Exfoliative Cytology of the Uterine Cervix

and Vagina
Leopold G. Koss, M.D.

The great interest in exfoliative cytology tion of cancer cells within smears of ex
at the present tinie is due chiefly to the foliated material prepared for microscopic
ease, efficiency and accuracy with which examination constitutes the basis of cyto
it can be utilized for the detection of early logic diagnosis.
cancer of the uterine cervix. With the help
TECHNICAL PRINCIPLES
of a qualified cytopathologist, the prac
ticing physician has within his reach a The correct recognition of cancer is
method of early diagnosis which can lead easier and more reliable if the cells are well
to the successful treatment of one of the preserved. Since drying will damage and
most serious and frequently fatal diseases distort the cells, it is important to obtain
of the female genital tract. While most in the best and niost rapid fixation of the
vasive cancers of the cervix may be readily specimen. For this reason, tue bottle con
recognized on clinical examination, the Wining the fixative should be opened
early stages of the disease and especially before tile smears are taken so that the
carcinoma in situ do not necessarily pro snieared slides can he fixed immediately.
duce appreciable changes in the appear The writer recommends the following
ance of the cervix. In fact, the diagnosis fixatives for genital smears, in order of
of early cervix cancer may be established preference:
(1) equalportions
of 95%
with certainty only by microscopic exami ethyl alcohol and ether; (2) 95% ethyl
nation of tissues or cells derived from the alcohol; (3) isopropyl alcohol (absolute).
lesion. Papanicolaou staining'3 is much pre
ferred to any other staining method be
Diagnostic Principles cause it assures cytoplasmic transparency
and cellular differentiation superior to any
BIOLOGIC PRINCIPLES
obtainable by other methods.
All epithelia of the body, whether nor
METHODS OF OBTAINING SMEARS
mal or cancerous, continuously shed cells
from their surfaces. Thus, cells desquamat The best possible accuracy in cytologic
ing from the uterine and vaginal epithelia diagnosis is achieved if both a vaginal and
may be readily obtained for microscopic a cervical smear are submitted to the labo
examination from the vaginal cul-de-sac. ratory. The smears should be evenly spread
In order to insure better diagnostic accu and placed immediately in the fixative.
racy, vaginal smears must he supplemented Obtaining smears at the time of menses
by s@nears obtained directly from the epi should be avoided. The patient should not
thelial surface of tile cervix. The recogni douche for 24 hours before the smears are
taken.
From the Memorial Hospital for Cancer and Allied
Diseases, New York, N. V. VAGINAL SMEARS —¿The vaginal smear
This work was performed wit/i f/se assistance of should he obtained as the very first pro
Mrs. Grace R. Dunce, B. S., Chief Cvtotechsnologist,
‘¿sfei,,orial Hospital for Cancer and Allied Diseases. cedure in the gynecologic examination.

