PIIS0893395222034123
PIIS0893395222034123
PIIS0893395222034123
In situ and invasive neoplastic glandular lesions of the uterine have cytologic correlates that must be
distinguished from a variety of benign and reactive conditions. Careful study of the cytologic features allows
discrimination in the majority of cases; however, the designation of ‘atypical glandular cells’ is reserved for
equivocal cases that cannot be readily resolved. In this article, the cytologic features of endocervical
adenocarcinoma in situ and invasive endocervical adenocarcinoma will be presented, highlighting their
correlation to the well-known histologic features. Variants of the usual type of endocervical neoplasms that have
important clinical and differential diagnostic features, including mucinous adenocarcinoma and clear-cell
carcinoma, will be discussed. There are a number of common cytologic mimics of endocervical neoplasms,
including tubal metaplasia, directly sampled (abraded) endometrium, and high-grade squamous intraepithelial
lesion involving endocervical glands. The cytologic features of these entities and their differentiation from
endocervical neoplasia will be explored. Finally, the role of ancillary studies such as human papillomavirus
testing in the management of glandular lesions of the cervix will be integrated into the discussion.
Modern Pathology (2016) 29, S1–S11; doi:10.1038/modpathol.2015.138
Over the past quarter century, there has been an incidence of glandular cancers, possibly secondary
increasing recognition of glandular lesions of the to more prevalent hormone use, greater attention has
cervix. Although well described in both invasive and been paid to their histo- and cytomorphologic
in situ forms, in clinical practice these types of cancers criteria and to the specific subtypes of glandular
were considered to be relatively rare, particularly cancer that can be identified. In addition, with the
when compared with their far more prevalent advent of improved Papanicolaou sampling devices,
squamous counterparts. In the period 1976–1980, which gather greater amounts of cytologic material
adenocarcinoma accounted for only 12% of all from the upper regions of the endocervical canal, the
invasive cervical cancers.1 In situ carcinoma, while cytologic features of adenocarcinoma in situ (AIS),
and of the less prevalent variants of invasive
identifiable in histologic material, was rarely
adenocarcinoma, have been better documented.
detected in cytologic specimens. With the sharp
This review will discuss the cytologic features of
decline in cancers of squamous type, which followed the most common ‘usual’ type of endocervical
widespread implementation of cervical cytology adenocarcinoma, in conjunction with its correlative
screening, and because of a perceived increase in histomorphologic features. There are a number of
common morphologic mimics of AIS and invasive
Correspondence: Dr DC Wilbur, MD, James Homer Wright adenocarcinoma, which can lead to overinterpreta-
Laboratory of Pathology, Department of Pathology, Massachusetts tion in cytology specimens. These entities and
General Hospital, Harvard Medical School, 55 Fruit Street, Warren their differential diagnostic features when compared
120, Boston, MA 02114, USA.
E-mail: [email protected]
with true cervical glandular neoplasia will be
Received 2 September 2015; revised 16 September 2015; accepted emphasized. A number of clinically important, but
5 November 2015 relatively rare, variants of invasive adenocarcinoma,
www.modernpathology.org
In situ and invasive cervical glandular lesions
S2 DC Wilbur
including villoglandular, clear-cell and mucinous As compared with squamous lesions, endocervical
types, have important clinical management ramifica- neoplasia arises more commonly in women of
tions when compared with the usual type. These higher socioeconomic status, with higher education
variants can and should be recognized in cytology levels, lower unemployment, higher incomes, and
specimens, and their clinical and cytologic features show less linkage to early sexual intercourse and
will be discussed. Finally, laboratory testing for smoking.7 Similar to their squamous counterparts,
human papillomavirus (HPV) is now commonly the linkage of endocervical glandular neoplasia to
advocated for the detection of squamous lesions. infection by high-risk HPV is also strong. Up to 100%
As the majority of cervical glandular lesions are also of AIS and 94% of invasive lesions have been shown
associated with HPV, testing for the virus may also to be positive for HPV when examined by sensitive
serve a purpose in their management regimens. The detection techniques.8 A unique feature of
use of HPV testing, however, differs significantly for endocervical neoplasia is a greater prevalence of
glandular neoplasia as compared with their squamous HPV type 18. In all, 50–75% of endocervical tumors
counterparts, and hence it is important to illustrate are HPV type 18 positive, with the remainder being
the differences and the rationale for its use. predominantly types 16 and the 18-associated
type 45.9 In particular, HPV 16 variants
(ie, Asian-American) have been shown to be
Endocervical neoplasia: general features associated with endocervical neoplasia. It has been
postulated (but not proven) that use of hormones
Endocervical adenocarcinoma has shown a small but increases the risk of endocervical neoplasia, similar
definitive increase in incidence over the past 40 to what has been proven for endometrial carcinomas.
