livrocitopapanicolaulaudo

Prévia do material em texto

Negative for Intraepithelial Lesion or Malignancy (NILM)
endometrial Cells
The Bethesda System: Historical 
Perspective
Terminology forms the basis for effective communication 
between the laboratory and clinician. The clinician is expected 
to provide relevant patient information to the laboratory. It is 
the laboratory’s responsibility to report results using terminol-
ogy that clearly conveys the diagnostic interpretation of the mor-
phologic findings. The use of a uniform diagnostic terminology 
facilitates communication by establishing a common language 
that, in theory, does not vary significantly from cytologist to 
cytologist or laboratory to laboratory. However, terminology is 
not static over time; rather, it evolves in parallel with increased 
understanding of the pathogenesis and biology of disease. The 
framework, therefore, must be flexible enough to incorporate 
advances in scientific knowledge without creating undue confu-
sion or complexity.
In 1988, the National Cancer Institute sponsored an open 
workshop—including cytotechnologists, pathologists, clini-
cians, and representatives of professional organizations—to 
develop a uniform descriptive terminology for cervical/vagi-
nal cytologic interpretation. The format that emerged became 
modifications were incorporated into the 1991 Bethesda System 
that streamlined the terminology.2 In addition, an ad hoc com-
mittee developed criteria for specimen adequacy and Bethesda 
interpretive categories, culminating in the first TBS atlas that 
outlined and illustrated the morphologic features.3 By the mid- 
to late 1990s, there was a significant penetration of the Bethesda 
System into cytopathology practice with approximately 90% of 
laboratories in the United States using the Bethesda terminol-
ogy for reporting of cervical/vaginal cytology.4
Among all the changes introduced by the implementation 
of TBS terminology into practice, none was more controversial 
than the category of atypical squamous cells (ASCUS). At that 
time, the majority of abnormal Pap tests reported annually in 
the United States, approximately 2.5 million, were interpreted 
as ASCUS and had highly variable management at considerable 
cost to the healthcare system. Another 1.2 million were inter-
preted as low-grade squamous intraepithelial lesion (LSIL).5 In 
an effort to determine the best management strategy (effective 
as well as cost-effective) for women with these equivocal and 
low-grade abnormalities, the National Cancer Institute (NCI) 
sponsored the ASCUS/LSIL Triage Study (ALTS), which was 
completed in 2001.6 This study, as detailed below, has allowed 
S e c t i o n A
Female Genital Tract
The Bethes
Contents
The Bethesda System: Historical Perspective
the 2001 Bethesda System
Report Format
Specimen Adequacy
Bethesda 2001 Specimen adequacy Categories
Squamous Cellularity
Quality Indicators
Management Guidelines
Impact on Laboratory practice
General Categorization
Interpretation/Result
known as The Bethesda System (TBS).1
Approximately two years later, a second meeting was con-
vened to critique and refine the terminology based on experience 
with the use of the system in actual laboratory practice. Minor 
C h a p t e r
da System for Reporting 
Cervical Cytology
ritu Nayar, David C Wilbur and Diane Solomon
epithelial Cell abnormalities: Squamous Cell
epithelial Cell abnormalities: Glandular Cell
Educational Notes/Suggestions
Ancillary Testing
Automated Review
Interobserver Reproducibility in Cervical Cytology
The Bethesda System and Reporting Anal-Rectal Cytology
Concluding Remarks
6
77
for a data-driven approach to management of these prevalent 
cervical cytologic abnormalities.
From its inception, the fundamental aim of the Bethesda 
 System has been to communicate clinically relevant information 
78
Diagnostic CytologyPART TWO
from the laboratory to the patient’s healthcare provider, using 
uniform, reasonably reproducible terminology which reflects 
the most current understanding of the biology of cervical neo-
plasia. Advances in the understanding of the biology of cervical 
cancer, results from clinical trials, the introduction of liquid-
based cytology, human papillomavirus (HPV) testing, and auto-
mated screening devices for cervical cytology led to the decision 
to convene the third Bethesda workshop in April 2001.
the 2001 Bethesda System
TBS 2001 Process
Approximately eight months prior to the workshop, nine 
forum groups consisting of 6 to 10 individuals with a 
breadth of expertise in the area of cervical cancer, were organ-
ized under the sponsorship of the NCI to formulate draft 
 recommendations. Internet bulletin boards were open to the 
worldwide cytology community for six months during the 
pre-conference process of review and discussion. Over 1000 
comments were considered in revising the pre-workshop 
drafts. The 2001 Bethesda workshop was co-sponsored by 
44 international professional organizations and attended by 
over 400 individuals, including pathologists, cytotechnolo-
gists, gynecologists, attorneys, patient advocates, and other 
healthcare workers involved in women’s health initiatives. 
The revised draft recommendations were presented by each 
forum group and after open discussions and voting by all 
participants, the 2001 Bethesda consensus terminology was 
finalized and published in 2002.7
Following the Bethesda workshop, the American Society for 
Colposcopy and Cervical Pathology (ASCCP) held a compara-
ble consensus workshop on patient management in September 
2001. This was also preceded by an Internet discussion, and 
resulted in the development of evidence-based management 
guidelines for abnormal cervical cytology corresponding to 
the 2001 Bethesda reporting format.8 The ASCCP management 
guidelines were subsequently updated at a consensus confer-
ence held in September 2006.9
After the initial publication of the Bethesda System 2001 ter-
minology (Table 6.1), the NCI approached the American Society 
of Cytopathology (ASC) to collaborate on publication of the 
second edition of the Bethesda Atlas10 and the development 
of an accompanying Bethesda System educational website.11 
Images chosen for the atlas and website underwent an exten-
sive selection/validation process, and included classic as well as 
morphologically difficult and “borderline” images, illustrated 
on both conventional and liquid-based preparations. A subset 
of images chosen for the Bethesda atlas were used to assess inter-
observer reproducibility in gynecologic cytology—the details of 
this Bethesda interobserver reproducibility project are described 
below.12
Report Format
The basic structure of TBS includes three elements, based on 
communication needs germane, but not limited, to gyneco-
logic cytology: (1) statement of specimen adequacy, (2) general 
categorization, and (3) descriptive terminology. The specimen 
type—conventional smear, liquid-based preparation, or other—
should also be stated in the report (Table 6.1).
Specimen Adequacy
Reporting of adequacy was an important quality assur-
ance measure introduced by the Bethesda System. The 1988 
Bethesda System incorporated a classification of three catego-
ries of specimen adequacy—satisfactory, less than optimal, 
and unsatisfactory—into the format of the report but did not 
outline specific morphologic criteria for evaluation of ade-
quacy. Participants at the 1991 Second Workshop, and oth-
ers in the cytopathology community,13,14 voiced the need for 
developing consensus guidelines. In response, following the 
second workshop, a Criteria Committee formulated the defi-
nitions for adequacy based on a combination of experience 
and review of an admittedly sparse scientific database. Three 
categories—“satisfactory,” “satisfactory but limited by…,” and“unsatisfactory”—based on estimates of overall squamous cel-
lularity, assessment of the transformation zone component, 
and the presence/extent of obscuring or limiting factors, were 
suggested in an initial attempt to develop a more standard-
ized approach to the evaluation of adequacy.15 It was empha-
sized that the indicated percentages should be used as general 
ranges, not strict numerical cutoffs and that patient-related 
clinical factors and previous cytologic findings should always 
be taken into consideration.
Bethesda 2001 Specimen Adequacy categories
In 2001, substantial changes were made to the adequacy compo-
nent of TBS. The previously used borderline adequacy category 
of “less than optimal” (1988)/“satisfactory but limited by…” 
(1991) was deleted in order to provide the clinician a clearer 
and more reproducible indication of the adequacy of the speci-
men.16 The classification recommended in TBS 2001 is either as 
“satisfactory” or “unsatisfactory”:
 • Satisfactory10—Satisfactory for evaluation (describe 
presence or absence of endocervical/transformation 
zone component and any other quality indicators, 
e.g., partially obscuring blood, inflammation, etc.). 
For “satisfactory” specimens, including information 
on transformation zone sampling and other adequacy 
qualifiers (obscuring elements, poor preservation, etc.) 
encourages specimen takers to pay greater 
attention to specimen procurement and handling. 
Any factors that compromise specimen quality can be 
mentioned in a note.
 • Unsatisfactory10—For unsatisfactory specimens the 
report should indicate whether the laboratory 
processed/evaluated the slide. Suggested wording is:
 – Rejected specimen—Specimen rejected/not 
 processed because (specify reason: specimen not 
labeled, broken slide, etc.).
 – Fully evaluated, unsatisfactory specimen—Speci-
men processed and examined, but unsatisfactory 
for evaluation of epithelial abnormality because 
of (specify reason: inadequate squamous component, 
obscuring blood, etc.).
Additional comments/recommendations may be made as 
deemed appropriate.
