400 Vol. 71, No.6
The Clinical Usefulness of the Vaginal Smear
MILTON ROSENTHAL, M.D., San Francisco
detection in a very small aliquot of the total vaginal
SUMMARY exudate. The oiher thing these investigators demon-
The chief value of the vaginal smear is as strated was that many malignant cells in isolated
a simple, rapid means of accelerating the spread are sufficiently distinctive to make possible a
diagnosis of cancer in those cases in which strong presumption or certainty of their nature. It
the presenting symnptoms or lesions are not was shown that a great deal might be learned and
sufficiently alarming to provoke immediate that benefits to patients might accrue because by
surgical investigation. this approach genital carcinoma might be discov-
Carcinoma of the endocervix, carcinoma in ered early, whereas formerly there was only the
situ, and carcinoma existing together with random biopsy applicable in obvious lesions to
inflammatory and scarifying lesions are most demonstrate their character. From the outset Papa-
likely to be discovered by the method. The nicolaou and Traut have emphasized that the effec-
very early ulcerative malignancy, before the tiveness of this method depended upon examination
fibrous proliferative reaction produces the of the smear by an experienced observer and sub-
classical raised edge and granular base, also stantiation by biopsy methods.
has been detected on the smear taken just There are several reasons for this somewhat sur-
before the cautery was applied. The small prising reliability of certain diagnostic hallmarks in
nubbin without ulceration also yields malig- isolated malignant cells. One is that the very severe
nant cells. These lesions almost always in- shrinking and wrinkling caused by suspension of
clude the surface epithelium. tissue in a fixative is avoided. On smears the cells
The positive smear has induced the imme- are spread out and flattened and frequently cover
diate repetition of biopsy or curettage when a greater area than they do even in situ. This can
the first sections have not confirmed a clinical be easily demonstrated by examining fresh teased
suspicion. It can prevent temporizing when preparations in saline. The greater detail that can
there is an easy explanation (such as estrogen be seen is startling-not quite as much as is seen
administration or menopausal irregularity) in dry smears of sternal marrow, for instance, but
for spotting and metrorrhagia. Examination much more than can be seen even after fixation in
of smears has corrected a number of false solutions especially designed to minimize shrinkage,
impressions caused by poor visualization. such as Zenker's. As a result, irregularity of nuclear
The smear method is irreplaceable for the outline is much more dependable as a criterion of
early diagnosis of recurrence after radiation. malignancy in smears than in sections. Another
Salvage of the patient is rarely to be expected unexpected advantage of smear staining is the very
from such a discovery, but planning of pal- remarkable constancy of hematoxylin uptake by the
liative therapy is greatly aided. nuclei. The variability of hematoxvlin staining in
sections seems to be partly due to the fact that when
a large number of cells are crowded together, espe-
THE rationale for cancer detection based upon cially if the nuclei are closely approximated, the
vaginal smears depends upon the fact that epi- excess hematoxylin is mechanically hindered from
dermoid cancer sheds superficial cells in numbers diffusing out into the acid differentiating solution.
far out of proportion to the size of the tumor. This The author has found it entirely unnecessary to con-
tendency to exfoliate from the surface is not at all trol differentiation visually, as is done with section
dependent upon the presence of ulceration; it is staining. The result of this constancy of staining is
probably due to a biological property of cancer cells that the appearance of hyperchromatism is a much
themselves. Cancer cells in tissue culture, for in- more valuable characteristic of isolated malignant
stance, have been observed to "break off" from the cells than it is in sections.
main clump and "metastasize" to other parts of the By repeatedly looking at corresponding smears
culture medium with greater ease than normal cells. and sections, it eventually becomes easy to envision
In essence the discovery of Papanicolaou and how cells in the one would look in the other. After
Traut was the demonstration of the frequency with two years of studying comparative material, the
which such cells appear, making possible their author is convinced of the value of having a smear
Presented before the Section on Obstetrics and Gyne-
as an aid in making a diagnosis if section examina-
cology at the 78th Annual Session of the California tion is equivocal. This is not meant to violate the
Medical Association, Los Angeles, May 8-11, 1949.