years.1 Reagan and Ng2 found in the early 1940s that The association noted between the Asian-American
endocervical adenocarcinoma comprised 5% of all variant of HPV type 16, which is known to have
cervical carcinomas. In recent years, this percentage increased susceptibility to estrogenic effects, and
has increased to 20–25%.3 Most recent studies have increased incidence in obese individuals provides
found incidence rates to be between 14 and 34% of additional support for this linkage.10
all cervical carcinomas.4 US SEER data gathered up A further link between endocervical and
to 2008 (Table 1) has shown that there has been a squamous carcinogenesis exists. About 50% of
slight increase in the absolute rate of endocervical endocervical tumors have a coexisting squamous
invasive carcinoma, but this rise has been lesion present.11 Such data might suggest that both
overshadowed by the decreased absolute incidence may arise from a common progenitor cell in the
of squamous carcinomas.5 In addition, international cervix, with differentiation along squamous,
databases have shown steady rates or declines in endocervical, or both lines based on such factors as
endocervical adenocarcinoma, raising questions HPV type, hormone use, and as-yet-undetermined
about the true trajectory of endocervical neoplasia cofactors.
prevalence.
Bousfield et al,6 in a large series comprising the
spectrum of endocervical neoplasia, found that age Table 2 Endocervical neoplastic categories and age6
and lesion severity were directly proportional,
similar to that which has been noted for squamous Age (years) Age range
neoplasia (Table 2). In addition, they found a
significant predilection for cervical adenocarcinomas EC ‘dysplasia’ 36 33–38
in young women, noting that 78% of cervical AIS 37 22–70
Microinvasive CA 41 28–66
carcinomas arising in the first three decades of life Invasive CA 49 26–85
were of endocervical origin.
The epidemiology of endocervical neoplasia Abbreviations: AIS, adenocarcinoma in situ; CA, cancer; EC,
differs from that of squamous neoplasia of the cervix. endocervical.
Abbreviations: ACA, adenocarcinoma; CI, confidence interval; EAPC, estimated annual percent change; N, number of cases; SCC, squamous cell
carcinoma; SEER, surveillance, epidemiology, and end results.
aData from the National Cancer Institute’s SEER Program.5
Clinically, preinvasive neoplasms of endocervical The addition of HPV testing has the potential to
origin are generally asymptomatic. They may be substantially improve detection of endocervical
visible on colposcopy if they involve the transforma- neoplasia. As noted above, virtually all of the most
tion zone, but higher in the canal they are often common types of AIS and invasive endocervical
non-visible. Even in invasive endocervical carcinomas, adenocarcinoma are HPV associated. Primary cervical
20–30% of colposcopies may be interpreted as cancer screening via the use of HPV cotesting (Pap
normal. Symptoms of abnormal uterine bleeding test plus HPV testing) or, more recently, by HPV
are most common; occasionally women have vaginal testing alone have shown more sensitive detection of
discharge or pain. In asymptomatic cases, the tumor glandular lesions.17 HPV primary screening, however,
is generally discovered via a routine cervical will not detect the rare forms of adenocarcinoma
cytology specimen; however, the sensitivity of that are not associated with HPV (eg, mucinous
detection is low in comparison with squamous carcinoma) as will be described below. For this
lesions, as would be expected by their location reason, some have advocated the use of cotesting, or
generally higher in the endocervical canal, above at least cytology at some (as-yet-unspecified) interval
the transformation zone. Australian cancer registry to achieve maximal screening benefit.