While unsatisfactory specimens which are processed and 
evaluated are not suitable for excluding an intraepithelial lesion 
the Bethesda System for Reporting cervical cytology
6
Table 6.1 the 2001 Bethesda System
SPeCimen TyPe
Indicate conventional smear vs liquid-based preparation vs other
AdequACy oF THe SPeCimen
 • Satisfactory for evaluation (describe presence or absence of endocervical/transformation zone component and any other 
quality indicators, e.g., partially obscuring blood, inflammation)
 • Unsatisfactory for evaluation … (Specify reason)
– Specimen rejected/not processed (specify reason)
– Specimen processed and examined, but unsatisfactory for evaluation of epithelial abnormality because of (specify reason)
GeneRAl CATeGoRizATion (oPTionAl)
 • Negative for intraepithelial lesion or malignancy
 • Other
 • epithelial cell abnormality: see Interpretation/result (specify squamous or glandular as appropriate)
inTeRPReTATion/ReSulT
Negative for intraepithelial lesion or malignancy
(when there is no cellular evidence of neoplasia, state this in the General Categorization above and/or in the Interpretation/result 
section of the report—whether or not there are organisms or other non-neoplastic findings)
OrGaNISMS
 • Trichomonas vaginalis
 • Fungal organisms morphologically consistent with Candida spp.
 • Shift in flora suggestive of bacterial vaginosis
 • Bacteria morphologically consistent with Actinomyces spp.
 • Cellular changes associated with herpes simplex virus
Other NON-NeOpLaStIC FINDINGS (OptIONaL tO repOrt; LISt NOt INCLUSIve)
 • reactive cellular changes associated with:
– Inflammation (includes typical repair)
– radiation
– Intrauterine contraceptive device (IUD)
 • Glandular cells status posthysterectomy
 • atrophy
Other
 • endometrial cells (in a woman > 40 years of age) (specify if “negative for squamous intraepithelial lesion”)
Epithelial cell abnormalities
SQUaMOUS CeLL
 • atypical squamous cells
– of undetermined significance (aSC-US)
– cannot exclude hSIL (aSC-h)
 • Low-grade squamous intraepithelial lesion (LSIL) (encompassing: hpv†/mild dysplasia/CIN 1)
 • high-grade squamous intraepithelial lesion (hSIL) (encompassing: moderate and severe dysplasia, CIS/CIN 2 and CIN 3)
– with features suspicious for invasion
 • Squamous cell carcinoma
GLaNDULar CeLL
 • atypical
– endocervical cells (NOS or specify in comments)
– endometrial cells (NOS or specify in comments)
– glandular cells (NOS or specify in comments)
 • atypical
– endocervical cells, favor neoplastic
– Glandular cells, favor neoplastic
 • endocervical adenocarcinoma in situ
 • adenocarcinoma
– endocervical
– endometrial
– extrauterine adenocarcinoma
– not otherwise specified (NOS)
Other malignant neoplasms (specify)
(continued)
79
80
Diagnostic CytologyPART TWO
along a diameter that includes the center of the 
preparation. The average number of squamous cells 
per field is thus estimated. One preliminary study 
constructive feedback provided in the written report, by tel-
ephone, or in a summary format comparing adequacy rates of 
n may increase the 
n by directing atten-
 otherwise uncertain 
nd date of last men-
nce of this informa-
ation; therefore, the 
atisfactory” in these 
suggested that LBPs containing 5,000–20,000 
 squamous cells should be considered as “borderline” 
cellularity.18
The reader is referred to the Bethesda atlas for cellularity 
tables and figures.10 These guidelines may change in the future 
based on additional data.
The TBS numeric criteria for cellularity may not be appli-
cable to vaginal specimens, cases with extensive cytolysis, cell 
clustering, and some cases of atrophy. Cytologists should utilize 
clinical information and their best judgment when interpreting 
peer group clinicians.
Clinical information
Providing pertinent clinical informatio
sensitivity and reliability of the evaluatio
tion to a clinical question or by clarifying
cytologic findings. At a minimum, age a
strual period should be provided. Abse
tion does not, however, preclude evalu
specimen may remain categorized as “s
circumstances.
or malignancy, the presence of endometrial cells in a women 40 
years or older, or the presence of organisms, can be reported in 
this context, since this information may prove to be clinically 
relevant for patient management.
As in prior Bethesda adequacy guidelines, if abnormal 
cells are detected, the specimen cannot be categorized as 
“unsatisfactory.”
Squamous cellularity
TBS 1991 required that well-preserved and well-visualized squa-
mous epithelial cells should cover more than 10% of the slide 
surface. In order to address adequacy on conventional as well as 
liquid-based preparations, and to improve interobserver repro-
ducibility, TBS 2001 went further to provide numerical estimates 
of what constitutes adequacy for squamous cellularity in cervi-
cal cytology preparations.
 • Conventional smears: An adequate conventional prepa-
ration should have a minimum of approximately 
8,000–12,000 well-preserved and well-visualized 
squamous cells. This minimum cell count should 
be estimated, not counted. The count includes both 
 nucleated mature and metaplastic squamous cells. The 
percentage of hypocellular areas, if present, should be 
estimated and the fields counted should reflect this 
proportion. The Bethesda atlas10 and website11 provide 
“reference images” of known cellularity at low (4×) 
magnification as a resource for cytologists to compare 
with the specimen being assessed.
 • Liquid-based preparations (LBP): An adequate LBP 
should have an estimated minimum of at least 5,000 
well-visualized, well-preservedsquamous cells.10,17 
 Estimation of cellularity is suggested in borderline 
cases by performing representative field counts. 
A minimum of 10 fields, usually at 40×, are assessed 
AnCillARy TeSTinG
provide a brief description of the test method(s) and report result so that it is eas
AuTomATed Review
If case is examined by automated device, specify device and result.
eduCATionAl noTeS And SuGGeSTionS (oPTionAl)
Suggestions should be concise and consistent with follow-up guidelines publish
relevant publications may be included).
Table 6.1 the 2001 Bethesda System—cont’d
such cases. At present there are no published studies specifically 
addressing the relationship between low cellularity and false-
negative rate.
Quality indicators
Patient/Specimen identification and Technical 
interpretability
Correct specimen identification is essential for evaluation and is 
required by the Clinical Laboratory Improvement Amendments 
of 1988 (CLIA ‘88). In addition to ensuring that the specimen 
corresponds to the correct patient, proper identification allows 
the laboratory to locate prior records and slides from the patient 
that may influence the current evaluation.
The cellular material must be well fixed and unobscured 
for interpretation. Minimal data regarding how obscuring 
factors affect the interpretive reliability of a cervical speci-
men are available. In order to be considered obscuring, the 
epithelial cell morphology must be uninterpretable. For 
example, although most cervical samples contain inflamma-
tory cells, moderate numbers do not generally obscure the 
nuclei of squamous cells. Even a large amount of inflamma-
tion or blood may be acceptable if it is spread thinly such 
that the intermixed epithelial cells can be easily visualized. 
In general, specimens with more than 75% of epithelial cells 
obscured are considered unsatisfactory. A variety of factors 
may compromise visualization of the cells. Heavy inflamma-
tion, blood, and extensive cytolysis are patient-related and 
independent of the sample taker. However, presence of air-
drying, thick uneven smears, or lubricant is often inversely 
correlated with the skill and experience of the clinician. 
With liquid-based preparations a number of these factors 
become less significant; however, appropriate collection/rins-
ing techniques need to be utilized. Clinicians who repeatedly 
obtain technically poor quality specimens may benefit from 
ily understood by the clinician.
ed by professional organizations (references to 
the Bethesda System for Reporting cervical cytology
6
Sampling of the Transformation zone
Presence of endocervical or squamous metaplastic cells forms the 
microscopic basis for the assumption that the transformation zone 
has been sampled. The numeric criterion for a transformation zone 
component, at least 10 well preserved endocervical or squamous 
metaplastic cells, did not change from TBS 1991; however, due 
to the widespread utilization of liquid-based preparations, single 
endocervical/metaplastic cells are acceptable and clusters are no 
longer required. This definition applies to specimens from both 
premenopausal and postmenopausal women having a cervix. In 
the situation of marked atrophy, where metaplastic and endocer-
vical cells often cannot be distinguished from parabasal cells, the 
laboratory has the option of making a comment regarding the diffi-
culty in assessing the transformation zone component. Patient fac-
tors, such as location of the transformation zone, age, pregnancy, 
and previous therapy, may limit the clinician’s ability to obtain an 
endocervical sample despite optimal collection technique.
Numerous cross-sectional studies have demonstrated that 
smears with endocervical and/or metaplastic cells have a signifi-
cantly higher frequency and higher grade of squamous epithe-
lial abnormality detected than do smears without such cells.19–22 
Paradoxically, short-term longitudinal studies of women whose 
initial negative smears lacked an endocervical component have 
shown no increase in abnormalities on repeat, satisfactory 
smears (as might be expected if the initial smears had a higher 
false-negative rate).23,24
Based on the above studies, TBS 2001 does not require the 
presence of a transformation zone component to categorize a 
specimen as satisfactory—adequate squamous cellularity is the 
only criterion. The absence of a transformation zone compo-
nent is considered to be a quality indicator. With the reported 
increase in endocervical carcinoma,25,26 the importance of the 
transformation zone component may undergo further evalu-
ation in the future, in order to ensure optimized screening 
 performance in the setting of endocervical neoplasia.
