Exploiting the
`"Toothpick Effect" of
`the Cytobrush by
`Plastic Embedding of
`Cervical Samples
`
`Mathilde E. Boon, M.D., Ph.D., M.I.A.C
`Pio Zeppa,.M.D.
`Elisabeth Ouwerkerk-Noordam, C.T.
`Lambrecht P. Kok, Ph.D.
`
`the combined spatida-Qjtobrush
`The introduction of
`sampling method has increased the diagnostic accuracy of
`cervical cytology. However,
`the smears from a small
`number of cases contained epithelial fragments dislodged
`by the "toothpick effect" of the Cytobrush. Due to their
`thickness,
`these epithelial fragments in the smears are
`very difficult (if not impossible) to scrutinize and are thus
`undiagnosable. The presence of only such fragments in
`smears led to false-negative diagnoses in two cases of
`invasive carcinoma. To solve this problem, these epithelial
`fragments were embedded in plastic, with thin sections
`
`prepared from the blocks. This paper presents the mor
`phologic features and diagnostic accuracy of 77 such
`problem cases (found among 50,000 cases with spatula-
`Cytobrush smears) to which this method was applied. In
`almost all cases,
`the diagnosis on the plastic-embedded
`sections matched the diagnosis on the colposcopically
`directed biopsy. Of these 77 patients, the biopsy diagnosis
`showed 5 severe dysplasias, 6 carcinomas in situ, 1
`squamous-cell carcinoma, 2 adenocarcinomas in situ and
`2 adenocarcinomas. In addition, data are presented con
`cerning the nuclear-size and shape-factor differences in
`smears versus plastic-embedded sections, and the stereo-
`logic consequences of smearing and cutting these epithe
`lial fragments are discussed. These plastic-embedded sec
`tions are well suited for use in quantitative microscopy, as
`well as for diagnostic purposes.
`
`The introduction of the combined spatuIa-Cytobnish
`sampling method for cervical cytology has dramati
`cally increased the proportion of smears containing
`endocervical cells submitted by general practi
`tioners.3 Since the implementation of this sampling
`technique in 1985, we have examined over 100,000
`smears prepared by this method. The Cytobrush is
`very successful in reaching and sampling transforma
`tion zones located in the endocervix, where the intra
`epithelial lesions, including carcinoma in situ, of the
`cervix often arise.6'14 Better sampling explains the
`great improvement in the detection of severe dyspla
`sia, carcinoma in situ and invasive carcinoma that the
`use of the Cytobrush has achieved.2'4
`Nonetheless, this technique also has a disadvan
`tage. The bristles of the Cytobrush easily dislodge
`epithelial fragments, both normal and abnormal. In
`many cases with cervical neoplasia, the smear con
`tains not only a monolayer of abnormal cells, but, in
`addition, many neoplastic epithelial fragments. These
`epithelial fragments, often containing 250 cells or
`more, cannot be spread on the glass slide. The nuclei
`overlap in these thick fragments and appear con-
`
`From the Leiden Cytology and Pathology Laboratory, Leiden,
`The Netherlands; the Cytopathology Service, II Faculty of Medi
`cine and Surgery, University of Naples, Naples, Italy; and the
`Institute for Theoretical Physics, Groningen University, Gro-
`ningen. The Netherlands.
`
`Dr. Boon is Senior Pathologist and Director, Leiden Cytology and
`Pathology Laboratory.
`
`Consiglio Nazionale delle Ricerche.
`Ms. Ouwerkerk-Noordam is Head Cytotechnologist, Leiden Cy
`tology and Pathology Laboratory.
`
`Dr. Kok is Senior Theoretical Physicist, University of Groningen.
`Address reprint requests to: Mathilde E. Boon, M.D, Ph.D.,
`M.I.A.C., Leids Cytologisch en Pathologisch Laboratorium,
`Postbus 16084, NL-230] GB Leiden,The Netherlands.