182
Lubricants should be avoided because they
interfere with staining and obscure cellular
detail. The vaginal smear can be effec
tively used for the determination of vari
ous changes within the epitheliurn of the
genital tract, but it fails in the detection of
in situ carcinoma of the cervix in about 30
per cent to 40 per cent of cases. However,
the vaginal snzear is indispensable in the
detection of endomneirial carcinoma.
CERVICAL SMEARS—¿Cervical smears
should be obtained under direct vision. All
lubricants must be avoided when introduc
ing the speculum. If there is difficulty in
inserting the speculum, particularly in
postmenopausal women with an extremely
atrophic and dry vaginal mucosa, the in
strument should be moistened with normal
saline solution.
There are two methods of obtaining
cervical smears: the cotton-tipped appli
cator may be used, or the wooden scraper,
as first suggested by Ayre, which is readily
prepared by cutting a tongue depressor to
the required shape. Both methods are ex
tremely efficient in disclosing the presence
of cervix cancer, whether in situ or inva
sive. The cotton swab, if not too thick,
may be introduced farther into the canal,
a fact occasionally of sonic importance.
The wooden scraper is potentially sonic
what traumatic, but the smears are richer
in cells. The use of either method is largely
a matter of personal preference.
The cervical smear is generally ineffec
tive in the detection of endometrial cancer.
ENDOCERVICAL ASPIRATION —¿Consider
able information may be obtained by aspi
rating the endocervical canal by means of
a small cannula attached to a syringe. The
method is used routinely in sonic centers.
It is very efficient in detecting endocervical
cancer but may fail if the lesion is located
on the portio or in the vagina. Endometrial
cancer may be diagnosed quite often in this
manner.
SELF-OBTAINED SMEARS —¿Several de
vices, such as the tampon,2 have been de
signed which can be introduced into the
vagina by the patient herself. The tanipon
can later he withdrawn by the patient or a
nurse and then sniearcd on slides. The ad
vantage of self-obtained smears lies in the
fact that a large number of screening pro
cedures can be performed on many women
under the supervision of a small medical
staff.
The disadvantages
ofthis
methodare
its relatively low sensitivity and the diffi
culty encountered in the interpretation of
material. The latter is due to the cellular
distortion resulting from smearing the
tampon.The methodalsodeprives
thepa
tient of an opportunity for a complete pel
vic examination. Finally, the introduction
and the withdrawal of the tampon may
cause discomfort for postmenopausal
women with atrophic vaginas. Self-ob
tained smears are recommended only if the
more conventional methods of obtaining
cytologic material cannot be applied. How
ever, if the choice is limited to this smear
or no smear at all, the self-obtained smear
may prove to be of a distinct advantage.
INFORMATION FOR LABORATORY
As with any laboratory procedure, it is
unwise to let the cytopathologist look at
the specimen without having necessary in
formation about the patient. In order to
protect both the clinician and the patholo
gist and to assure the patient of maximum
diagnostic reliability, the following infor
mation should be provided with every
specimen: (I) age; (2) menstrual history
and date of last menstrual period; (3)
clinical findings and diagnosis, and (4)
past history of treatment, particularly cau
tery, surgery, radiation, chemotherapy,
etc., even if not applied to the genital tract.
Scope and Value of Cytology
Considerable skill is required for the
correct interpretation of cytologic speci
mens. Mere fluency in histologic diagnosis
is not sufficient criterion of diagnostic
ability in the field of exfoliative cytology,
but it constitutes an excellent background
for further training and experience.
Knowledge of normal cytology is an
essential
prerequisite
for attemptingto
make the diagnosis of cancer. Besides serv
ing as a tool for cancer diagnosis, the
smears may provide a variety of types of
useful information. These will he men
tioned briefly for completeness and clarity.
183
NORMAL CYTOLOGY
The vagina and the vaginal portion of
the cervix (the portio) are lined by strati
fied squamous epithelium which normally
does not become keratinized (Fig. 1). The
superficial squamous cells which desqua
mate from such an epithelium during the
child-bearing age have abundant cyto
plasm and small nuclei (Fig. 2). Occa
sionally cells from the deeper epithelial
layers may be present in smears. Such cells
have less cytoplasm, are generally smaller
and, depending on the layer of origin, are
classified as interniediate, parabasal or
basal. At certain times, such as prior to
sexual maturity, during the first weeks
postpartum, or after the menopause, the
maturation of the squamous epithelium
may not go beyond the stage of parabasal
cells (Fig. 5).
The endocervical canal is lined with a
layer of tall, mucin-producing columnar
cells. The numerous endocervical glands
are lined in a similar nianner (Fig. 3). The
characteristic columnar appearance of en
docervical cells is evident in cervical
smears (Figs. 2 and 4). Endocervical cells