Management Guidelines
The ASCCP has published management guidelines for Pap test 
specimen adequacy and quality indicators.9 The preferred man-
agement for unsatisfactory Pap tests is a repeat test within a short 
interval of 2–4 months. Unsatisfactory cases are unreliable for 
detection of an epithelial abnormality; furthermore a longitu-
dinal study found that unsatisfactory Pap tests are more often 
from high-risk patients and have significantly more SIL/cancer 
on follow-up than patients with satisfactory index Paps.27 The 
guidelines suggest a repeat Pap test in 12 months for most women 
who lack a transformation zone component or whose cytology 
is partially obscured unless there is a history of prior adequate/
negative Pap tests. Indications for considering earlier repeats are 
also outlined and depend on additional patient risk factors.28 
For quality assurance, it may be prudent to have the patholo-
gist review unsatisfactory cases prior to final sign-out because of 
the clinical implications of such a report and the association of 
obscuring blood/inflammation with invasive cancers.29
impact on Laboratory Practice
The incorporation of specimen adequacy as an integral part 
of the cervical cytology report has been acknowledged as one 
of the most important contributions of TBS. The impact on 
 laboratory practice has been dramatic. Surveys conducted by 
CAP revealed that in 1990 only 35% of responding labora-
tories routinely reported specimen adequacy; by 1992, this 
figure increased to 85%.30 A 1991 CAP survey found that 
most laboratories reported unsatisfactory specimen rates of 
0.5–1.0%. By the year 2003, a CAP survey assessing Bethesda 
implementation and reporting rates31 found that 73.6% of 
responding laboratories had eliminated the use of the “sat-
isfactory but limited by…” category and the 2001 TBS mini-
mum squamous cellularity criteria had been adopted by 
85.3%. Experts had predicted an increase in unsatisfactory 
rates with the use of TBS 2001 criteria. While some studies 
have indeed reported this (up to a tenfold increase on con-
ventional smears32), the CAP survey31 and other reports from 
the United States33 and Europe34 did not show an increase 
in the unsatisfactory rate after conversion to TBS 2001 ade-
quacy criteria. Possibilities suggested by the authors include 
improved sampling and preparation methods, related pre-
dominantly to liquid-based methodology, or, alternatively, 
lack of attention to the TBS 2001 criteria.31
General Categorization
The general categorization is a clerical device to aid clinicians and 
their office staff in triaging patients/prioritizing cases for review 
and to assist laboratories in compiling statistical information.
There are 3 headings used under the general category:
 1. “Negative for intraepithelial lesion or malignancy” 
for specimens in which an epithelial abnormality is 
not identified. Organisms and other benign/reactive 
cellular changes can be reported in the Interpretation 
under this category.
 2. “Other” may be utilized for cases in which there is 
no clear cytologic abnormalitybut the findings may 
warrant follow-up/investigation based on patient risk, 
for example endometrial cells in a woman 40 years of 
age or older.
 3. “Epithelial cell abnormality” may be utilized for squa-
mous or glandular epithelial abnormalities. Specify as 
far as possible which type is present.
If more than one diagnostic entity is present—for exam-
ple, an infectious process and an epithelial abnormality—the 
specimen should be categorized according to the most clini-
cally significant lesion; in this example, epithelial cell abnor-
mality. However, the general category should not replace 
narrative (descriptive) terminology for communicating the 
interpretation/result. Some laboratories also extend the con-
cept of a general or summary categorization to nongynecologic 
 specimens.
interpretation/Result
The prior Bethesda system use of the term “diagnosis” was 
replaced by “interpretation/result” in 2001. Workshop partici-
pants felt that cervical cytology is a screening, not a diagnostic 
test, that provides the clinician with information on morpho-
logic findings that need to be integrated with the patient’s 
other clinical findings for a final diagnosis and subsequent 
 management.7
81
82
Diagnostic CytologyPART TWO
negative for intraepithelial Lesion or Malignancy 
(niLM)
The Bethesda 2001 category of NILM is used to report non-
neoplastic findings in the absence of an intraepithelial lesion 
or malignancy. This term is used both as a general categoriza-
tion and as an interpretation and incorporates the reporting 
of organisms and other non-neoplastic findings such as reactive 
cellular changes (Table 6.1). The NILM category replaces the two 
prior Bethesda categories of “within normal limits” (WNL) and 
“benign cellular changes” (BCC). The basis of this change was 
to clearly communicate to the physician that despite any other 
“benign” changes reported, the Pap test is “negative” or without 
evidence of cervical intraepithelial neoplasia or malignancy.
Clearly, the main purpose of cervical cytology screening is the 
detection of cervical squamous cell carcinoma and its precur-
sors; however, reporting the findings of organisms or reactive 
conditions can make an important contribution to patient care. 
This documentation can facilitate patient triage, provide clinical-
cytologic correlation, and focus attention on cytomorphologic 
criteria during microscopic screening and interpretation of 
 cervical cytology.
The category of “infections” was changed to “organisms” 
in TBS 2001 since the presence of some organisms represents 
colonization rather than a clinically significant infection. Excel-
lent specificity and reproducibility can be achieved for the 
cytopathologic interpretation of fungal elements, Trichomonas 
vaginalis, Actinomyces, and herpes simplex virus, by application 
of reproducible morphologic criteria. The interpretation of 
Chlamydia spp. is not listed in TBS because of the acknowledged 
low diagnostic accuracy of routine cytology for this organism 
and because of the availability of other, more accurate detec-
tion methods. TBS lists the organisms that should be reported; 
however, the laboratory is advised to discuss the relevance of 
reporting organisms and other non-neoplastic findings with 
their clinicians and come to a decision about what to report 
under the NILM category.
Cells manifest reactive morphologic changes in response to 
a variety of traumatic insults such as infection, inflammation, 
and radiation. Reparative processes, radiation, atrophy, and intra-
uterine contraceptive devices are examples of entities that induce 
cellular changes that may mimic intraepithelial lesions or even 
cancer. Severe reactive/reparative changes are difficult to dis-
tinguish from neoplastic changes and such interpretations are 
well known to have lower reproducibility than classic repair.35 
It is, however, important to recognize benign reactive features 
in order to avoid overinterpretation and resulting false-positive 
interpretations. A CAP report indicates that reparative changes 
tend to be easier to recognize on LBP, yielding less false positives 
than on conventional smears.36
Keratotic cellular changes—hyperkeratosis, parakeratosis, 
and dyskeratosis—are descriptive terms that do not clearly 
communicate a diagnostic interpretation and are not included 
in TBS. The classification of such changes as benign/reactive 
or dysplastic should be based on the cytoplasmic and nuclear 
alterations present and reported under the appropriate general 
category/interpretation.
Occasionally, benign-appearing glandular cells may be seen 
in post-hysterectomy patients that can have a wide variety of 
sources, including adenosis, metaplasia, and prolapse of the 
remaining fallopian tube after a simple hysterectomy.37,38 This 
finding can be communicated to the clinician under the NILM 
category and per current ASCCP guidelines does not require 
further follow-up.9 Other non-neoplastic changes that may be 
reported under NILM include atrophy and tubal metaplasia.
Details regarding the morphology of these entities are dis-
cussed elsewhere in this book.
endometrial cells
TBS 1991 recommended that benign-appearing endometrial 
cells in postmenopausal women be reported as an “epithelial 
cell abnormality” based on the increased risk for endometrial 
adenocarcinoma (6%) and endometrial hyperplasia (12%) on 
a meta-analysis.3,39,40
In TBS 2001, a new category was included to report the pres-
ence of benign-appearing endometrial cells in women aged 40 
years or older.7,10 The basis for including this new category in TBS 
2001 was twofold: (a) review of the published literature showed 
an exceedingly low rate of significant lesions in anyone less than 
40 years of age, and (b) pathologists may lack clinical informa-
tion on menstrual dates/menopausal status, hormone therapy/
tamoxifen, abnormal bleeding, and other endometrial carcinoma 
risk factors. It is important to include in the interpretation whether 
the cytology is “negative for squamous intraepithelial lesion.”
Only exfoliated, intact endometrial cells should be reported 
under the “other” category. As described in Bethesda 2001, the 
exfoliated groups of endometrial cells may be of epithelial and/
or stromal origin; morphological distinction of these two cell 
types is usually not possible. Directly sampled lower uterine seg-
ment or abraded stromal cells/histiocytes, when present alone, 
should not be reported under this category. Atypical endome-
trial cells should be reported as an epithelial glandular cell 
abnormality.1,10
The prevalence of benign-appearing endometrial cells cervi-
cal in Pap tests from women aged 40 years or older is difficult 
to assess due to differences in study designs, but has been esti-
mated to range from 1–3/100 to 1/1600 or less.39 After adoption 
of TBS 2001, there have been many reports in the cytology litera-
ture that have shown minimal risk associated with this interpre-
tation, especially in premenopausal women.40 This TBS category 
has been controversial for clinicians and initially resulted in an 
increase in endometrial biopsies.