data showed that upon retrospective review of
cytology following a histologic diagnosis of AIS,
the false-negative rate was about 50%.12 Causes for Endocervical AIS
false-negative examinations were predominantly
sampling (meaning no identifiable lesion cells were AIS generally presents without symptoms. It may be
found on the slide), and also interpretation (meaning inapparent on colposcopic examination as it most
that upon review, lesional cells were present that often involves the endocervical canal. It is most often
had been missed or misinterpreted as benign or unifocal and hence continuous, although rare exam-
reactive). For invasive endocervical adenocarcinoma, ples of multifocal disease may be identified.18 It is
the false-negative rate of preceding cytologic therefore most often first detected via screening
examination was found to range from 45 to 74%, procedures, most commonly Papanicolaou testing
with about 20% being due to interpretation errors.13 or by the less specific process of HPV testing. HPV
Suffice it to say that cytology alone is significantly testing is more sensitive, but requires colposcopy
less sensitive for glandular, as opposed to squamous and biopsy to identify the disease. On histopathology
lesions, and that education and experience is specimens, AIS presents as a hyperchromatic
required to achieve best detection sensitivity. The epithelium involving both superficial and deep
cytologic criteria for the diagnosis of AIS were not glands. These hyperchromatic cells may be admixed,
well documented even as late as the first edition of with benign endocervical epithelium showing a
the Bethesda System in 1988,14 in which it was sharp transition (Figure 1). In contrast to the normal
subsumed under the category of ‘glandular atypia’. simple endocervical epithelium, AIS shows
At the time of the second edition in 2001,15 AIS was pseudostratification of enlarged hyperchromatic
specifically incorporated as a diagnostic entity
because in the ensuing years, the morphologic
criteria had been well studied and documented,
and they had been found to be reasonably well
reproduced between observers. At that time, it was
hoped that improvements in detection sensitivity
from the use of new sampling devices and enhanced
recognition and proper categorization would ensue.
Fortunately, this appears to have occurred as recent
data suggests that up to 80% of women having either
AIS or invasive endocervical adenocarcinoma will
have an abnormal cervical cytology specimen
ranging from at least ‘atypical’ (for an incomplete,
but compelling morphologic pattern) to a more
specific interpretation.13 In addition, the prevalence
of AIS in registry data is increasing at a greater rate
than are the precursor squamous lesions (sevenfold
increase in SEER data in the past 20 years).16 This
increase is almost certainly due to a compilation of
the factors noted above, including improved
sampling devices, proper sampling technique, Figure 1 Endocervical adenocarcinoma in situ: low magnification
introduction of liquid-based cytology preparation of endocervical adenocarcinoma in situ involving a single gland.
The gland in the lower right shows a hyperchromatic epithelium
allowing for improved sample preservation and in stark contrast to the normal endocervical epithelium lining the
presentation, and finally from better recognition of non-involved glands and surface. This hyperchromatic area is
the cytologic features of glandular neoplasia. often the first identifiable feature.
Figure 2 Endocervical adenocarcinoma in situ: note the pseudos- Figure 4 Normal endocervical cells: a ‘picket fence’ configuration
tratified nuclei with numerous mitotic figures present. recapitulating the histology of Figure 3.