are uncommonly seen in vaginal smears.
The point of transition between the endo
cervical glandular cells and the squamous
epithelium of the portio is referred to as
the squamocolumnar junction. The junc
tion is usually located in the general vicin
ity of the external os.
BENIGN CERVIX LESIONS
INFLAMMATION —¿ Trichomonas vagi
na/is, a parasite which is a very coninion
cause of vaginitis and cervicitis, may be
readily identified in cervical and vaginal
smears (Fig. 6). Candida albicans (Mo
nilia) may also be readily recognized. The
cytologic changes caused by severe inflam
niatory processes, especially those due to
trichomonads, may present some difficul
ties in interpretation.9 Cellular atypias due
to inflammation may, in rare instances, be
suspected of being carcinoma; conversely,
cancer cells may be mistaken for inflani
matory cells or remain concealed within
the debris and inflammatory exudate in
smears.
184
Occasionally it is a wise course to clear
up the infection by comiservative measures
and then repeat the cytologic examination.
Drastic therapeutic measures, such as elec
trocautery, applications of silver nitrate,
etc., should be avoided as they may only
contribute to the confusion of the cytologic
picture.
EROSION —¿The so-called cervical ero
sion is a sharply demarcated area of red
ness which appears on the surface of the
portio adjacent to the external os. The le
sion is often caused by the presence of
transparent endocervical mucosa replacing
stratified squamous epitheliuni on the sur
face of the cervix. The vessels in the under
lying stroma are visible through the thin
mucosal layer and account for the red ap
pearance of the area. The term eversion or
ectropion of the endocervical mucosa is
more descriptive and should be applied to
this entity. A cytologic smear is quite effec
tive for differentiating a superficially ulcer
ating carcinoma from eversion. The ever
sion sheds only benign columnar cells of
the endocervical type, often arranged in
papillary clusters.
LEUKOPLAKIA —¿Areas of excessive and
abnormal keratin formation on the sur
face of the squamous epithelium of the
cervix appear clinically as white patches
and have therefore been called leukoplakia
(Fig. 7). Leukoplakia may be due to a
variety of mechanical factors, such as pro
lapse, pessary, etc. It may also be pres
ent on an area of cervix previously treated
by cauterization. Leukoplakia results in
the presence of keratinized anuclear
“¿squames―
in smears (Fig. 8). The rela
tively uncommon keratinizing cervix can
cers may appear clinically as foci of leuko
plakia. Such cancers shed abnormal cells
and may be differentiated cytologically
from the benign lesions.
EFFECTS OF TREATMENT ON SMEARS
CAUTERY—¿Cauterization of the cervix
may cause a marked distortion of cells to
the point of their being confused with can
cer. Since these effects may persist for as
long as six weeks, the pathologist should
always be informed that the procedure was
previously carried out.
 RADIATION —¿Radiation niay cause an
immediate and a late effect. The immediate
effect of radiation is manifested by cellular
and nuclear enlargement, vacuolization,
multinucleation and nuclear abnormalities.
The late effect of radiation may persist for
a great many years; the writer has ob
served one case in which it was apparent
19 yearsaftercompletionof radiationpatients with lesions of the cervix epi
treatment. Considerable cellular and nu
clear distortion which are occasionally
present may be misinterpreted as cancer,
unless one is aware of the possibility of
such changes.
ANTIBIOTICS —¿Certain broad-spectrum
antibiotics when applied topically niay
produce massive desquamation of the epi
theliuni and thus conceal the presence of
cancer. No antibiotics should he used topi
ca1l@for at least one @nomzth before smears
are taken.
CARCINOMA OF THE CERVIX
CLASSIFICATION—¿Although two differ
ent types of epithelium occur within the
cervix, most cervix cancers originating on
the portio or in the endocervical canal are
of the squanious or epidermoid type. per cent of cases of in situ cancer and re
Adenocarcinoma of the cervix is relatively
uncommon but may occasionally be found
side by side with epidermoid cancer. In
sonic instances, mixed or muco-epider
moid fornis of cancer may occur which
contain elements of both epidermoid and
mucus-producing cancer.
IN SITU CARCINOMA —¿Considerable evi able to dislodge sonic of the minute foci
dence is now available that invasive carci
noma of the cervix is preceded by in situ
carcinoma. In situ carcinoma is best de
fined as a form of cancer still confined to
the epithelium.
In situ epidermoid carcinonia is char
acterized histologically by a profound up
heaval in the structure of the squamous
epithelium (Fig. 9). There is a loss of
orderly stratification and niaturation. The
component cells vary in size and display
marked nuclear abnormalities. Mitotic fig
ures, normal or abnormal, niay be readily
observed. An extension of the process into
the endocervical glands is not infrequent
and does not constitute evidence of inva
sion.
During the last 50 years, evidence has
accumulated that in situ carcinoma of the
cervix can progress to invasive cervix can