`
`Dr. Zeppa is Assistant Pathologist, II Faculty of Medicine and
`Surgery, University of Naples, and recipient of a grant from the
`
`Received for publication January 29, 1990.
`
`Accepted for publication July 3, 1990.
`
`0001-5547/91/3501-057/$02.00/0 © The International Academy of Cytology
`Acta Cytologica
`
`57
`
`

`

`i
`
`SSiii
`
`'
`
`RR
`
`Booji el nl
`
`Acta Cytologica
`
`densed. The chromatin structure can be evaluated
`¦ only if the nuclei at the edge of these fragments are
`somewhat spread. These neoplastic epithelial frag
`ments are also found in smears prepared with a
`spatula, but less frequently. The cytotechnologist may,
`in addition, find areas in these smears that contain a
`well-spread monolayer of aberrant cells whose nu
`clear features can be evaluated for diagnosis.
`After introducing the combined spatula-Cytobmsh
`sampling method in our laboratory, which processes
`specimens from more than 500 general practitioners,
`two false-negative diagnoses were made in cases with
`poorly differentiated squamous-cell carcinoma. In
`both cases, the cancer cells in the smears were present
`exclusively in the thick fragments; a monolayer of
`cancer cells was not present in these smears. Because
`these fragments were also present in large quantities
`in our correctly diagnosed cancer cases, we realized
`that the bristles of the Cytobrush have a "toothpick
`effect," easily dislodging these epithelial fragments
`from the endocervical canal. This is well illustrated by
`the scanning electron microscopy pictures of Glenthoj
`et al.8
`A solution to the problem of the epithelial frag
`ments was found by embedding them in plastic and
`preparing thin sections thereof. This solution pre
`sented itself because we were involved in research
`concerning the final dimensions of nuclei in plastic-
`embedded sections.7 Using these standardized tech
`niques, large nuclei can be obtained for precise eval
`uation of the nuclear details. By this embedding
`method, the disadvantage of the Cytobrush (the dis
`lodging of epithelial fragments) is turned into an
`advantage.
`
`a
`
`% JV
`
`Figure 1
`An undiagnosable epithelial fragment in a Cytobrush smear.
`The nuclear details are not visible (Papanicolaou stain, x430).
`
`Table I Plastic Embedding Procedure
`
`1. Place the cell suspension in a clean glass tube.
`2. Centrifuge at 1,500 RPM for 10 minutes.
`3. Decant the supernatant. Dry the sediment by putting the
`tubes upside down.
`4. Incubate with the GMA monomer, solution A, for at least
`45 minutes.
`5. Repeat steps 2 and 3.
`6. Mix solution A and B in the right proportions.
`7. Stir well for 2 minutes.
`8. Centrifuge the mixture at 1,500 RPM for 2 minutes.
`9. Place the cells in a mould, using a pipette. Add some more
`plastic if necessary.
`10. Put the blockholder on top of the material and cover it
`with paraffm.
`11. Allow the plastic to harden for at least one hour.
`
`Our cytotechnologists were instructed in the selec
`tion of cases whose smears contained undiagnosable
`epithelial fragments (Figure 1) but lacked a mono
`layer of abnormal cells. Only 0.05% of the spatula-
`Cytobrush smears (representing a very small propor
`tion of our patients) fell in this category. This paper
`reports our experience in using plastic embedding to
`solve this problem, thus capitalizing on the toothpick
`effect of the Cytobrush.
`
`Materials and Methods
`Patients
`
`From over 50,000 smears prepared by the spatula-
`Cytobrush sampling method, 77 (about 0.1%) con
`tained epithelial fragments and lacked a monolayer of
`abnormal cells. In each case, the general practitioner
`was asked to take two additional Cytobrush samples,
`one from the ectocervix and one from the endocervix,
`and suspend the cells in Leiden fixative (containing
`ethyl alcohol and polyethylene glycol). The samples
`were sent to the laboratory by mail in small plastic
`containers; thus, the samples were fixed at least for
`eight hours.