The work of this project is supported from funds from
very proper practice of employing the biopsy as the
the U. S. Public Health Service. yardstick whereby the reliability of the smear is
From the Cytology Laboratory, Departments of Obstet-
rics and Gynecology, and Pathology, University of Cali-
measured. Nor does it mean that the biopsy need
fornia Medical School. not determine the diagnosis. That the smear is not
December, 1949 USEFULNESS OF THE VAGINAL SMEAR 401
a substitute for biopsy has been emphasized by all of in situ carcinoma to spread on the surface lining over a
observers, and it is to be hoped there is no more large area. There may be no ulceration or induration.
misunderstanding on that score. However, it is true CASE No. 2: A 53-year-old visitor at the anti-cancer
that, as the examiner gains experience and confi- clinic had no symptoms and the cervix looked perfectly
dence, there are some cells which are so distinctive normal. The rest of the pelvic examination was negative
and familiar to him that no number of negative except for a mild atrophic vaginitis. The smears showed
biopsies would shake his belief in the presence of almost nothing other than carcinoma cells, even though they
were taken from material in the posterior fornix. Section of
cancer. Learning about such cells is a subjective the cervix showed that the carcinoma was linin, the canal.
and individual matter and depends upon how much The thin layer of cancer could be associated with no indura-
material has been examined. There is another large tion. A considerable nunmber of smears were taken, to be
group of cells which are not diagnostic but which used for teaching purposes, and they showed literally thou-
are derived from canicer 99 per cent of the time. sands of malignant cells.
Therefore, if these cells appear, the patient is sus- The point may be made that the two cases reported
pected of cancer and should be followed for a very are of the kind in which cancer is the least likely
long time if the first diagnostic procedures do not to be suspected on the basis of either presenting
demonstrate it. In this respect the vaginal smear symptoms or pelvic examination; but the lesion
method is different from most laboratory proce- sheds copious cells, and it is in such cases the
dures; rapport should exist between the clinician chance of cure is best.
and cytologist. This does not mean that the cytolo- CASE No. 3: A 52-year-old woman, one year past the
gist requires the clinical information. On the con- menopause, complained of slight mucoid discharge for the
trary, it has been found to be better that the previous five months. Pelvic examination disclosed a 1 cnm.
cytologist know nothing about the patient until the smooth translucent polyp protruding from the os and it
cells have been evaluated; otherwise subjective was thought that this was the source of the discharge.
errors will creep in. In the past, advocates of the Several small fibromyomata were also palpated. Vaginal
method have had to use only the material which smears showed striking clusters of cells with nuclear pre-
demonstrated the reliability of the diagnosis. They dominance, crowding irregularity, and hyperchromatisni.
Curettage revealed adenoacanthoma in the uterine cavity.
have had to refrain from demonstrating the addi- After intracavitary radium treatment, the panhysterectomy
tional group of cases in which the test is a more specimen showe(d no remnant of carcinoma.
senisitive but only a presumptive indication of ma- CASE No. 4: A 32-year-old woman had had a supra-
lignancy. It cart be said that there are diagnostic cervical hysterectomy and right salpingo-oophorectomy he-
smears and there are suspicious smears, and as cause of fibromyomata and endometrial cysts. Subsequently,
experience is gained the former category enlarges an examination showed what was described as a "clean ero-
and the latter increases in usefulness. sion" on the cervix. A smear was taken before cauterization
There are several situations in wxhich the vaginal and abundant malignant cells were found. Biopsy showed
smear has an outstanding and special application. an early non-invading carcinoma.
The first has already been suggested when equivo- The greatest value of the vaginal smear has been
cal biopsies were referred to. Those gynecologists the acceleration of the diagnosis of cancer in
who observe many patients with cancer are regret- women whose presenting symptoms or lesions were
tably familiar with the small cervical lesion or even not sufficiently alarming to provoke surgical diag-
the larger one without well defined edges for which nostic measures. Without doubt, in most cases the
the biopsy report is an admirable piece of literature diagnosis would have been made eventually, but
but without a title. Added to this is the lesion which many months have been saved, and not only by
looks like cancer but which upon biopsy is not pathognomonic cells on the smears, but also by
shown to be cancer. It is instructive to consider suspicious and even atypical cells. It is not the
that a smear is a sampling of the entire surface of function of the vaginal smear to save the trouble
the lesion and the biopsy section is a radial plane, of biopsy or curettage. On the contrary, it increases
ten micras thick, through the lesion. A positive the frequency of these procedures.
smear does not wholly contradict a negative biopsy.