cer. Stoddard, in 1952, found 42 such
cases in the literature. Numerous addi
tional cases have been observed since that
time. In a vast prospective study, Petersen
followed (without major treatment) 126
theliuni that correspond quite closely to
in situ carcinoma and related abnormali
ties. Approximately one third of these pa
tients developed invasive cervix cancer
within nine years of observation. The pe
riod of evolution of in situ carcinoma to
invasive carcinonia may be quite long;
Galvin has reported a period of 16 years
in one case. Statistical evidence demon
strates that patients with invasive cervix
cancer are, on the average, five to 10 or
more years older than patients with in situ
carcinoma.'9
It is iniportant to note that in sonic in
stances in situ carcinomas of the cervix
cannot again be demonstrated after the
initial biopsy or after very superficial treat
ment. This was found in approximately 25
lated lesions (which had no significant
treatment) followed in our institution for
several years.° It is felt that this course of
the disease is secondary to the removal of
the major portion of the lesion by biopsy
forceps. It is likely that, in a favorable
case, the regenerating healthy epitheliuni is
of cancerleftbehind.There are also
numerous instances on record in which in
situcarcinomawas foundincidentally in
uteri removed surgically for other reasons5
and in autopsy material from women who
died of unrelated causes. Thus, many
women with in situ carcinoma may lose
their uteri or their lives from other causes
before developing invasive cervix cancer.
The total evidence accumulated to date
suggests that invasive cervix cancer is pre
ceded by in situ carcinoma in most, if not
all, cases. On the other hand, not all in
situ cancers will necessarily progress to in
vasion within the lifespan of the bearer.
Since it is not possible to prognosticate
the outcome of in situ carcinoma on the
185
strength of the histologic examination,9
nor to ascertain the duration of an in situ
carcinoma prior to detection, the lesion
should be considered a potentially dan
gerous one. Its very slow evolution, how
ever, remnoves it automatically frommi the
category of surgical emnergencies.
The prognosis of treated in situ card
nonia of the uterine cervix is excellent. No
cases of metastases from in situ carcinoma
have been reported, according to our
knowledge. Very nearly 100 per cent of
cases treated by one of the many ap
proaches available survive five years or
longer without evidence of disease. Only
sporadically are cases reported in which
a recurrence of the disease was noted.―
We have observed a few such cases with a
recurrence in the vagina after the initial
treatment by hysterectomy.
Since the prognosis of invasive cervix
cancer is not nearly so favorable, it is ob
vious that cervical cancer should be diag
nosed and treated electively in the in situ
stage. Undoubtedly, even the keenest
methods of investigation will occasionally
fail to detect early cervix cancer, and it
may well be that in an exceptional case the
course of the disease is so fulminating that
there is no time for early detection. How
ever, these situations are so uncommon
that they should not detract from the value
of a sustained search for early cervix can
cer, preferably in the in situ stage.
It must be emphasized here that in situ
carcinoma may he present as a “¿cancerous
coating― at the periphery of invasive can
cer. Because of differences in the thera
peutic approach to the two forms of cervix
cancer, it is essential that sufficient tissue
evidence be obtained prior to therapy in
order to insure that no invasion is present.
EPIDERMOID CANCER —¿The cytologic
diagnosis of epidermoid cancer is based on
recognizable cellular changes in smears.
These changes pertain chiefly to the nuclei
and, to a lesser degree, to the cytoplasni of
cancer cells.
The nuclear changes readily noted are:
enlargement: abnormally dark staining or
hyperchrornasia, due to an increase in
sonic nucleoprotei ns (desoxyribonucleic
acid); irregularities of size, shape and
186
chrornatin pattern; presence of abnormal
nucleoli; and abnormally high mitotic or
premitotic activity. As a result of nuclear
enlargement without a corresponding in
crease in the surrounding cytoplasm, the
surface ratio of the nucleus to the cyto
plasm is changed in favor of the nucleus.
The cytoplasmic changes comprise:
cytoplasmic irregularities, abnormal kera
tin formation and development of bizarre
shaped cells. There niay also be a marked
variation in cell size (Figs. 10 and 14).
DIFFERENTIAL DIAGNOSIS —¿It is often
possible to niake the diagnosis of in situ
carcinoma on cytologic grounds.'7 In addi
tion to cells of frank cancer as defined
above, in situ epidermoid cancers shed
cells which characteristically have a well
differentiated and normally abundant cyto
plasm surrounding an abnormal nucleus.
Such cellswere calleddyskaryotic
by
Papanicolaou'4 and may be classified into
superficial, intermediate, parabasal, and
endocervical types, depending on their
layer of origin as indicated by their cyto
plasmic features (Fig. 12). The presence
of these cells indicates that a certain ten
dency toward epithelial maturation may he
