch Surg. 2007;392(1):75–81.
70. Miettinen M, Furlong M, Sarlomo-Rikala M, Burke A, Sobin LH, Lasota J.
Gastrointestinal stromal tumors, intramural leiomyomas, and leiomyosarcomas in
the rectum and anus: a clinicopathologic, immunohistochemical, and molecular
genetic study of 144 cases. Am J Surg Pathol. 2001;25(9):1121–1133.
71. Hassan I, You YN, Shyyan R, et al. Surgically managed gastrointestinal
stromal tumors: a comparative and prognostic analysis. Ann Surg Oncol. 2008;
15(1):52–59.
72. Nigri GR, Dente M, Valabrega S, et al. Gastrointestinal stromal tumor of
the anal canal: an unusual presentation. World J Surg Oncol. 2007;5:20.
73. Tsai MC, Lin JW, Lin SE, Chen HH, Lee CM, Hu TH. Prognostic analysis of
rectal stromal tumors by reference of National Institutes of Health risk categories
and immunohistochemical studies. Dis Colon Rectum. 2008;51(10):1535–1543.
74. Ranganathan V. Primary rectal Hodgkin’s lymphoma: initial manifestation
of HIV. Am J Gastroenterol. 1996;91(1):180–181.
75. Simpson L, Taylor S, Piotrowski A. Uncommon presentations of Hodgkin’s
disease, case 3: Hodgkin’s disease, rectal presentation. J Clin Oncol. 2004;22(1):
196–198.
76. Ioachim HL, Dorsett B, Cronin W, Maya M, Wahl S. Acquired
immunodeficiency syndrome-associated lymphomas: clinical, pathologic, im-
munologic, and viral characteristics of 111 cases. Hum Pathol. 1991;22(7):659–
673.
77. Sapp M, Perez-Ordonez B, Brenneman F, Imrie K, Morava-Protzner I, Lim
MS. EBV-associated perianal Hodgkin’s disease in an HIV-positive individual.
Am J Hematol. 2001;66(1):42–45.
78. Guler ML, Daniels JA, Abraham SC, Montgomery EA. Expression of
melanoma antigens in epithelioid gastrointestinal stromal tumors: a potential
diagnostic pitfall. Arch Pathol Lab Med. 2008;132(8):1302–1306.
Tumors of the Anal Canal—Shia 1611