`
`Preparing Plastic-Embedded Sections
`The procedure for plastic embedding is given in Table
`I while the components of the solutions used are
`given in Table EL The plastic blocks were ready for
`cutting after one hour. Sections of 1 pm thickness were
`cut and stained by a modified Papanicolaou method.
`Standard diagnostic criteria were applied to the sec
`tions, but with care to identify differences between
`plastic-embedded and smeared cells.
`
`Morphometry
`To ascertain the differences in cell images between
`
`

`

`Volume 35
`Number 1
`jnnuary-Febnmry 3992
`
`Embedding Versus Smearing of Cytobrush Samples
`
`59
`
`Table 11 Components of Solutions A and B
`
`Solution A: Glycolmethacrylate with 200 ppm
`-
`hydroquinon (Merck-
`Schuchardt No. 800588)
`2-Butoxyethanol (Merck-
`Schuchardt No. 801554)
`Benzoylperoxide (with 20% H20)
`(Baker No. 9104)
`Solution B: Polyethylene glycol 400 (Fluka-
`AG No. 39430)
`.
`N,N-Dimethylaniline (Fluka-AG
`No. 39430)
`
`90 ml
`
`10 ml
`
`0.5 g%
`.
`15 parts
`
`1 part
`
`smears and plastic-embedded sections, nine cases
`with a known histologic diagnosis were chosen for
`morphometry. Since it was anticipated that the differ
`ences might depend on the cell type, various lesions
`were selected: four cases with a negative histologic
`diagnosis and one case each of severe dysplasia,
`squamous carcinoma in situ, invasive squamous-cell
`carcinoma, adenocarcinoma in situ and invasive ade
`nocarcinoma. For these nine cases,
`the cytologic
`smears and the corresponding plastic-embedded sec
`tions were used for morphometry. In each slide, 20
`nuclei
`in the abnormal epithelial
`fragments were
`measured with a MOP Videoplan (Kontron, Munich,
`Germany) equipped with a graphics tablet. A magni
`fication of 400 x was used. The following parameters
`were calculated for each nucleus: AREA, FORM ELL
`(the ratio between the short axis and the long axis)
`and FORM PE (4 7t * AREA/perimeter2). The mean
`values and standard deviations (SDs) of the parame
`ters were calculated for each slide.
`
`4S*
`
`¦m'
`
`Ch
`
`i-V
`
`J
`
`Figure 2
`An epithelial fragment showing moderate dysplasia with
`clumping of chromatin. Small nucleoli are also visible. Two bi-
`nucleated cells and some koilocytotic halos suggest an HPV in
`fection (plastic-embedded section, Papanicolaou stain, x430).
`
`Figure 3
`An epithelial fragment showing carcinoma in situ. The nuclei
`have coarse and irregularly distributed chromatin. The nucleoli
`are evident. Note the mitosis in the third upper layer of the
`fragment (plastic-embedded section, Papanicolaou stain, x430).
`
`Results
`Diagnostic Criteria
`
`Compared to those in smears, the nuclei in the thin
`sections were less hyperchromatic. The abnormal
`epithelial cells were mainly found in the tissue frag
`ments, particularly in cases with severe abnormalities
`(severe dysplasia, carcinoma in situ and invasive car
`cinoma). The contrast between heterocliromatin and
`euchromatin was more pronounced in the sections as
`compared to formalin-fixed nuclei. The nucleoli and
`deviations in their size and shape were well pre
`sented. Aberrations in nuclear shape were easily de
`tected in the thin sections. The mitoses and their
`location in the epithelial fragment were also impor
`tant criteria in the sections.