Figure 3 Normal endocervical epithelium: this epithelium con- Figure 5 Normal endocervical cells: ‘honeycomb’ architecture for
sists of a single layer of columnar cells with basal nuclei and normal endocervical cells when viewed end-on. Note the
mucinous caps. Contrast to the pseudostratified epithelium of significant pleomorphism of the endocervical nuclei, which is a
adenocarcinoma in situ (AIS) present in Figure 2. characteristic reactive change, not to be taken in and of itself as a
feature of neoplasia.
nuclei, numerous mitotic figures (often apical), and axis (Figure 5). In the latter, nuclei are evenly
apoptosis (Figure 2). dispersed with little or no overlap, and the mucous
To understand the cytologic features of AIS, those caps of the cells are identified in focal planes above
of the normal endocervical epithelium must also be (or below) the nuclei. The appearance of normal
well appreciated. In the transformation zone, endocervical cells changes as upper portions of the
endocervical epithelium has a simple columnar endocervical canal are sampled. As the endocervical
configuration with relatively uniform nuclei present merges with the endometrial mucosa, the epithelium
at the basal portion of each cell with an overlying cap becomes pseudostratified with generally smaller
of abundant frothy mucinous cytoplasm (Figure 3). nuclei (Figure 6). In addition, tubal metaplastic
This histologic appearance translates in a cytologic change becomes more common, which can in and
specimen to groups of two configurations: (1) the of itself create differential diagnostic difficulties
linear ‘picket fence’ arrangement with columnar (discussed below).
cells lining up completely analogous to the histology The cytologic features of AIS translate directly
as presented (Figure 4); and (2) the two-dimensional from the histopathologic features noted above, and
‘honeycombed’ sheet, when abraded cells are are compiled in Table 3. Cytologic specimens
observed ‘end on’ viewing down the long columnar are cellular with increased numbers of dense
Figure 7 Endocervical adenocarcinoma in situ: (a) nuclear protrusion at the margins of the group—the so-called ‘feathering’ and (b) coarse
granularity of the chromatin and the irregularity of the nuclear envelope.
Figure 8 Endometrioid variant of endocervical adenocarcinoma: Figure 9 Endocervical adenocarcinoma, usual type: some
This entity can be difficult to identify because of the small size of features of adenocarcinoma in situ (AIS) remain with more severe
the nuclei in contrast to the usual types. Feathering is still evident. cytologic change, architectural gland formations, and stromal
These cells are often misinterpreted as being from benign reaction.
endometrial sampling or exfoliation.
Figure 11 Villoglandular variant of endocervical adenocarcinoma: Figure 13 Lobular endocervical hyperplasia: this is a precursor
this designation is reserved for grade 1 lesions with single layers of lesion to mucinous (gastric-type) adenocarcinoma of the
abnormal epithelium covering stromal papillae. endocervix. Abundant complex mucinous epithelial cells are
arranged in a lobular pattern without significant cytologic atypia
are key features of this entity.
Figure 12 Villoglandular variant of endocervical adenocarcinoma: Figure 14 Mucinous (gastric type) endocervical adenocarcinoma:
the cytology shows slender papillary groups consisting of ‘grade 1’ the malignant counterpart of lobular endocervical hyperplasia.
neoplastic cells. Atypical mucinous epithelial cells are arranged in small- to
medium-sized glandular units penetrating the cervical wall with
little to no stromal reaction.
HIK1083), and commonly shows STK11 mutations presence of large numbers of endocervical
(chromosome 19p associated with Peutz–Jeghers mucinous-type cells should prompt a closer
syndrome).22 Mucinous carcinoma has a known examination for cytologic atypia (nuclear enlargement
noninvasive precursor lesion, lobular endocervical and irregularity), which is invariably present in some
glandular hyperplasia (Figure 13). In histologic cells. In addition, the pyloric mucin will show a
specimens, mucinous carcinoma commonly presents golden-brown appearance in the Papanicolaou stain
as tall mucin-producing glands penetrating the and may hence be very informative if present
cervical wall with minimal if any stromal reaction (Figure 15). Primary clear-cell carcinoma in the
and minimal cytologic atypia, hence the synonyms endocervix is rare. Its histology is similar to other
‘adenoma malignum’ or ‘minimal-deviation sites in the female genital tract. The far more
adenocarcinoma’ (Figure 14). Cytologically, common entity of endocervical microglandular
mucinous carcinomas are difficult to identify hyperplasia can be a good mimic of clear-cell
because of their morphologic similarity to benign carcinoma. Both entities have closely packed glands,
mucinous endocervical epithelium. However, the cystic glands, and infiltrating inflammatory cells.