It simply indicates the need for another biopsy. The preparation of the vaginal smear is ex-
tremely simple and takes about 15 seconds. If a
CASE REPORTS lesion is present on the cervix it should be scraped;
CAsE No. 1: A 55-year-old woman without gynecologi- if not, material from the posterior fornix or on the
cal symptoms had a circular reddened area about the cer- lower blade of the speculum is used. The smear
vical os, which was considered to be simple erosion. The need not be uniformly thin a thickness just short
smear showed innumerable malignant cells of the squamous of complete opacity is a good rule. The slides
type. The biopsy showed no epithelium except for a few should then be placed immediately in an equal
rete pegs indistinguishable from normal downward exten- mixture of 95 per cent alcohol and ether before
sions. (One liestitates to say this because anyone but a there is any drying. If there is no ether handy, 95
pathologist is apt to laugh at it, but it is quite possible per cent alcohol is almost as good. After fixation
to wipe the carcinoma right off the surface of a biopsy
specimen.) Further biopsy material demonstrated carci- for an hour, the slide may be allowed to dry and
noma, and a section of the endocervix showed where most it retains its stainability for weeks. It is not neces-
of the carcinoma was. This was an in situ carcinoma al- sary to transport the slides in fixative to the lab-
though the glands were penetrated. It is rather characteristic oratory, and the author has found no advantage in
402 CALIFORNIA MEDICINEV Vol. 71, No. 6
coating them with glycerine as has been advocated compromise with this principle in the diaginosis of
by others. uterine cancer, although it may occasionally be
The method should not be considered as a screen- justified in cancer of the lung when the tumor is
ing test to "rule out cancer." In the author's series, located beyond the view through the bronchoscope.
approximately 4 per cent of proved cases have been A specialized application of the vaginal smear
missed on the smear. It is true that some of these diagrnosis is the detection of recurrence after radia-
smears were technically faulty, having been too tion. This is usually only of prognostic value but
scanty, or taken where bleeding was brisk so that may frequently aid in planning palliative therapy.
nothing but blood was present on the smear. Never- Case detection at anti-cancer clinics in the east-
theless, "Rule out carcinoma" is a very unsuitable ern United States has been low-about one case in
inscription on the request sheet. 500 visitors. Data from similar clinics for which
CASE REPORT examinations are done by the University of Califor-
CASE No. 5: A 32-year-old woman complained of ten- nia Cvtological Laboratory have not yet been ana-
derness in the right lower quadrant. On pelvic examination lyzed. It has been stated that an undue amount of
a small amount of white discharge was seen in the external time and labor is required for the yield, but the
os. Smears from the posterior fornix showed many malignant technicians at the Cytological Laboratory have not
cells, and upon reexamnination a reddened granular area on found such material particularly burdensome. A
the posterior lip was noted. It proved malignant on biopsy. well trained, experienced technician can screen a
In most of the similar cases of poor visualization slide without unusual cell components in five min-
in which reexamination was instigated by positive utes and when the anti-cancer slides are received in
smears, there were lesions on the posterior lip. It small groups and distributed among several tech-
has been difficult to assess the "provocative" value nicians there is no great effort required. No increase
of the vaginal smear in the presence of banal ap- of staining time is needed at all; the slides are
pearing lesions of the cervix because the descrip- stained in batches and it takes no longer to stain
tion of the lesion on the second examination, after 20 than to staiii one.
the positive smear report was given, has frequently Technicians can be trained in about four months
been much more suggestive than the description to recognize suspicious cells. Their task is to mark
given after the first examination. them for evaluation by the cytologist. Because of
At present, patients are not treated for cancer the time required, it is not practicable for a path-
unless the biopsy warrants it. There is 11o need to ologist to do his own screening.