It may be useful to add an educational note to this interpre-
tation in order to clearly communicate to clinicians that this 
interpretation has an increased risk of neoplasia, but the risk 
is low, especially in premenopausal women and those without 
endometrial carcinoma risk factors, and that clinical correlation 
with other risk factors and symptoms is necessary. Examples 
of educational notes for this interpretation can be found in the 
second edition of the Bethesda atlas.10
The 2006 ASCCP guidelines provide additional guidance and 
suggest that for asymptomatic women who are documented by 
clinical history to be premenopausal, with benign appearing 
endometrial cells, endometrialstromal cells, or histiocytes; no 
further evaluation is required. For documented postmenopausal 
women with endometrial cells, on the other hand, endometrial 
assessment is suggested, regardless of symptoms.9
epithelial cell Abnormalities: Squamous cell
Squamous intraepithelial lesion (SIL) encompasses the morpho-
logic spectrum of noninvasive squamous epithelial abnormali-
ties associated with HPV infection. Since the Bethesda System 
the Bethesda System for Reporting cervical cytology
6
was introduced in 1988, this spectrum has always been divided 
into low-grade (LSIL) and high-grade (HSIL) categories. LSIL 
encompasses changes referred to as “HPV effect,” “koilocytosis,” 
and mild dysplasia/cervical intraepithelial neoplasia (CIN 1). 
HSIL includes moderate dysplasia (CIN 2) and severe dysplasia/
carcinoma in situ (CIN 3). The basis for this bipartite classifica-
tion of SIL in TBS is based on the principles that this division 
(a) better reflects natural history and clinical management and 
(b) has better intra- and interobserver reproducibility than does 
a three-tiered reporting system.
Atypical Squamous Cells (ASC)
The term ASCUS was initially introduced into the earliest ver-
sion of the Bethesda System to reflect the reality and limitations 
of light microscopy in classifying borderline cytologic changes. 
The use of multiple ASCUS qualifiers such as “not otherwise 
specified” (NOS), “favor reactive,” and “favor SIL/dysplasia” led 
to overuse of this category and by 1996, ASCUS interpretations 
accounted for a mean of 5.2% of all cervical cytology reports in 
the United States.41 ASCUS interpretations caused dilemmas for 
clinicians due to the lack of standardized follow-up and vari-
ability of outcomes.
With advances in the understanding of the biology of HPV 
infections and results from various natural history studies,42 as 
well as from the NCI ALTS trial,6 the focus of cervical cancer 
screening has shifted from detecting and treating any CIN to 
focusing on treating high-grade CIN. Based on this concept, in 
TBS 2001, the term ASCUS was replaced by ASC, which has a 
narrower definition and only two qualifiers: atypical squamous 
cells of undetermined significance (ASC-US) and atypical squa-
mous cells, cannot exclude HSIL (ASC-H).7 A subclassification 
was aimed at having greater clinical utility by clearly separating 
equivocal findings into those that are worrisome for HSIL in dis-
tinction from other types of ASC. As a general guide, the major-
ity of ASC interpretations should fall into the ASC-US qualifier 
(90–95%) with only 5–10% into the ASC-H category.7,10
ASC is not a single biologic or interpretive entity: it encom-
passes a spectrum of cellular changes reflecting a variety of 
pathologic processes that for one reason or another cannot be 
more definitively categorized. Specifically, ASC should be used 
for changes suggestive of SIL, that are either quantitatively or 
qualitatively insufficient for a definitive interpretation. For a 
cell to be classified as ASC, it should show squamous differen-
tiation, an increase in nuclear cytoplasmic ratio, and minimal 
nuclear changes.10 In each case of ASC, the cytopathologist must 
consider the summation of the morphologic abnormalities in 
terms of quantity and severity within the context of the clinical 
information provided.
Atypical Squamous Cells of Undetermined Significance (ASC-US)
Most often, ASC-US involves noninflammatory changes in 
squamous cells with mature, superficial/intermediate-type cyto-
plasm. Nuclear enlargement is approximately two-and-a-half to 
three times the area of a normal intermediate squamous nucleus, 
but the chromatin remains evenly distributed without signifi-
cant hyperchromasia. Nuclear outlines are smooth and regular, 
although there may be variation in nuclear size. The differen-
tial diagnosis is usually between a reactive change versus LSIL 
but the change(s) quantitatively or qualitatively fall short of 
establishing a definitive interpretation of LSIL. Round or ovoid 
cells that resemble large metaplastic or small intermediate cells 
may also be classified as ASC-US. In liquid-based preparations, 
the cells may appear smaller and rounder compared to conven-
tional smears. The cells in question should always be compared 
to “normal”-appearing intermediate cells on the same slide. 
In distinguishing reactive changes, cells that demonstrate pale 
round nuclei and even chromatin distribution favor an interpre-
tation as NILM rather than ASC.
Atypical Squamous Cells, Cannot Exclude High-Grade Squamous 
Intraepithelial Lesion (ASC-H)
The ASC-H category is useful for changes suggestive of, but fall 
short of a definite interpretation of HSIL. The differential includes 
HSIL and mimics of HSIL. A variety of patterns can be recognized:
 1. Small cells with a high nuclear to cytoplasmic ratio 
or “atypical (immature) metaplasia. Nuclear abnor-
malities such as abnormal shapes, hyperchromasia, 
and chromatin irregularity favor HSIL over benign 
metaplasia.”
 2. Crowded sheet pattern or so-called hyperchro-
matic cell groups. Dense cytoplasm, polygonal cell 
shape, and distinct cell borders favor squamous over 
endocervical cells. This cell pattern includes a broad 
differential from normal (atrophy, endometrial cells) 
to neoplastic (endocervical adenocarcinoma, HSIL, or 
HSIL involving glands) changes.
 3. Atypical cells in the setting of atrophy, atypia seen 
following radiation therapy, poorly preserved endo-
metrial cells or histiocytes, and intrauterine device 
users may all show cellular changes that are difficult 
to distinguish from HSIL. In such situations, a 
 designation as ASC-H may be appropriate.
laboratory Reporting of ASC
Subsequent to the publication and dissemination of TBS 1988/ 
1991, many clinicians felt overwhelmed by ASCUS interpreta-
tions in their patient practices. This phenomenon was not limited 
to the United States or to TBS; greatly increased rates of minor 
degrees of abnormality have been observed in countries that do 
not use TBS.43 The reasons underlying this real or perceived ASCUS 
explosion were twofold: (1) The constant specter of medical–
legal litigation has lowered the threshold for diagnosis of cellu-
lar abnormalities in many laboratories;9,31 and (2) atypical cases 
historically may have been camouflaged in vague terms such as 
“inflammatory atypia,” “benign atypia,” “borderline HPV,” and 
“koilocytotic atypia.” The aggregation of all such equivocal cases 
under one heading highlighted the subjective, interpretative 
nature of cytopathologic diagnosis, something long understood 
by laboratorians but not always recognized by clinicians. Some 
contend that in TBS 1991, ASCUS merely replaced the old Pap 
Class 2 or “inflammatory atypia” designations. However, a study 
by Sidawy and Tabbara demonstrated that by using criteria 
similar to those outlined above, almost two-thirds of 88 smears 
previously interpreted as “inflammatory atypia” could be reclas-
sified as reactive; only 3 out of 57 cases (5%) had CIN (all low 
grade) on follow-up colposcopic biopsies. In contrast, among 
the smears that fulfilled ASCUS criteria, 61% correlated with 
 colposcopic biopsies positive for CIN.44
Laboratory rates of ASC will vary depending on the patient 
population, the diagnostic criteria used, and the experience 
and skill of the microscopist(s). If used appropriately, ASC 
should be an infrequent designation employed only when cel-
lular changes elude a more definitive interpretation. Although 
83
84
Diagnostic CytologyPART TWO
there is no “correct” percentage rate of ASC, benchmarks were 
provided when ASCUS was introduced in the Bethesda termi-
nology. In a low-risk population, it was suggested that the rate 
of ASCUS should be less than 5%. For laboratories that serve 
high-risk populations(e.g., sexually transmitted disease clinics 
or colposcopy clinics), the rate of ASCUS could be higher, but 
by 1991 guidelines should not exceed two to three times the rate 
of SIL; thus the ratio of ASCUS/SIL suggested was in the range 
of 2–3:1. A 1993 CAP survey focusing on laboratory utilization 
of ASCUS found that 86% of responding laboratories used the 
term ASCUS and the median ASCUS rate was 2.8%, with 90% of 
laboratories reporting rates of less than 9%. The median ASCUS/
SIL ratio was 1.7; for 90% of laboratories, the calculated ratio 
was less than 3.6.41 In 2003, a follow-up CAP survey on Bethesda 
2001 implementation and reporting rates showed a decrease in 
the average ASC/SIL ratio (from a median of 2.0 in 1996 to 1.4 
in 2002). This can be explained by increased LSIL detection on 
LBP and also possibly by using Bethesda 2001 criteria more 
stringently.31
Sherman and colleagues, in a study correlating cytopatho-
logic diagnoses with detection of HPV DNA, also found that use 
of TBS criteria reduced the percentage of inconclusive “atypi-
cal” smears. Overall, a consistent relationship between high-risk 
HPV detection and TBS diagnostic categories was evident. High-
risk HPV types were detected in 10% of negative smears, 30% 
of ASCUS, and 60% of SIL specimens. Based on these data, the 
authors proposed using high-risk HPV testing as an objective 
quality assurance measure to assess the performance of a cytopa-
thology laboratory.45 These results were substantiated by ALTS.5,6
It is well established that ASC-US is one of the least repro-
ducible cytologic interpretations.12,46 Various quality assurance 
monitors may be utilized to evaluate the laboratory’s utilization 
of ASC. These include the following:
 1. Correlation of ASC-US cases with high-risk HPV posi-
tivity rates; results from ALTS indicate that this should 
be in the range of 40–60%, or in essence that ASC-US 
is a 50–50 proposition between SIL (usually LSIL) and 
cellular changes unrelated to HPV;47, 48
 2. Correlation of ASC cases with results of colposcopi-
cally directed biopsy;
 3. Review of ASC cases by a second cytopathologist; and
 4. Calculation of ASC/SIL ratio.
The ASC-HPV+/ASC ratio closely mirrors the ASC/SIL ratio. 
