Diagnostic Cytology
the interrupted nuclear membrane due to irregular sedimenta-
tion on the membrane of the coarsely granular chromatin. It is
rare to find eosinophilic-staining nucleoli, but chromocenters
or "false" nucleoli are rather common (see Fig. 8.87). "True"
nucleoli may be obscured by the coarse hyperchromatic chro-
matin but are definitely present, because they are related to the
high proliferation rate of the lesion. The absence of nucleoli
would be in contradiction to the relatively high proliferative
activity, which is evident from the presence of mitoses even in
the uppermost layers of the epithelium. Macronucleoli, which
are usually present in nuclei from invasive cancer cells, are only
rarely seen in carcinomas in situ.
Relative Nuclear Area.
The relative nuclear area (nuclear area in
relation to cytoplasmic area) with an average value of 50% bet-
ter reflects the primitive character of these cells from carcinomas
in situ than just the size of the cell or the nucleus.
Key features of carcinoma in situ
• Larger number of abnormal cells than other grades of
• Cells arranged predominantly in syncytial aggregates
with indistinct cell borders;
• Some isolated cells;
• Small cells with little cyanophilic cytoplasm and with-
out signs of maturation;
• Round-to-oval nuclei occupy half of cellular area;
• Bare nuclei are commonly noted;
• Even to irregular chromatin distribution with fine to
coarsely granular texture;
• Nuclear membrane is commonly disrupted and irregu-
lar (cerebriform); and
• Micronucleoli rarely noted—obscured by dense
Reagan and Hamonic14 reported the presence of inflamma-
tory exudate in specimens from cancers in situ in 68% of cases,
compared with 32% in specimens from dysplasias. In our own
material from a population-screening program, the percentage
of smears showing an inflammatory exudate in cases of carci-
noma in situ and severe dysplasia (85.1%) was not significantly
different from the percentages of smears found with less severe
changes or without signs of epithelial abnormalities.72
Biologic Significance
It is generally agreed that the majority of lesions classified as
carcinoma in situ are actively proliferating lesions that, if left
alone, will finally evolve into invasive cancer. However, with cur-
rently available knowledge and techniques, it is virtually impos-
sible from cytologic, histologic, or clinical data to predict which
reaction will regress and which will progress to a more severe
It is generally accepted that all invasive cancers of the uterine
cervix develop from a carcinoma in situ or from severe dyspla-
sia. However, massive evidence shows that not all carcinomas in
situ or severe dysplasias progress into an invasive process; many
regress into a lesion of less severity or disappear completely. This
means that these lesions diagnosed as intraepithelial cancer are
essentially not malignant in nature.
Many studies have tried to uncover the characteristics of
those carcinomas in situ that progress to an invasive process,
to discriminate these from the lesions that show regression but
that appear identical on light microscopy. Data on the follow-up
of severe dysplasias and carcinomas in situ are not abundant,
because most lesions are immediately treated when diagnosed.
Available data indicate that only a relatively small number of
lesions eventually evolve to an invasive squamous cell carci-
Prospective studies designed to monitor the course of car-
cinoma in situ suggest a relatively slow evolution to invasive
carcinoma. Analyses of incidence rates and mortality indicate
that the length of the preinvasive stage is on the order of 12
to 15 years. Reports of so-called fast-growing cancers, which are
claimed to have passed very quickly through a premalignant
preinvasive phase, as should be evidenced by a nondiagnos-
tic cytologic examination shortly before the diagnosis of the
invasive process, appear with some periodicity in the literature.
However, on careful analysis of available evidence, the accuracy
of the preceding cytologic diagnosis usually does not stand,
because false-negative diagnoses due to sampling errors cannot
be precluded. In our own registry of premalignant and malig-
nant lesions of the uterine cervix, encompassing 17 years before
and during a population-screening program, difference in mean
age between cancer in situ and invasive cancer was 13.7 years
(carcinoma in situ: mean age 39.2 years; invasive cancer: mean
age 52.9 years).
Developm ental C arcinom a in Situ
Reagan and colleagues79 described the morphologic charac-
teristics of a group of epithelial lesions that they considered
to be developmental carcinoma in situ. Comparable cytologic
features in women who ultimately were proved to have carci-
noma in situ were described by Koss and Durfee.44 Compared
with classic carcinoma in situ, cells more frequently lie sin-
gly than in arranged syncytial masses. When present in aggre-
gates, cells more often had a sheet-like rather than a syncytial
arrangement. In almost all instances, a significant admixture
of dysplastic cells was found in the smears. Average cell size
was between the mean sizes for dysplastic cells and carcinoma
in situ cells. Also, for the distribution of cell shapes, values
proved to be between the values found for dysplasias and those
for carcinomas in situ. Nuclear chromatin was predominantly
finely granular and unevenly distributed. The lesions usually
showed a more orderly growth pattern and a relatively low
mitotic activity.
The cellular characteristics of this so-called developmental car-
cinoma in situ were reminiscent of cells found in proliferating
reserve cells. The mean age at which this developmental cancer
in situ was found antedated the mean age of classic carcinoma
in situ by 5 years.
Fawdry found "early recurrences" of carcinoma in situ within
a year after hysterectomy in 0.9% of women who had a hyster-
ectomy after a diagnosis of carcinoma in situ.150 During subse-
quent long-term follow-up, invasive squamous cell carcinomas
of the vaginal vault were found in 0.3% of patients. These fig-
ures emphasize the importance of cytologic follow-up by vagi-
nal vault smears in women who have had a hysterectomy for a
severe epithelial abnormality of the cervix.
Co-occurrence of Dysplasia and Carcinoma in Situ
An area in which discrepancies in the grading of lesions still
occurs is the discrimination between severe dysplasia and car-
cinoma in situ. Although for the sake of a better understand-
ing of the biologic behavior of intraepithelial lesions it is still
previous page 177 ComprehensiveCytopathology 1104p 2008 read online next page 179 ComprehensiveCytopathology 1104p 2008 read online Home Toggle text on/off