However, clinically they have very different reliable cytologic or histologic identification is not
presentations: clear-cell carcinoma in the older possible because of the wide range of non-neoplastic
age group with symptoms and a mass, and conditions that can present in the endocervical
microglandular hyperplasia in young reproductive canal. Hence, the 2014 Bethesda System for cytologic
age patients who are asymptomatic and have no interpretation prefers the use of the term ‘atypical
mass, except occasionally a benign-appearing glandular cells’ for cytologic presentations, which
endocervical polyp. Clear-cell carcinoma shows meet some but not all of the criteria for AIS.24
high-grade nuclear atypia, whereas microglandular Some cases of atypical glandular cells will have
hyperplasia shows bland nuclei with characteristic follow-up showing true endocervical neoplasia,
subnuclear vacuoles (Figure 16). but a significant number will show either squamous
Some investigators have suggested that neoplastic neoplasia or benign reactive outcomes. The ability
lesions less than AIS can be detected, and have to distinguish between neoplastic and benign
suggested a designation of ‘endocervical dysplasia’ presentations can be difficult and hence further
or ‘low-grade endocervical neoplasia’.23 Although clinical evaluation is recommended for all such
most agree that such ‘dysplastic’ lesions exist, instances.
Figure 15 Mucinous (gastric type) endocervical adenocarci- Figure 17 High-grade squamous intraepithelial lesion: when
noma: the cells contain abundant frothy mucinous epithelium high-grade squamous intraepithelial lesion (HSIL) involves
and are arranged in groups or strips. Nuclear atypia is endocervical glands, its appearance in cytologic preparations can
prominent. often show similarities to glandular lesions (see Figure 18).
Figure 16 (a) Clear-cell carcinoma: marked cytologic atypia in densely packed glands with cleared or eosinophilic cytoplasm.
(b) Microglandular hyperplasia: densely packed glands can show papillary structures with little nuclear atypia. The diffuse basal vacuoles
are prominent.
Cytologic mimics of endocervical metaplasia can coexist with AIS, hence the presence
neoplasia of cilia in some cells does not preclude that other
atypical cells in the specimen may be neoplastic.
There are a number of important mimics of Key to the differential between tubal metaplasia and
endocervical neoplasia found in cervical cytologic AIS is examination of the chromatin pattern. Fine
specimens. These include high-grade squamous granularity favors tubal metaplasia, whereas coarse
intraepithelial lesions (HSILs), tubal metaplasia, granularity favors AIS (Figure 20).