However, the ASC-HPV+/ASC ratio offers the additional advan-
tage of identification of aberrant trends where ASC and SIL are 
both being misinterpreted, which may allow ASC/SIL ratios to 
remain within “acceptable” ranges despite the erroneous trend.49 
After implementation of LBPs, many laboratories have reported 
an increase in SIL rates over the increase in ASC rates, such that 
lower ASC/SIL ratios are being seen in many laboratories.31 The 
prior 2–3:1 suggested ratio for ASC/SIL may undergo revision as 
future benchmarking results are gathered.
Clinical management of ASC
Women with ASC have a low prevalence of invasive cancer, esti-
mated at 0.1–0.2%.50 The prevalence of CIN 2/3 is substantially 
higher in women with ASC-H (37–40%)51,52 than in those with 
ASC-US (11.6%).51 ASC-US/high-risk HPV-positive cases over 
2-years follow-up in ALTS have the same cumulative risk of CIN 
2/3, about 27–28%, as a cytologic LSIL.47 In contrast women 
in the ALTS who were ASC-US/high-risk HPV-negative showed 
a very low (1.4%) absolute risk of subsequently detected CIN 
3 or worse and no cancers were detected in the two-year study 
period, similar to women with negative cytology in the absence 
of HPV testing.53
The ASCCP consensus guidelines8,9 for ASC follow-up have 
seen widespread penetration into US practice. For ASC-US, HPV 
DNA testing, repeat cytological testing, and colposcopy are all 
acceptable methods for managing women over 20 years of age. 
When liquid-based cytology is used, reflex oncogenic or high-
risk HPV DNA testing is the preferred management approach. 
ASC-US/high-risk HPV positive women should be managed in a 
fashion similar to those with LSIL. In adolescents (20 years and 
younger), follow-up with annual cytology is suggested due to 
the high prevalence of HPV in this age group and the low risk 
of persistence.9
ASC-H on the other hand needs more aggressive follow-up, 
with colposcopic evaluation at the first interpretation. HPV test-
ing is not recommended for triage of ASC-H. However, if a CIN 
2/3 lesion is not identified at colposcopy, follow-up with either 
HPV testing at 12 months or cytology testing at 6 and 12 months 
is recommended.9
Squamous intraepithelial lesions (Sil)
In TBS, LSIL and HSIL encompass the spectrum of precursors to 
squamous carcinoma of the cervix. Unlike CIN and dysplasia 
classifications that maintain HPV as a separate diagnostic cate-
gory, low-grade SIL incorporates changes of HPV as well as mild 
dysplasia/CIN 1. High-grade SIL includes moderate dysplasia/
CIN 2, severe dysplasia/CIN 3, and carcinoma in situ/CIN 3. 
Cytologists are of course free to append degrees of dysplasia or 
CIN classifications to a SIL interpretation.
Conceptual Basis for Two-Tiered Terminology of SIL
Previous terminology classifications—degrees of dysplasia and 
grades of CIN 1 to 3—have emphasized the morphologic con-
tinuum of squamous lesions that was thought to reflect a con-
tinuous process in the development of cervical cancer. Natural 
history studies42 and HPV research have since established that 
HPV infection is a necessary cause for cervical carcinogene-
sis;55,56 however, even most oncogenic or high-risk HPV types 
cause transient low-risk lesions that regress, and cervical carci-
noma develops in a small subset of persistent/progressive HPV 
infections.57 It is estimated that approximately 70% of cervical 
cancers are associated with HPV 16 or 18.58
The two-tiered LSIL/HSIL Bethesda approach attempts to 
morphologically distinguish minor from significant lesions; 
however, morphology is an imperfect reflection of biologic 
potential. Low-grade lesions, particularly those that persist, may 
progress or be associated with the development of high-grade 
lesions, and some high-grade lesions may regress.57 Some have 
questioned setting TBS division of LSIL and HSIL at the break-
point of CIN 1–2, or mild/moderate dysplasia, arguing that 
some CIN 2/moderate dysplasias should be considered low-
grade lesions. However, because some CIN 2 lesions represent 
high-grade disease processes, conservatism dictated its inclusion 
into the more severe TBS category to ensure maximal sensitivity 
of the process.
Morphologic Features
An interpretation of LSIL based on cellular changes asso-
ciated with HPV requires nuclear as well as cytoplasmic 
 abnormalities. Nuclear changes may include enlargement with 
the Bethesda System for Reporting cervical cytology
6
hyperchromasia or pyknosis, and chromatin smudging and 
wrinkling of nuclear contours. Cytoplasmic changes consist 
of a well-defined perinuclear cavity, associated with peripheral 
thickening of the cytoplasm or cytoplasmic orangeophilia, and 
rounding of cellular contours. Specimens with subtle changes 
that fall short of definitive LSIL may be categorized as ASC-US. 
Cytoplasmic vacuolization (pseudokoilocytosis) alone, in the 
absence of any nuclear atypia, is considered a benign change 
and should not be classified as LSIL or ASC-US.
Intraepithelial precursors of squamous cell carcinoma 
present a spectrum of morphologic changes within which one 
is able, in most cases, to classify lesions as LSIL or HSIL; how-
ever, occasional “borderline” cases occur. In the CAP Interlabo-
ratory Comparison Program in Cervicovaginal Cytology (PAP), 
the discrepant rate between low- and high-grade lesions ranged 
from 9.8 to 15% for cytotechnologist, pathologist, laboratory, 
and all responses.59 Cytology and histology may also be discrep-
ant; 15–25% of women with LSIL cytology are found to have 
histologic CIN 2/3 on further work-up.48,59 Features thatfavor a 
high-grade lesion include increased numbers of abnormal cells, 
higher nucleus to cytoplasmic ratios, greater irregularities in the 
outline of the nuclear envelope and nuclear chromatin distri-
bution, and increased number of chromocenters. The appear-
ance of the cytoplasm may also assist in determining whether 
a borderline case is low- or high-grade SIL. LSIL changes typi-
cally involve “mature,” intermediate, or superficial type cyto-
plasm with well-defined polygonal borders. Cells of HSIL have 
a more immature type of cytoplasm, either delicate and lacy or 
dense/metaplastic, with rounded cell borders. Lesions previ-
ously termed “pleomorphic dysplasia,” “keratinizing dysplasia,” 
or “atypical condyloma,” which are composed of single cells 
or clusters of cells with enlarged hyperchromatic nuclei and 
abundant but abnormally keratinized cytoplasm, are always 
considered HSIL.
HSIL, Cannot Exclude Invasion
In rare cases of HSIL, invasive carcinoma may be difficult to 
exclude. Examples include atypical keratinized cells without 
diathesis/necrosis in the background or cases in which the back-
ground is suspicious but malignant cells are not seen.10 This 
terminology may be used in such cases to communicate the 
increased concern to the clinician.
Squamous Cell Carcinoma
Squamous cell carcinoma is defined as an invasive malig-
nant tumor with squamous differentiation. TBS does not 
subdivide squamous cell carcinoma into keratinizing and non-
 keratinizing types, although the atlas does discuss the morphol-
ogy separately. On LBPs, tumor diathesis may be more difficult 
to recognize; and as such, in the United States a trend toward 
undercalling squamous cell carcinoma as HSIL, especially on 
LBP, has been noted.60
Management of SIL
Low-grade SIL. Based on natural history studies of HPV infection, 
it is clear that the majority of cytologically detected LSIL regress 
within an average of two years.42 After implementation of liquid-
based cervical cytology, there has been a steady increase in the rate 
of LSIL in the United States—in 2003 the median rate was 2.4%.31 
Anecdotal experiences suggest that this has further increased with 
the use of location guided screening. In ALTS, the HPV positivity 
rate in LSIL was 83%; a meta-analysis published in 2006 reported 
a 76.6% positivity rate.61 Thus HPV DNA testing is not suggested 
for initial triage of LSIL. Initial colposcopy identifies preva-
lent CIN 2 or greater in 18% of women with LSIL; subsequent 
follow-up over 2 years identified another approximately 10% 
CIN 2/3, irrespective of whether the initial colposcopy was 
 negative or showed histologic CIN 1.54,62
Colposcopy is recommended for managing LSIL; exceptions 
include adolescents, postmenopausal, and pregnant women. If 
no lesion is identified or colposcopy is unsatisfactory, endocer-
vical sampling is recommended for non-pregnant women. If 
CIN 2/3 is not detected, post-colposcopically, either HPV test-
ing at 12 months or repeat cytology at 6 and 12 months is sug-
gested. In adolescents with LSIL, initial colposcopy and/or HPV 
testing is not recommended; they should be followed by annual 
 cytologic testing. Further details can be found in the ASCCP 
management guidelines.9
High-grade SIL. The median percentile reporting rate of HSIL 
in the United States is estimated at 0.5%,31 and approximately 
2% of women with HSIL cytology have invasive carcinoma.63 
Follow-up of cytologic HSIL carries a significant risk of a CIN 
2/3—a single colposcopy identifies 53–66% of prevalent CIN 
2/3; and CIN 2/3 is found in 84–97% of women who proceed to 
a loop electrosurgical procedure (LEEP).64 Thus both colposcopy 
and LEEP are acceptable for management of cytologic HSIL.9
epithelial cell Abnormalities: Glandular cell
Background
Cervical cytology is primarily a screening test for cervical squamous 
intraepithelial lesions and squamous cell carcinoma; cytology may 
have lower sensitivity for detection of glandular lesions has lower 
sensitivity due to limitations in sampling and interpretation.