`The diagnostic criteria distinguishing the vari
`ous grades of intraepithelial neoplasia (mild dyspla
`sia, moderate dysplasia, severe dysplasia and car
`cinoma in situ) were the same as those used in
`paraffin-embedded sections, with the added advan
`tage that the chromatin changes were very well pre-
`• sented (Figures 2 and 3). In fragments of carcinoma in
`situ, mitotic figures were noted in different levels
`of the neoplastic epithelium. The diagnosis of inva
`sive squamous-cell carcinoma was based on the fol
`lowing (cytologic) criteria of the epithelial fragments:
`marked irregularity of the distribution of chromatin
`("empty" areas in the nucleus), relatively large abnor
`mally shaped nucleoli, marked variation in nuclear
`shape (Figure 4) and the presence of mitotic figures.
`In addition, necrosis was a diagnostic feature of
`invasive carcinoma. The diagnosis of adenocar-
`
`(cid:9)
`

`

`60
`
`Boon et al
`
`Acta Cytologica
`
`: i (
`
`l|^,7
`
`'tk
`
`¦
`
`*£
`
`Figure 4
`An epithelial fragment from a case of invasive squamous-cell
`carcinoma- Note the irregular shape of the nuclei, the coarse
`dumps of chromatin and the prominent and irregular nucleoli.
`An atypical mitosis (arrow) is also present (plastic-embedded
`section, Papanicolaou stain, x430).
`
`Figure 6
`In this fragment, the stroma cells are covered by benign squa
`mous cells. The stroma cells have small monotonous nuclei and
`scanty cytoplasm (plastic-embedded section, Papanicolaou
`stain, x430).
`
`cinoma was made when there were cylindrical
`cells with abnormally shaped nuclei with large nu
`cleoli (Figure 5). The stromal cells could be easily dis
`tinguished from the epithelial cells (Figure 6), which
`was often impossible in tissue fragments in the cor
`responding smears. Stromal cells covered with nor
`mal epithelial cells were often encountered in cases
`with a negative diagnosis on the plastic-embedded
`section.
`
`Figure 5
`Two epithelial fragments from a case of adenocarcinoma of the
`endocervical epithelium. The fragment on the right has features
`of adenocarcinoma in situ; the one on the left has features of ad-
`enocardnoma. Note the abnormal nuclear shape, the prominent
`nucleoli and the dispersed chromatin pattern in the adenocard-
`noma fragment (plastic-embedded section, Papanicolaou stain,
`x430).
`
`Diagnoses
`The diagnoses made on the 77 plastic-embedded
`sections are summarized in Table HI. In 30 (38.9%) of
`the cases, no abnormality was detected in the section.
`Dysplasia was diagnosed in 37 (48%) of the cases (28
`mild and moderate dysplasias and 9 severe dyspla
`sias). The diagnosis was squamous-cell carcinoma in
`situ in five (6.5%) of the cases, adenocarcinoma in situ
`in four cases (5.2%) and invasive squamous-cell car
`cinoma iir one case (1.3%).
`In 18 cases, cervical pathology was found in the
`histologic samples taken by the gynecologist (Table
`III). Tire diagnostic accuracy for severe dysplasia,
`squamous-cell carcmoma in situ and invasive
`squamous-cell carcinoma was good, but an invasive
`adenocarcinoma was diagnosed in two of the four
`cases with a cytodiagnosis of adenocarcinoma in situ
`on the plastic-embedded section. The plastic-
`embedded sections of these two cases were reviewed;
`in one case, epithelial fragments were identified with
`diagnostic features of invasive adenocarcinoma that
`were underestimated at the time of the initial diag
`nosis (Figure 6).
`
`Morphometry
`The results of the morphometry are given in Table IV.