and cervical endometriosis/directly sampled Direct sampling of endometrial tissue, be it from
endometrium (Table 3). cervical endometriosis or lower portions of the
When HSIL involves the endocervical gland necks uterine corpus, can create difficulty in cytologic
(Figure 17), the cells sampled may show some differential diagnosis. Abraded endometrium can
features of endocervical neoplasia, including present in densely packed glandular groups, which
densely packed hyperchromatic groups, palisading can show pseudostratification, marked hyperchromasia,
of cells, and pseudocolumnar configurations. Common and coarse chromatin granularity. If the endometrial
features also include mitoses, apoptotic debris, and tissue is cycling or disordered, it can be mitotically
nuclear atypia. Important differential diagnostic active and rarely show apoptosis (Figure 21). The
features include lack of typical architectural of AIS presence of large organoid groups of cells forming
described above, meaning no rosettes, feathering, or tubules, isolated groups of endometrial stromal cells,
pseudostratified strips. In addition, if HSIL in a gland or stromal cells attached to the edges of the organoid
is suspected, review of the background for classic groups is key to this differential diagnosis. Stromal
isolated HSIL cells is warranted—their presence cells attached to organoid groups can mimic
makes HSIL far more likely as a cause for the dense feathering, but close examination will show the
groups identified (Figure 18). mesenchymal nature of these cells with wispy
Tubal metaplasia is a benign metaplastic process cytoplasm and small nuclear size. Overall, nuclei of
very common to both the endometrial and upper directly sampled endometrium will be much smaller
than those of AIS and attention to this particular
endocervical epithelium. Classic studies have shown
detail should allow their distinction (Figure 22).
that the newer sampling devices used in cervical
cytology (broom and brush devices) sample many
cells from the upper endocervical canal.25 Tubal
Conclusions
metaplasia recapitulates the epithelium of the
normal fallopian tube and is comprised of ciliated Endocervical neoplasia is increasing relatively,
columnar cells, mucinous peg cells, and slender and also in absolute incidence, when compared
intercalated cells. This combination can lead to with squamous neoplasia. Endocervical lesions
dense hyperchromatic groups, with pseudostratified are less likely to show symptoms or be colposcopically
architecture (Figure 19). Mitoses can be present, but evident. Hence, cytology and primary HPV testing
apoptotic debris is not typically seen. Cilia are an have a very important role in their detection.
important clue to its benign nature. However, tubal At present, all glandular atypias and definitive
Figure 20 (a) Tubal metaplasia: there are several mitotic figures present (arrows). (b) Tubal metaplasia: hyperchromatic crowded group
with mitotic figures (arrows). As in Figure 19, the chromatin pattern of the nuclei is smooth and evenly distributed in contrast to
adenocarcinoma in situ (AIS).
Figure 21 Directly sampled endometrium: mitotic figures Figure 22 Directly sampled endometrium: a sharply defined
are noted in the proliferative phase (arrow). Note the small hollow organoid tube of endometrial epithelial cells is present in
size of the nuclei and the presence of endometrial stromal the top of the field. A large fragment of spindled endometrial
fragments. stromal cells is noted at the bottom of the field. The presence of
both elements is useful in differentiating these cells from
adenocarcinoma in situ (AIS).
Compendium of Diagnostic Cytology, 6th (edn). 14 Kurman RJ, Solomon D. The Bethesda System for
Chicago, IL, USA: Tutorials of Clinical Cytology, 1988, Reporting Cervical/Vaginal Cytologic Diagnoses.
pp 146–153. New York, NY, USA: Springer, 1994.
3 Adegoke O, Kulasingam S, Virnig B. Cervical cancer 15 Solomon D, Nayar R. The Bethesda System for
trends in the United States: a 35-year population-based Reporting Cervical/Vaginal Cytologic Diagnoses, 2nd
analysis. J Womens Health 2012;21:1031–1037. (edn). New York, NY, USA: Springer, 2003.
4 Vizcaino AP, Moreno V, Bosch FX et al. International 16 Ward KK, Shah NR, Saenz CC et al. Changing
trends in the incidence of cervical cancer: I. Adeno- demographics of cervical cancer in the United States
carcinoma and adenosquamous cell carcinomas. Int J (1973–2008). Gynecol Oncol 2012;126:330–333.
Cancer 1998;75:536–545. 17 Hildesheim A, Berrington de Gonzalez A. Etiology and
5 Tambouret RT, Wilbur DC. Glandular Lesions of the prevention of cervical adenocarcinomas. J Natl Cancer
Uterine Cervix. Cytopathology with Histologic Inst 2006;98:292–293.