expected to exist within the cervix lesion, a
property that is usually reserved for in situ
carcinoma. While dyskaryotic cells also
may be present in smears of invasive can
cer, they will be few in number. These
criteria, if judiciously used by experienced
observers, may be quite helpful in ruling
out invasion on cytologic grounds. They
are not infallible, however. Histologic con
firmation should be made before treatment
is started.
TREATMENT OF IN SITU CANCER —¿
Treatment of any neoplastic lesion de
pends on its prognosis. The lesion of
known benign behavior requires merely
local removal. Lesions known to metasta
size are, as a rule, treated radically by sur
gery or radiation in the hope that they will
he removed or destroyed as completely as
possible before distant spread has oc
curred; invasive cervix cancer belongs to
this category of lesions.
In situ carcinoma without invasion is a
lesion quite comparable to Bowen's dis
ease of the skin and, as a general rule, is
curable by simple surgical removal. There 12). These lesions have been classified
is, however, a considerable difference of by various authors as atypical hyperplasia,
opinionastowhetherthisremovalshould dysplasia, borderline atypias, koilocytotic
take place by means of local treatment of or warty atypias, etc. Their behavior varies
the cervix or by hysterectomy. While the markedly and cannot be predicted on cyto
writer does not feel qualified to settle this logic or histologic grounds. Sonic lesions
controversy, he would like to point out niay disappear without any known treat
thathe hashad an opportunity toobserve ment or after a biopsy, but some of these
numerous cases of in situ carcinoma, previ lesions persist and are, in all 1i@c1ihood,
ously treated by extensive conization or stages in the genesis of cervix cancer'2
trachelectomy, that did not show any and should be classified as precancerous.
evidence of recurrence over periods rang Such epithelial abnormalities may be
ing from three to five years.6 Cytologic fol found adjacent to in situ and even invasive
low-up is mandatory after limited therapy. cancer, and the possibility of this associa
On the other hand, there is evidence tion should be kept in mind at all times.
that this type of treatment may not be Also, invasive cervix cancer has been
sufficient in sonic instances, because the known to follow such lesions and it would
lesion may be beyond the reach of the be erroneous to leave these lesions un
conization knife.'6 Several such cases have treated unless the patient can be watched
beenobserved
by thewriter.
Thisobserva very careful/v over a period of years for
tion would tend to support the view held signs of developing carcinoma.
by some physicians that a total hysterec The treatment should he guided by the
tomy with a wide vaginal cuff is the treat degree of abnormality observed in histo
nient of choice for extensive in situ can logic material. Conservative treatment,
cer. Whichever mode of therapy is chosen, such as local removal of the borderline le
in situ carcinoma is not a .surt,lcal emer sions by extensive cauterization or coniza
gency and may be treated electively and tion, appears to be quite adequate and pre
without urgency. serves the reproductive function of pa
IN SITU CARCINOMA AND PREGNANCY —¿ tients in the child-bearing age.
Cumulative evidence at this time indicates It should be pointed out also that the
that in situ carcinonia in the pregnant nomenclature
andclassification
ofsuchle
woman is essentially the same lesion as in sions vary from laboratory to laboratory
situ carcinoma in the nonpregnant wom and depend largely upon the individual
an.'' In keeping with our knowledge of the pathologist's approach to the problem.
slowevolution
and excellent
prognosis
of Such lesions may readily become “¿shop
in situ carcinoma, it appears completely ping lesions―,and if they are examined by
permissible to let the patient with this le different authorities they will be given vari
sion complete her pregnancy. The writer ous labels ranging from “¿atypia―
to “¿in
situ
has knowledge of several patients with carcinoma―. The writer believes that the
untreated in situ carcinonia who went nomenclature of such early neoplastic le
through pregnancy and vaginal delivery sions is immaterial, provided they are
without any undue effects. It would seem, recognized and not dismissed as of little
therefore, thateven intheseinstances in importance.
situcancershouldbe treated electively,at ADENOCARCINOMA —¿ Mucus-producing
a tinie selected by the physician after con adenocarcinomas shedvacuolated colum
sultation with his patient. nar cells with markedly abnormal nuclei.
BORDERLINE LESIONS —¿Cytology is an Papillary adenocarcinonias shed papillary
extremely sensitive method for detecting clusters which often cannot be differen
abnormalities of cervix epithelium. A tiated from other fornis of adenocarci
certain percentage of abnormal smears nonia originating anywhere within the
is due to the presence of epithelial lesions genital
tract,
or from adenocarcinonias
which by our present criteria cannot be arising elsewhere and becoming metastatic
classified as in situ cancer7― (Figs. II and tothegenital
tract
(Figs.
15and 16).The