Reporting Glandular Cells in TBS 2001
In TBS 2001, the term atypical glandular cells of undetermined 
significance (AGUS) has been eliminated to avoid confusion, 
particularly among clinical staff, with ASC-US. Abnormal glan-
dular cells should be subclassified when possible as endocervical 
or endometrial; otherwise the generic term “atypical glandular 
cells” should be used. It is also advisable to use the qualifiers 
“not otherwise specified” or “favor neoplastic” for endocervi-
cal and glandular cells to convey the level of concern about any 
abnormality identified. The qualifier “favor reactive” from TBS 
1991 has been eliminated as follow-up studies show that results 
are similar to those in the NOS category, and as such, this quali-
fier provides no useful predictive value. Atypical endometrial 
cells are not further qualified due to difficulty in doing so and 
lack of reproducibility of the morphologic criteria. Adenocarci-
noma in situ is a separate interpretive entity in TBS 2001, having 
been well described and shown to have moderately good repro-
ducibilty since the 1991 TBS version.65
Atypical Glandular Cells (AGC)
As with its squamous ASC counterpart, this designation applies 
to glandular cells that demonstrate changes beyond those 
encountered in benign reactive processes, yet which are insuf-
ficient for an interpretation of in situ or invasive adenocarci-
noma. This interpretation should be further qualified, where 
possible, to indicate whether the cells are thought to be of 
endocervical or endometrial origin. This category includes a 
broad morphologic spectrum ranging from atypical-appearing, 
85
86
Diagnostic CytologyPART TWO
reactive processes all the way to adenocarcinoma in situ (AIS). 
Therefore, lesions falling into this category should be further 
subclassified, if possible, according to whether a neoplas-
tic process is favored or the changes are non-specific (NOS). 
Specific comments may be added to the interpretation if perti-
nent clinical findings and/or history are available and relevant 
(polyps, IUD, etc.).
Atypical Endocervical Cells, NOS
Endocervical cells can show a variety of changes associated 
with benign/reactive processes in the endocervical canal. 
 Reactive endocervical cells can show some pleomorphism 
of cell size as well as nuclear enlargement, multinucleation, 
and prominent nucleoli; however, there is usually a honey-
comb or sheet-like pattern and nuclei remain round and the 
chromatin bland. Such changes are usually recognized as NILM 
and not included in the AGC category. Cells that show cyto-
logic changes beyond those recognized easily as reactive such 
as significant nuclear enlargement/crowding, hyperchromasia, 
loss of mucin, and loss of polarity should be considered for 
inclusion in the atypical endocervical cells, NOS category. Such 
changes may be seen in conditions such as tubal metaplasia, 
radiation therapy, endocervical polyps, and microglandular 
hyperplasia, and in IUD users, but also in neoplastic condi-
tions in a small percentage of cases. This category therefore 
includes changes that are in excess of those attributable to a 
reactive/reparative condition but which fall short of those seen 
in glandular neoplasia.
Atypical Endocervical Cells, Favor Neoplastic
These cells are characterized by cellular strips and rosettes 
demonstrating elongated, overlapping nuclei with moderately 
coarse chromatin and hyperchromasia. The peripheral border 
of the glandular clusters may be “feathered,” with protruding 
nuclei, in contrast to the smooth communal border typical of 
glandular fragments. In LBPs cells are more rounded and three-
 dimensional. Cellular changes, while suspicious for in situ or 
invasive adenocarcinoma, are quantitatively or qualitativelyinsufficient for an outright interpretation as such.
Atypical Endometrial Cells
These are usually small groups of cells with slightly enlarged 
nuclei, and variable prominence of nucleoli and nuclear 
hyperchromasia. Their distinction from cytologically benign 
endometrial cells is based primarily on the criterion of 
increased nuclear size. When dealing with LBPs, it is important 
to keep in mind that menstrual/shed endometrium is often 
well preserved and may show nuclear size and shape pleomor-
phism and the presence of nucleoli. The differential of atypi-
cal endometrial cells is broad and may include endometrial 
 polyps, endometritis, IUD associated changes, hyperplasia, 
and carcinoma.
Endocervical Adenocarcinoma (In Situ and Invasive)
Endocervical AIS is a high-grade endocervical neoplastic lesion 
that cytologically demonstrates nuclear enlargement, hyper-
chromasia, stratification, and mitotic activity. Invasive carci-
noma overlaps cytologically with AIS, but may show features 
of invasion, including prominent nucleoli and tumor diathesis. 
The possibility of a coexisting squamous lesion should always 
be carefully assessed when a glandular lesion is detected, due to 
the high rate of coexistence of SIL in cases with AIS.66–68
Endometrial Adenocarcinoma
The cytologic features are directly related to the histologic grade 
of the tumor, with well-differentiated cases yielding malignant 
cells with minimal atypia and poorly differentiated tumors 
being obviously malignant. Tumor diathesis is often difficult to 
appreciate, particularly in LBP. In general endometrial lesions 
yield fewer cells than do directly sampled endocervical lesions.
Extrauterine Adenocarcinoma
A clean background and tumors whose cytologic features are not 
characteristic of uterine/cervical tumors should raise the possi-
bility of metastasis. Diathesis is usually not seen unless there 
is direct extension from the rectum or bladder with associated 
tissue destruction.
diagnostic difficulties
Criteria indicating invasion—tumor diathesis and macronucle-
oli—may be absent in the majority of well-differentiated, early 
adenocarcinomas. It also can be difficult to differentiate SIL 
with gland involvement from AIS. HSIL/CIS involving endocer-
vical glands may yield round cell clusters with smooth periph-
eral contours showing group polarity and “columnar” shape 
of individual cells, thus mimicking a glandular abnormality.69 
Additionally, SIL and AIS may coexist in up to 50% of cases, and 
at conization, a high proportion of AIS specimens demonstrate 
concurrent SIL.68
Benign entities such as tubal metaplasia, directly sampled 
lower uterine segment (LUS) endometrial cells, and cervical 
endometriosis may all morphologically mimic AIS. Fragments of 
tubal metaplasia may demonstrate crowded sheets of glandular 
cells with enlarged nuclei as well as cell fragments with nuclear 
palisading and nuclear overlap, mimicking some of the mor-
phologic features of AIS.70 However, rosette formation is uncom-
mon in tubal metaplasia, and the nuclear chromatin tends to be 
more finely granular. The most helpful findings though, when 
present, are cytoplasmic terminal bars and cilia.
Directly sampled endometrial tissue may mimic AGC or glan-
dular neoplasia. Inadvertent sampling of the LUS may occur 
because of closer approximation of the LUS to the cervical os 
following cone biopsy71 or with aggressive use of endocervical 
brushes. In contrast to spontaneously exfoliated endometrial cells, 
which typically shed as tight ball-like clusters, direct brushing of 
endometrial tissue yields large cellular fragments. These fragments 
often recapitulate their native three-dimensional architecture 
with branching tubular glands enmeshed in stroma composed of 
round to spindle-shaped cells.72 Glandular cells show crowding 
with overlapping round nuclei and scant cytoplasm. Peripheral 
palisading may be evident. The low power recognition of branch-
ing glands and glandular-stromal complexes is an important clue 
to avoid confusion with AGC or glandular neoplasia.
Conventional smears with a diagnosis of adenocarcinoma 
consistently identified correctly by CAP interlaboratory glass 
slide program participants were significantly more likely to 
have more abnormal cells, larger abnormal cells, larger nuclei, 
marked atypia, and hyperchromasia than cases that performed 
poorly.73 Glandular lesions have a slightly different morphol-
ogy on LBPs; specifically the cells may be flatter, feathering less 
prominent, and diathesis more difficult to appreciate.10 Details 
are discussed elsewhere in this book. However, as for squamous 
lesions, there have been reports showing increased detection 
of glandular abnormalities on LBPs compared to conventional 
smears.68,74,75
the Bethesda System for Reporting cervical cytology
6
management
Atypical glandular cells are estimated to be reported in only 0.2% 
of cervical cytology tests in the United States.31 While AGC may be 
associated with benign and reactive conditions such as endocervi-
cal/endometrial polyps, it is clear from several studies that AGC is 
a “high-risk” interpretation compared to ASC; the reported rate of 
neoplasia in follow-up of AGC ranges from 9 to 38%.67,68,76,77
Due to the high risk of a significant lesion associated with 
a cytologic interpretation of AGC, colposcopy with endocervi-
cal sampling is recommended for women with all subcatego-
ries of AGC and AIS. In women over 35 years of age, additional 
endometrial sampling is recommended. Endometrial sampling 
is also recommended for women under the age of 35 with clini-
cal indications suggesting that they may be at risk for neoplas-
tic endometrial lesions, such as unexplained vaginal bleeding 
or conditions suggesting chronic anovulation. In women with 
atypical endometrial cells, both endometrial and endocervical 
sampling should be done initially. In 2006 ASCCP suggested 
that while HPV testing alone is not appropriate for initial triage 
of any subcategory of AGC or AIS, HPV DNA testing at the time 
of colposcopy is preferred in women with atypical endocervical, 
endometrial, or glandular cells NOS, and the results should be 
utilized in overall patient management.9
educational notes/Suggestions
The use of educational notes/comments is optional. If these are 
used by the pathologist/laboratory, it is suggested that they be 
concise, be phrased in the form of a suggestion, not a directive, 
and be substantiated by published guidelines from professional 
organizations.7 Examples can be found in the second edition of 
the Bethesda atlas.10
Ancillary Testing
If ancillary testing, such as high-risk HPV, has been performed, 
whether the report is issued concurrently with the cervical cytol-
ogy result or as an addendum/separate report will depend on 
the laboratory’s information system, turnaround time for such 
testing, and clinical expectations. The methodology utilized for 
the ancillary test should be specified. Suggestions for reporting 
of molecular tests are provided in the Bethesda atlas.10
Automated Review
For cervical cytology preparations that undergo computer-only 
or computer-assisted review, the type of instrument used and 
any result should be included in the report. In addition, if there 
was no “human” review of the slide, this should be made clear 
in the report.