`In all cases,
`the measured nuclei in the cytologic
`smear were much smaller than those in the corre
`sponding plastic-embedded section. The size differ
`ence was most pronounced in the case with severe
`dysplasia (52 sq pm versus 141 sq pm) and least
`pronounced in one of the negative cases (35 sq pm
`
`

`

`Volume 35
`Number 1
`Jauuary-Febrmry 1991
`
`Embedding Versus Smearing of Cytobrush Samples
`
`61
`
`Table III Findings in 77 Cases with Uudiagnosnble Epithelial Fragments in the Cytobrush Smears
`
`Diagnosis on plastic-embedded sections
`Mild/moderate
`dysplasia
`
`CIS
`
`AIS
`
`Severe
`dysplasia
`
`Squamous-
`cell carcinoma
`
`Total
`
`Histologic diagnosis
`
`. Negative
`
`Negative
`Mild/moderate dysplasia
`Severe dysplasia
`CIS
`AIS
`Squamous-cell carcinoma
`Adenocarcinoma
`No histology or negative
`follow up
`
`Total
`
`0
`0
`0
`0
`0
`0
`0
`
`30
`
`30
`
`4
`5
`0 .
`0
`0
`0
`.0
`
`19
`
`28
`
`1
`2
`4
`2
`0
`0
`0
`
`0
`
`9
`
`0
`0
`1
`4
`0
`0
`0
`
`0
`
`5
`
`*
`
`0
`0
`0
`0
`2
`0
`2
`
`0
`
`4
`
`0
`0
`0
`0
`0
`1
`0
`
`0
`
`1
`
`5
`7
`5
`6
`2
`1
`2
`
`49
`
`77
`
`versus 62 sq jim). The SDs for AREA are given as a
`percentage of the mean value of AREA.
`
`Discussion
`
`Light microscopic examination requires that three-
`dimensional cells be presented in two dimensions. In
`cytology, the reduction from three to two dimensions
`is accomplished by stretching the cells over the glass
`slide. Immediately after the cells are placed on the
`glass slide, both nucleus and cytoplasm spread and
`become flatter. To what extent the nucleus of a single
`cell flattens depends on its rigidity, on the elasticity of
`its membrane, and on the external forces, including
`surface tension, gravity and background material. For
`instance, for nuclei with identical volumes in thyroid
`aspirates, the sizes of "free" flattened nuclei were 1.5
`times those of corresponding "nonfree" nuclei (which
`were restricted in their spreading by the colloid in the
`aspirate).5 The spreading of nuclei in epithelial frag
`ments is restrictively influenced by the neighboring
`cells. The flattening of cell nuclei is severely ham
`pered in the larger epithelial fragments.
`When the epithelial fragments are embedded in
`plastic and thin sections are made, the three dimen
`sions of the cells are reduced to two without the
`nuclear shrinkage that occurs when cells are embed
`ded in paraffin.12'13'15'16 The combination of the
`Leiden fixative, containing ethyl alcohol and poly
`ethylene glycol, with plastic embedding in GMA is
`optimal. Due to the formation of a crude network of
`coagulated proteins by the fixative, the monomer can
`easily infiltrate into the cells.10
`We used morphometry to illustrate the differences
`of the nuclei in the epithelial fragments in the smears
`versus the sections. It is important to realize that,
`when we measure nuclei
`in cytologic smears, we
`
`measure their "shadows" on the glass slide while we
`measure the profile of a thin slice of the nucleus in the
`plastic-embedded sections (Figure 7). We considered
`size parameters and form parameters. If one cuts a
`population of egg-shaped nuclei in making a section,
`the average of the cut areas is approximately two
`thirds of the average of the areas of their projection
`("shadow") on the glass slide. Thus, one is tempted to
`think that the section values for AREA will be approx
`imately 0.7 times the smear values. However, for the
`nine cases studied, an average ratio of 1.9 was found
`instead of 0.7.
`The reason lies in the differences of the two prep-
`
`smear
`
`plastic
`
`Figure 7
`Nuclei of the same three-dimensional size (with a long axis A
`and short axis B) and with the same orientation in epithelial
`fragments can often give different two-dimensional images. In
`the smear (left), the resulting image is that of the projection
`("shadow") on a plane (glass slide). In the plastic-embedded
`section (right), the nucleus is cut and the image results from
`one thin slice only. Note that the image of the nucleus is larger
`in the smear than in the plastic-embedded section. However, in
`our material, the nuclear images were larger in the plastic-
`embedded sections (see Table IV) because there are additional
`factors involved that influence the size of nuclei in the smears
`(see text).