187
muco-epidermoid variety of cervix cancer
sheds primarily cells of the epidermoid
type among which may be found vacuo
lated, mucus-containing cancer cells.
RECURRENT CERVIX CANCER—Cytology
offers
an excellent
opportunity
forthede
tection of recurrent carcinoma of the cer
vix, irrespective of whether surgery or
radiation was used to treat the primary
tumor. The cells of recurrent carcinoma
are usually free of any radiation effect. It
is of interest to note that carcinonia of the
cervix recurring after treatment by radia
tion appears occasionally as in situ carci
noma.8
Confirmatory Biopsy
A n@'cytologic report suggesting or diag
mzosing cervix cancer of any t)'pe mnust be
comzfirmned by biopsy before definitive tizer
apy is instituted. The reasons for this pro
cedure may be briefly summarized as fol
lows:
1. Cytologic study, regardless of how
accurately it is performed, gives only urn
ited information about the location or ex
tent of the lesion. It is of particular impor
tance to realize that invasive cancer may
be found where none was suspected on
cytologic grounds.
2. While a lesion of the cervix may be
observed clinically and the cytologic report
may indicate the presence of cancer, the
lesion may prove to be benign; the cancer
cells may have originated from an area
not noted on clinical examination.
3. The relatively short period of time
that the cytologic techniques have been in
USC accounts for a certain niargin
nostic error that niay vary according to
the exaniiner's experience. The role of in
flammatory processes, radiation, cauteriza
tion, etc., as potential sources of cytologic
error, have already been discussed.
It has to he realized that cytologic
smears may reflect epithelial alterations
which are geographically extremely small
and measure only a few millimeters in di
ameter. Therefore, the epithelium should
not be daniaged prior to biopsy. Any ener
getic scrubbing of the cervix almnost in
variably results in the remnoval of the epi
188
theliu!ml (1/1(1 theme/ore mnay render the
biopsy comnpletely valueless. Only the most
gentle preparations should be applied to
the cervix.
MULTIPLE BIOPSIES
Multiple biopsy specimens should not
be taken blindly and thrown together into
a bottle. Each specimnen should he kept in
a separate bottle, numbered and geo
graphically designated so that an opinion
can be formed as to the extent of the lesion
present. Also, such biopsy specimnens
Silould always include the external os of
the cervix which is the area of greatest con
centration of in situ cancer.4 One way of
obtaining geographically designated speci
mens is the so-called four-point biopsy
method: one specimen is taken from a
point in each of the four quadrants of the
cervix—for example, at 12, 3, 6 and 9
o'clock.4
Schiller's iodine test is very helpful in
discovering areas of epithelial abnormal
ity, including in situ carcinoma.2' Such
areas are poor in glycogen and therefore
do not stain with iodine.
The chief advantage of cervix biopsies
is that they can be performed as an office
procedure and do not require hospitaliza
tion. A disadvantage is the relatively lim
ited amount of information that small tis
sue fragments can give, especially as re
gards the existence of invasive carcinoma.
COLD KNIFE CONIZATION
This procedure consists of removing a
conical portion of cervix in such a way
that the base of the cone surrounds the
of diag external os and the apex is within the endo
cervical canal. The cone should always be
niarked by means of a suture at a pre
designated area, such as, at 12 o'clock. All
of the tissue obtained by conization must
be examnined histologically; otherwise,
grave omissions and diagnostic errors niay
occur.
Conization by nieans of electrocautery
is definitely not advisable because it in
jures the integrity of the tissues which
must be examined microscopically.
Conization offers a very good sampling
of material from the cervix. If a large cone
fails to disclose invasive cervix cancer, the tumor differs substantially from that of
chances are negligible that such a lesion cervixcancer.
exists within the cervix. The procedure Carcinomas of the vulva may also occa
may also be curative in sonic cervix le sionally he the source of abnormal cells.
sions. The chief disadvantages of coniza
CYTOLOGY
tion are the necessity of hospitalization,
however brief, and the fact that it may Cytologic presentation of vaginal or
produce considerable bleeding. vulval squamous carcinoma is indistin
guishable from that of cervical carci
CURETTAGE
noma. The niain difference is the distribu
If the previously described methods of tion of cells within the sniears: cells of
diagnostic confirmation of cervix lesions cervix cancer will be abundant in the cervi
fail, an endocervical or endometrial curet cal smear and scanty in the vaginal smear,
tage niay occasionally disclose the exist while cells of vaginal cancer will he scanty
ence of the lesion. in the cervical smear and abundant in the
vaginal smear. Therefore, if no lesion can
FAILURE OF CONFIRMATION
he found in the cervix, despite the presence
Ina certain
percentage
ofcases,
usually of cells of epidermoid or squamous carci
not exceeding 0.5 per cent of all suspicious noma, it is imperative to investigate the
or positive smears, efforts at confirniation vagina.
of the existence of a cervix lesion may fail. It should be kept in niind that vaginal
Under these circunistances, it is advisable lesions niay occasionally he hidden behind
to initiate a search for a vaginal lesion (see the cervix. In such circunistances, cx
below), an endometrial or ovarian lesion, amination under anesthesia, utilizing the
or a nietastatic lesion. If this search is con Schiller test for guidance, or obtaining
sistently negative and the smears have he smears (appropriately labeled) from vari
conic negative after extensive biopsies, re ous areas of the vaginal mucosa may prove
cutting and examining the biopsy material to be distinctly advantageous.
previously obtained niay reveal the exist Early carcinoma of the vagina niay be
ence of a niinute focus of in situ carcinoma quite inconspicuous, appearing clinically
which was not present in the initial sec as a red or white patch. Biopsies of such
tions of tissue. areas may reveal the source of abnormal
cells.
Diagnosis of Vaginal Cancer In view of the serious prognosis of
vaginal carcinonia, immediate and appro
Primary malignant tumors of the vagina priate therapy should be instituted when
fall essentially into two groups: the botry such a lesion is discovered. Under no cir
oid sarcoma, a variety of rhabdomyosar cumstanccs, however, should therapy he
coma occurring chiefly in the vaginas of applied before the lesion is localized and
children, and the carcinonias affecting then evaluated by biopsy.
women usually after 40 years of age. Cy
tology has no place in the diagnosis of METASTATIC TUMORS
hotryoid sarcoma, but it niay be very suc Mctastases
from distant
sites
or exten
cessfully applied to the detection of carci sion of cancer from adjacent organs may
noma of the vagina. also be observed in the genital tract of the
Carcinomas of the vagina are usually of female. The writer has noted cancer cells
the epidernioid or squanious type. Very in genital smears in several cases of mam
little is known at the present time about niary carcinoma as well as in other meta
the duration of the in situ stage since the static tumors having their primary origin
lesion is comparatively rare. However, elsewhere. It is important to emphasize
even very small, locally invasive cancers that primary cervical or vaginal carcinoma
of the vagina may produce distant metas may occur in the presence of cancer else
tases and therefore the behavior of this where in the body.