interobserver Reproducibility 
in Cervical Cytology
In an effort to improve standardization, clarity, and reproduc-
ibility of cervical cytology reporting, the second edition of the 
Bethesda atlas10 emphasized more detailed morphologic criteria 
and had many more images, which were complimented by addi-
tional images on the Bethesda website.11 In addition, as part of 
the ASC–NCI Bethesda Project, a web based interobserver repro-
ducibility study was designed to gauge cervical cytologyrepro-
ducibility prior to publication of the atlas and website. A range 
of classic and borderline images (77) were included for inter-
pretation; approximately 651 cytotechnologists and pathologists 
worldwide participated in the study. It was apparent from the 
results that the morphology presented was more important in 
classifying images correctly than were professional or academic 
degrees, or other variables assessed. In this study, exact agree-
ment with the TBS panel was relatively low (57%), although 
agreement was 84.1% at the threshold of distinguishing NILM 
from non-negative. Participants achieved a higher sensitivity for 
correctly classifying high-grade squamous lesions than that for 
high-grade glandular lesions. The details of this study have been 
published12 and all the images and associated histograms of 
participants’ responses are available for review on the Bethesda 
website.11
The Bethesda System and Reporting 
Anal-Rectal Cytology
Anal cancer is considered an appropriate target for cytologic 
screening in selected high-risk populations. The anatomic com-
monality of the anal–rectal canal and the cervical mucosa is 
reflected in that both have a transformation zone. HPV is a com-
mon risk/etiologic factor for cancers of the anus and cervix and 
subsequently the morphology of cytology samples from both 
sites is comparable. It follows that sampling devices, preparation 
techniques, and morphologic interpretation using the Bethesda 
system terminology utilized for cervical cytology can readily be 
applied for anal–rectal cytology screening.
Adequacy criteria for anal–rectal cytology are based, at 
present, on limited personal experiences. As a guide, minimum 
adequacy cellularity should be in the range of 2000–3000 nucle-
ated squamous cells for conventional smears and for LBP sam-
ples 1–2 nucleated cells/high-power field for ThinPrep (20 mm 
diameter) preparations and 3-6 nucleated squamous cells/high-
power field for SurePath (13 mm diameter) preparations.10
Normal elements seen in anal–rectal specimens include 
nucleated, anucleate, and metaplastic squamous cells, rectal 
columnar cells, fecal matter, and mucus. A comment should 
be included in the report about the presence of a transformation 
zone component. Cytomorphologic criteria are quite similar 
to those utilized for cervical cytologic interpretation; how-
ever, there is a higher incidence of poor preservation, cellular 
 degeneration, and cytoplasmic keratinization/parakeratosis, 
and classic koilocytes are less frequently identified.10,78
When targeting high-risk groups, the rate of epithelial abnor-
malities noted is far higher than that reported for cervical cytol-
ogy.78,79 Reports from the United States suggest that anal–rectal 
cytology screening is sensitive but has low specificity for predict-
ing the grade of the lesion, with a tendency to under-represent 
the grade of squamous abnormality. While it has been shown 
that screening high-risk patients by cytology is effective, the 
present impediments to the success of early detection of anal 
cancer by this method include limited clinical expertise and 
means for the subsequent treatment/follow-up of these patients, 
and the high risk of complications associated with excisional 
procedures at this site.
87
88
Diagnostic CytologyPART TWO
can Pathologists interlaboratory compari-
son program in cervicovaginal cytology. 
Arch Pathol Lab Med 2000;124:203-211.
In: Wied GL, Keebler CM, Rosenthal DL, et 
al (eds) Compendium on Quality Assurance, 
Proficiency Testing and Workload Limitations 
 25. Smith HO, Tiffany MF, Qualls CR, et al. 
The rising incidence adenocarcinoma 
relative to squamous cell carcinoma of 
 5. ASCUS LSIL Triage Study (ALTS) Group. 
Results of a randomized trial on 
management of cytology interpretations 
of atypical squamous cells of undeter-
mined significance. Am J Obstet Gynecol 
2003;188:1383-1392.
 6. Schiffman M, Adrianza ME. ASCUS-LSIL 
Triage Study. Design, methods and char-
acteristics of trial participants. Acta Cytol 
2000;44(5):726-742.
 7. Solomon D, Davey D, Kurman R, et al. 
for the Forum Group Members and the 
Bethesda 2001 Workshop. The 2001 
Bethesda System—terminology for 
reporting results of cervical cytology. 
JAMA 2002;287:2114-2119.
 8. Wright TC Jr, Cox JT, Massad LS, et al. 
ASCCP-Sponsored Consensus Con-
ference. 2001 Consensus Guidelines 
for the management of women with 
cervical cytological abnormalities. 
JAMA 2002;287:2120-2129.
 9. ASCCP consensus guidelines. Am J Obstet 
Gynecol 2007; 197(4):346-355.
in Clinical Cytology. C
Cytology, 1995: 90-9
 16. Birdsong G. Pap sm
understanding satis
Diagn Cytopathol 20
 17. Studeman KD, Ioffe
et al. Effect of cellul
ity of detecting squa
liquid-based cervica
2003;47(4):605-610
 18. Bolick DR, Kerr J, St
of cellularity in the 
grade and low grade
thelial lesions. Acta 
923 (abstract).
 19. Bos AB, van Ballegoo
Akker-van Marle E, e
status is not predicti
cervical cancer in the
smears. Am J Clin Pa
 20. Elias A, Linthorst G,
significance of endo
diagnosis of cervica
Acta Cytol 1983;27:2
hicago: Tutorials of 
4. 
ear adequacy: is our 
factory … or limited? 
01;24:79-81.
 OB, Puszkiewicz J, 
arity on the sensitiv-
mous lesions in 
l cytology. Acta Cytol 
.
anely BE, et al. Effect 
detection rates of high 
 squamous intraepi-
Cytol 2002;46:922-
ijen M, van den 
t al. Endocervical 
ve of the incidence of 
 years after negative 
thol 2001;115:851-855.
 Bekker B, et al. The 
cervical cells in the 
l epithelial changes. 
25-229.
the uterine cervix in the United States: 
a 24-year population based study. 
Gynecol Oncol 2000;108:397.
 26. Zheng T, Holford TR, Ma Z, et al. The 
continuing increase in adenocarcinoma 
of the uterine cervix: a birth control 
phenomenon. Int J Epidemiol 
1996;25:252-258.
 27. Ransdell JS, Davey DD, Zaleski S, 
et al. Clinicopathologic correlation of 
the unsatisfactory Pap smear. Cancer 
Cytopathol 1997;81:139-143.
 28. Davey D, Austin M, Birdsong G, et al. 
ASCCP patient management guidelines: 
Pap test adequacy and quality indicators. 
J Low Genit Tract Dis 2002;6:195-199.
 29. Nielsen ML, Davey DD, Kline TS. 
Specimen adequacy evaluation in 
 gynecologic cytopathology. Diagn 
Cytopathol 1993;9:394-403.
 30. Davey DD, Woodhouse S, Styer P, et al. 
Atypical epithelial cells and specimen 
adequacy. Arch Pathol Lab Med 
2000;124(2):203-211.
Concluding Remarks
Cervical cytology has seen many changes since its introduction 
in the 1960s—liquid-based sampling techniques, automated 
preparation, computer-assisted screening, HPV DNA testing, 
and more recently dual testing, which combines cytology screen-
ing with HPV testing. HPV vaccines entered the market in 2006, 
and are likely to further decrease the incidence of invasive squa-
mous cell carcinoma of the cervix and its precursor lesions. 