Correlates. New York, NY, USA: Springer, 2015. 18 McCluggage WG. New developments in endocervical
6 Bousfield L, Pacey F, Young Q et al. Expanded glandular lesions. Histopathology 2013;62:138–160.
cytologic criteria for the diagnosis of adenocarcinoma 19 Howitt BE, Herfs M, Brister K et al. Intestinal-type
in situ of the cervix and related lesions. Acta Cytol endocervical adenocarcinoma in situ: an immuno-
1980;24:284–296. phenotypically distinct subset of AIS affecting older
7 Korhonen MO. Epidemiological differences between women. Am J Surg Pathol 2013;37:625–633.
adenocarcinoma and squamous cell carcinoma of the 20 Wilbur DC, Colgan TJ, Ferenczy AS et al. Chapter 7.
uterine cervix. Gynecol Oncol 1980;10:312–317. Glandular tumours and precursors, part of tumours
8 Zielinski GD, Snijders PJ, Rozendaal L et al. The of the uterine cervix. In: Kurman RJ, Carcangiu ML,
presence of high-risk HPV combined with specific Herrington CS et al. (eds). WHO Classification of
p53 and p16INK4a expression patterns points to Tumours of Female Reproductive Organs, 4th (edn).
high-risk HPV as the main causative agent for Lyon, France: IARC, 2014, pp 183–194.
adenocarcinoma in situ and adenocarcinoma of 21 Hagiwara T, Kaku T, Kobayashi H et al. Well-differen-
the cervix. J Pathol 2003;201:535–543. tiated villoglandular adenocarcinoma of the uterine
9 Clifford G, Franceschi S. Members of the human cervix: assessment of cytological features by histo-
papillomavirus type 18 family (alpha-7 species) share logical subtypes. Acta Cytol 2013;57:61–68.
a common association with adenocarcinoma of 22 Kusanagi Y, Kojima A, Mikami Y et al. Absence of
the cervix. Int J Cancer 2008;122:1684–1685. high- risk human papillomavirus (HPV) detection in
10 Tornesello ML, Losito S, Benincasa G et al. Human endocervical adenocarcinoma with gastric morphology
papillomavirus (HPV) genotypes and HPV16 variants and phenotype. Am J Pathol 2010;177:2169–2175.
and risk of adenocarcinoma and squamous cell 23 Mathers ME, Johnson SJ, Wadehra V. How predictive is
carcinoma of the cervix. Gynecol Oncol 2011;121: a cervical smear suggesting glandular neoplasia?
32–42. Cytopathology 2002;13:83–91.
11 Ayer B, Pacey F, Greenberg M et al. The cytologic 24 Nayar R, Wilbur DC. The Bethesda System for Report-
diagnosis of adenocarcinoma in situ of the cervix uteri ing Cervical/Vaginal Cytologic Diagnoses, 3rd (edn).
and related lesions. I. Adenocarcinoma in situ. Acta New York, NY, USA: Springer, 2015.
Cytol 1987;31:397–411. 25 Babkowski RC, Wilbur DC, Rutkowski MA et al. The
12 Ruba S, Schoolland M, Allpress S et al. Adenocarcinoma effects of endocervical canal topography, tubal meta-
in situ of the uterine cervix: screening and diagnostic plasia, and high canal sampling on the cytologic
errors in Papanicolaou smears. Cancer 2004;102: presentation of non-neoplastic endocervical cells.
280–287. Am J Clin Pathol 1996;105:403–410.
13 Krane JF, Granter SR, Trask CE et al. Papanicolaou 26 Massad S, Einstein MH, Huh WK et al. 2012 Updated
smear sensitivity for the detection of adenocarcinoma consensus guidelines for the management of abnormal
of the cervix. A study of 49 cases. Cancer (Cancer cervical cancer screening tests and cancer precursors.
Cytopathol) 2001;93:8–15. J Low Genit Tract Dis 2013;17:S1–S27.