189
 Summary ablydevelopintoinvasive
cervixcarci@
noma within the lifespan of the patient.
Our present general knowledge pertain 5. The presence of in situ carcinoma
ing to the early stages of cancers of the in limited biopsy material calls for further
cervix and the vagina may be summarized investigation of the cervix to rule out the
as follows: concomitant presence of invasive cervix
1. By application of cytology to the cancer.
detection of cervix cancer in the in situ 6. If invasion has been ruled out, in
stage, it is within the reach of the medical situ carcinoma ceases to be a surgical
profession to reduce very significantly the emergency. It can be treated electively and
mortality rate due to this disease. without urgency.
2. All, or nearly all, of the carcinomas
7. In the presence of cytologic findings
which suggest epidermoid
of the cervix arc preceded by the stage of or squamous
carcinonia in situ. carcinoma, a thorough search should he
instituted for carcinoma of the vagina or
3. In situ carcinonia may remain sta
tionary for periods varying from one to 10
the vulva, if the lesion cannot be localized
years or longer. If adequately treated, the
within the cervix.
prognosis is excellent. 8. Metastatic cancers from sites other
4. The evidence presently available
than the genital tract may occasionally be
suggests invari the cause of positive genital smears.
thatnotallinsitucancers

Legends

Fig. 1. Normal squamous epithelium of cer
vix and vagina. (“<
233)
Fig 2. Mature squamous cells as seen in cervi
cal and vaginal smears. In the center there is
also a cluster of ciliated endocervical
(@;<935)
Fig. 3. Normal endocervical mucosa. Note
tall mucus-secreting cells. (X 233)
Fig. 4. Normal endocervical cells in smears.
(X 935)
Fig. 5. Appearance of a smear in advanced
postmenopausal atrophy. Note small size of
squamous cells and numerous inflammatory
cells in the background. (X 935)
Fig. 6. Tricluomnomuasinfestation in vaginal
smear. The parasite is an oval, greenish gray
cells. body with a small peripheral nucleus. Squa
mous cells show a moderate nuclear atypia.
(X 935)
Fig. 7. Hyperkeratosis of cervix with epithe
hum presenting clinically as a white patch or
leukoplakia. (X 233)
Fig. 8. Anuclear squamous cells originating
from the surface of leukoplakic mucosa.
(X 935)

190
@ ‘¿I.. -%... - -
.,-
@ f -@.‘. @‘¿? •¿.-c@z@:@

Fig. 1. Fig. 2.

4@

@ I
•¿s a I
.4,

¶@ @. ‘¿p

Fig. 3.
: Fig. 4.
.@:

4 4

Is.'
I ,

..@ S •¿

@ @44.

C'..
4 0' 4,

@ •¿ •¿
—¿â€¢â€˜
‘¿..—
@..

Fig. 7. Fig. 8.

191
 4, . , . .1' -

@ @4S•'•'• ( .@,,
@ .-@ .1,.,@

@ ‘¿â€˜@!,@“ -. 4

@ ‘¿p.c,,, @..
@ d.,@-

@ 1: ‘¿44 ,,,
4
,, ,

.1i@• @.

@ L:,. ,., 1) , .@4 ‘¿@‘

‘¿â€œ1.
•¿â€˜:@.@

@
4_I, .,.q@
@,.@...@ . . / .4f@
‘¿.‘@, .11 ‘¿ -‘.@_ ,,@:

•¿ ,. -
p

‘¿p.

/4 —¿

@
@ I
@•¿
‘¿
.‘ •¿14
@h

Fig. hi. “¿.12.

4.'. , $ .@.‘

@ 4', 4 ‘¿4
@
@ 4 4 4 .4 5 .@ :p. 5 4

—¿ ,m.',

Fig. 13. Fig. 14.

....,‘:.@••)‘4

Fig. 15.