Cervical cancer screening guidelines have undergone significant 
changes after implementation of LBP and HPV testing and with 
advances in the understanding of cervical neoplasia.80 It is pre-
dicted that the number of cervical cytology tests performed will 
decrease significantly in the future if there is compliance with 
these guidelines.81
The Bethesda System has met the goals that were conceived 
at the time of its implementation in 1988. It has seen successful 
penetration into cervical cytology reporting worldwide, allow-
ing laboratories to use consistent terminology in conveying 
results to clinicians and thus enabling comparison of studies 
across many countries and health care systems. The use of the 
Bethesda ASC-US terminology prompted the NCI-sponsored 
ALTS trial, the results of which have significantly impacted the 
management of equivocal and low-grade cervical cytologic 
abnormalities.The Bethesda terminology has been updated 
twice—in 1991 and 2001 since its inception in 1988 in order 
to keep pace with the advances in our understanding of cervi-
cal cancer and evolving technologies in cervical cancer screening 
and prevention. The Bethesda System also provided the basis for 
the ASCCP to develop consensus guidelines for management of 
cervical cytologic abnormalities as defined by TBS. This proc-
ess of re-evaluation and revision will continue in the future in 
order to provide the most accurate, reproducible, and relevant 
terminology. Optimal communication and ultimately patient 
care outcomes will therefore ensue.
References
 1. National Cancer Institute Workshop. The 
1988 Bethesda System for reporting cervi-
cal/vaginal cytologic diagnoses. JAMA 
1989;262:931-934.
 2. National Cancer Institute Workshop. The 
Bethesda System for reporting cervical/
vaginal cytologic diagnoses. Acta Cytol 
1993;37:115-124.
 3. Kurman RJ, Solomon D. The Bethesda 
System for Reporting Cervical/Vaginal 
Cytologic Diagnoses: Definitions, Criteria 
and Explanatory Notes for Terminology and 
Specimen Adequacy. New York: Springer-
Verlag, 1994.
 4. Davey D, Woodhouse S, Styer P, et al. 
Atypical epithelial cells and specimen 
adequacy: current laboratory practices of 
the participants in the College of Ameri-
 10. Solomon D, Nayar R (eds) The Bethesda 
System for Reporting Cervical Cytology, 2nd 
edn. New York: Springer, 2004. 
 11. Bethesda System Web Atlas. Available at 
the American Society of Cytopathology 
website: http://www.cytopathology.org
 12. Sherman ME, DasGupta A, Schiffman M, 
et al. Bethesda Interobserver Reproduc-
ibility Study (BIRST). Cancer Cytopathol 
2007;111:15-25.
 13. Bottles K, Reiter RC, Steiner AL, et al. 
Problems encountered with The Bethesda 
System: The University of Iowa experi-
ence. Obstet Gynecol 1991;78:410-414.
 14. Kline TS, Solomon D. Guidelines for 
specimen adequacy: A plea for workable 
definitions. Diagn Cytopathol 1991;7:1-2.
 15. Solomon D, Henry M. Specimen adequacy. 
 21. Mauney M, Eide D, Sotham J. 
Rates of condyloma and dysplasia in 
Papanicolaou smears with and 
without endocervical cells. Diagn 
Cytopathol 1990;6:18-21.
 22. Mitchell H, Medley G. Influence of 
endocervical status on the cytologic 
prediction of cervical intraepithelial neo-
plasia. Acta Cytol 1992;36:875-880.
 23. Mitchell H, Medley G. Longitudinal study 
of women with negative cervical smears 
according to endocervical status. Lancet 
1991;337:265-268.
 24. Mitchell HS. Longitudinal analysis 
of histologic high grade disease after 
negative cervical cytology according to 
endocervical status. Cancer Cytopathol 
2001;93:237-240.
the Bethesda System for Reporting cervical cytology
6
 31. Davey DD, Neal MH, Wilbur DC, et al. 
Bethesda 2001 Implementation and 
reporting rates: 2003—practices of 
participants in the College of American 
Pathologists Interlaboratory Comparison 
Program in cervicovaginal cytology. Arch 
Pathol Lab Med 2004:28(11):1224-1229.
 32. Fidda N, Miron J, Rodgers WH, et al. 
Impact of the new Bethesda system 2001 
on specimen adequacy of cervicovaginal 
smears. Diagn Cytopathol 2004;30: 
235-239.
 33. Quddus MR, Sung CJ, Eklund CM, et al. 
ASC:SIL ratio following implementation 
of 2001 Bethesda system. Diagn Cytopathol 
2004;30(4):240-242.
 34. Prandi S, Beccati D, Aloysio GD, et al. 
Applicability of the Bethesda system 
2001 to a public health setting. Cancer 
Cytopathol 2006;108(5):271-276.
 35. Colgan TJ, Woodhouse SL, Styer PE, et al. 
Reparative changes and the false- 
positive/false-negative Papanicolaou 
test. Arch Pathol Lab Med 2001;125(1): 
134-140.
 36. Snyder TM, Renshaw AA, Styer PE, et al. 
for the Cytopathology Resource Com-
mittee. Altered recognition of reparative 
changes in ThinPrep specimens in the 
College of American Pathologists Gyne-
cologic Cytology Program. Arch Pathol 
Lab Med 2005;129(7):861-865.
 37. Ponder TB, Easley KO, Davila RM. 
Glandular cells in vaginal smears from 
posthysterectomy patients. Acta Cytol 
1997;41:1701-1704.
 38. Bwetra C. Columnar cells in posthyster-
ectomy vaginal smears. Diagn Cytopathol 
1992;8(4):342-345.
 39. Greenspan DL, Cardillo M, Davey DD, et al. 
Endometrial cells in cervical cytology: 
review of cytologic features and clinical 
assessment. J Low Gen Dis 2006;10(2): 
111-122.
 40. Browne TJ, Genest DR, Cibas ES. The 
clinical significance of benign-appearing 
endometrial cells on a Papanicolaou test 
in women 40 years or older. Am J Clin 
Pathol 2005;124(6):834-837.
 41. Interlaboratory Comparison Program in 
Cervicovaginal Cytology. 1993 PAP sup-
plemental questionnaire on Laboratory 
Practice: ASCUS. College of American 
Pathologists, 1994.
 42. Ho GYF, Bierman R., Beardsle L, et al. 
Natural history of cervicovaginal papil-
lomavirus infection in young women. 
N Engl J Med 1998;338:423-428.
 43. Raffle AE, Alden B, Mackenzie EFD. 
 Detection rates for abnormal cervical 
smears: What are we screening for? 
 Lancet 1995;345:1469-1473.
 44. Sidawy MK, Tabbara SO. Reactive 
change and atypical squamous cells 
of undetermined significance in 
Papanicolaou smears: A cytohisto-
logic correlation. Diagn Cytopathol 
1993;9:423-429.
 45. Sherman ME, Schiffman MH, Lorincz AT, 
et al. Toward objective quality assurance 
in cervical cytopathology. Am J Clin Pathol 
1994;102:182-187.
 46. Stoler MH, Schiffman M. Atypical 
 Squamous Cells of Undetermined 
 Significance-Low-grade Squamous 
Intraepithelial Lesion Triage Study (ALTS) 
Group. Interobserver reproducibility 
of cervical cytologic and histologic 
interpretations: realistic estimates from 
the ASCUS-LSIL Triage Study. JAMA 
2001;285(11):1500-1505.
 47. The ALTS Group. Results of a randomized 
trial on the management of cytology 
interpretations of atypical squamous cells 
of undetermined significance. Am J Obstet 
Gynecol 2003;188:1383-1392.
 48. Stoler MH. New Bethesda terminology 
and evidence based management guide-
lines for cervical cytology findings. JAMA 
2002;287(16):2140-2141.
 49. Ko V, Nanji S, Tambouret R, et al. Testing 
for HPV as an objective measure for 
quality assurance in cervical cytology. 
Cancer Cytopath 2007;111:67-73.
 50. Jones BA, Novis DA. Follow-up of abnor-
mal gynecologic cytology: a College of 
American Pathologists Q-probes study of 
16 132 cases from 306 laboratories. Arch 
Pathol Lab Med 2000;124:665-671.
 51. Sherman ME, Solomon D, Schiffman M. 
Qualification of ASCUS: a comparison of 
equivocal LSIL and Equivocal HSIL cervi-
cal cytology in the ASCUS LSIL Triage 
Study. Am J Clin Pathol 2001;116:386-394.
 52. Sordon M, Dilworth HP, Ronnett BM. 
Atypical squamous cells, cannot exclude 
high-grade squamous intraepithelial 
lesion. Diagnostic performance, human 
papillomavirus testing, and follow-up 
results. Cancer Cyopathol 2006;108:32-38.
 53. Safaeian M, Solomon D, Wacholder S, 
et al. Risk of precancer and follow-up 
management strategies for women with 
human papillomavirus-negative atypical 
squamous cells of undetermined 
significance. Obstet Gynecol 
2007;109:1325-1331.
 54. Cox JT, Solomon D, Schiffman M. Pro-
spective follow up suggests similar risk 
of subsequent CIN 2 or 3 among women 
with CIN 1 or negative colposcopy and 
directed biopsy. Am J Obstet Gynecol 
2003;188:1406-1412.
 55. Nubia Muñoz N, Bosch FX, Sanjosé 
SD, et al. Epidemiologic classification 
of human papillomavirus types associ-
ated with cervical cancer. N Engl J Med 
2003;348(6):518-527.
 56. Lorincz AT, Reid R, Jenson AB, et al. 
Human papillomavirus infection of the 
cervix: Relative risk associations of 15 
common anogenital types. Obstet Gynecol 
1992;79:328-337.
 57. Wright TC, Schiffman