192
 References

1. Ayre, I. E.: The vaginal smear; “¿precancer―cell prevalence and posipregnancy persistence. Cancer
studies using a modified technique. Am. I. Obst. & 9:1195-1207, 1956.
Gynec. 58:1205-1219, 1949. ii. May, H. C.: Carcinoma in-situ of tile vagina
subsequent to hysterectomy for carcinoma in-situ of
2. Bader, G. M.; Simon, T. R.; Koss, L. G., and
the cervix. An,. J. Obst. & Gynec. 76:807-811, 1958.
Day, E.: A study of the detection-tampon @nethiod as
a screening device for uterine cancer. Cancer /0:332- /2. McKay, D. G.; Terjanian, B.; Poschyachinda, D.;
337, 1957. Younge, P. A., and Hertig, A. T.: Clinical and patho
logic significance of anaplasia (atypical hyperplasia)
3. Cuyler, W. K.; Kauftnann, L. A.; Palumbo, L., and of the cervix uteri. Ohst. & Gynec. 13:2-21, 1959.
Carter, B.: Cytologic studies in malignant lesions of
the vagina. Study 1. Primary squamous cell carci 13. Papanicolaou, G. N.: A new procedure for stain
noma. Sung., Gynec. & Obst. 96:115-117, 1953. ing vaginal smears. Science 95:438-439, 1942.
4. Foote, F. W., Jr., and Stewart, F. W.: The ana 14. Papanicolaou, G. N.: Cytologic diagnosis of ute
tomical distribution of intraepithelial epidermoid car rine cancer by examination of vaginal and uterine
cinomas of the cervix. Cancer 1:431-440, 1948. secretions. Am. J. Clin. Path. 19:301-308, 1949.
15. Petersen, 0.: Precancerous changes of the cervi
5. Howard, L. H., Jr.; Erickson, C. C., and Stoddard, cal epithelium in relation to manifest cervical carci
L. D.: A study of the incidence and histogenesis of noma; clinical and histological aspects. In: Acta
endocervical metaplasia and intraepithelial carci radiol. (supp. 127), 1955; pp. 1-168.
noma; observations on 400 uteri remnoved for non
cervical disease. Cancer 4:1210-1223, 1951. /6. Przybora, L. A., and Plutowa, A.: Histological
topography of carcinoma in situ of tile cervix uteri.
6. Jordan, M. J.; Bader, G. M., and Day, E.: A Cancer 12:263-277, 1959.
rational approach to the management of atypical
lesions of the cervix. Am. J. Obst. & Gynec. 72:725- 17. Reagan, I. W., and Hicks, D. J.: A study of in
739, 1956. situ and squamous-cell cancer of the uterine cervix.
Cancer 6:1200-1214, 1953.
7. Koss, L. G., and Dun/ce, G. R.: Unusual patterns 18. Reagan, J. W.; Hicks, D. I., and Scott, R. B.:
of squamous epitheliumn of the uterine cervix; cyto Atypical hyperplasia of uterine cervix. Cancer 8:42-
logic and pathologic study of koilocytotic atypia. Ann. 52, 1955.
N. Y. Acad. Sc. 63:1245-1261, 1956.
19. Stern, E.: Rate, stage, and patient age in cervical
8. Koss, L. G.; Melamned, M. R., and Daniel, W. W.: cancer; an analysis of age specific discovery rates for
In-situ epidermnoid carcinoma of the cervix and vagina atypical hyperplasia, in situ cancer, and invasive can
following radiotherapy for cervix cancer. To be pub cer in a well p:pulation. Cancer 12:933-937, 1959.
lished.
20. Stoddard, L. D.: The problem of carcinoma in
9. Koss, L. G., and Wolinska, W. H.: Trichomnonas situ with reference to the human cervix uteri. In:
vaginalis cervicitis and its relationship to cervical McManus, I. F. A.: Progress in Fundamental Medi
cancer; a histocytological study. Cancer 12:1171- cine. Philadelphia. Lea and Febiger. 1952; pp. 203-
1193, 1959. 260.
10. Marsh, M., and Fitzgerald, P. 1.: Carcinoma in 21. Younge, P. A.: Cancer of the uterine cervix; a
situ of the human uterine cervix in pregnancy; preventable disease. Obst. & Gynec. 10:469-481, 1957.

Legends

Fig. 9. Epidermoid carcinoma in situ of the mic differentiation and nuclear abnormalities.
cervix. Note abnormal make-up of the epi (X 935)
thehium as compared with Fig. 1. (X 233) Fig.13.Invasive
epidermoid
carcinoma
of
Fig. 10. Appearance of a cervical smear from cervix. ( 233)
the same case as Fig. 9. Note marked nuclear
abnormalities. (X 935) Fig. 14. Cells from the same case as Fig. 13.
Note nuclear abnormalities and nearly total
Fig.11.Borderline atypiaof cervix epithe absence of any cytoplasmic differentiation.
hum. Observe nuclear abnormalities and good
(X 935)
epithehial stratification. Lesion does not war
rant the diagnosis of in situ carcinoma but Fig. 15. Adenocarcinoma of the endocervix.
may precede it, or may be associated with it. (X 233)
(X 233) Fig.16.Cells
fromsamecaseasFig.15.Note
Fig. 12. Dyskaryotic squamous cells from the nuclear abnormalities and especially promi
same case as shown in Fig. II. Note cytoplas nent nucleoli. (X 935)

193