`
`

`

`62
`
`Boon et n\
`
`Acta Cytologica
`
`Table IV Morphometric Data on Nuclei in Epithelial Fragments in Cytobrush Smears and Corresponding Plastic-Embedded Sections
`
`Area
`
`Mean
`(sq pm)
`
`54.431
`67.206
`
`53.802
`81.229
`
`45.816
`74.561
`
`35.501
`62.402
`
`SD (%)
`
`21.8
`18.6
`
`15.6
`16.0
`
`23.1
`13.7
`
`31.9
`22.1
`
`Area
`ratio:
`section/
`
`smear
`
`1.235
`
`1.511
`
`1.627
`
`1.758
`
`Form ELL
`
`Form PE
`
`SD
`
`0.148
`0.095
`
`0.083
`0.075
`
`0.139
`0.113
`
`Mean
`
`0.891
`0.925
`
`0.925
`0.937
`
`0.927
`0.877
`
`SD
`
`0.077
`0.046
`
`0.078
`0.029
`
`0.053
`0.071
`
`Mean
`
`0.705
`0.752
`
`8.380
`13.002
`
`0.772
`0.660
`
`0.774
`0.775
`
`Sample
`
`Diagnosis
`
`? N
`
`o abnormality
`
`? N
`
`o abnormality
`
`? N
`
`o abnormality
`
`? N
`
`o abnormality
`
`Smear
`Section
`
`Smear
`Section
`
`Smear
`Section
`
`Smear
`Section
`
`Case
`
`no.
`
`1
`
`2
`
`3
`
`4
`
`0.121
`0.083
`
`0.104
`0.155
`
`0.097
`0.097
`
`0.085
`0.127
`
`0.093
`0.084
`
`0.087
`0.113
`
`0.931
`0.937
`
`0.917
`0.911
`
`0.934
`0.910
`
`0.943
`0.836
`
`0.899
`0.895
`
`0.948
`0.630
`
`0.052
`0.022
`
`0.052
`0.071
`
`0.042
`0.053
`
`0.030
`0.100
`
`0.038
`0.048
`
`0.092
`0.137
`
`0.784
`0.731
`
`0.755
`0.758
`
`0.837
`0.742
`
`0.746
`0.715
`
`0.818
`0-657
`
`2.682
`
`1.506
`
`2.517
`
`2.151
`
`2.156
`
`52.765
`141.514
`
`54.699
`82.388
`
`32.459
`81.712
`
`40.214
`86.498
`
`45.762
`98.646
`
`28.2
`30.8
`
`25.3
`36.8
`
`17.9
`34.9
`
`27.9
`29.8
`
`21.1
`35.1
`
`? D
`
`ysplasia
`
`? C
`
`IS
`
`? I
`
`nvasive carcinoma
`
`? A
`
`IS
`
`? A
`
`denocarcinoma
`
`Smear
`Section
`
`Smear
`Section
`
`Smear
`Section
`
`Smear
`Section
`
`Smear
`Section
`
`5
`
`6
`
`7
`
`8
`
`9
`
`aratory techniques. The shrinkage occurring during
`fixation depends on the percentage of alcohol in the
`fixative: the higher the percentage, the greater the
`shrinkage. The main components of both the spray
`fixative used for the smears and the Leiden fixative
`used for the plastic technique are the same: alcohol
`and polyethylene glycol. However, the percentage
`alcohol component is 70% in the spray fixative and
`50% in the Leiden fixative; thus, the spray fixative
`induces greater shrinkage. The shrinkage due to alco
`hol can be counteracted by the polyethylene glycol
`replacing the water in the cell, resulting in an increase
`in the nuclear volume. For optimal effect, the cells
`must be surrounded by polyethylene glycol, which is
`the case in the suspension step of the plastic tech
`nique. In the smear technique, the cells are in contact
`with polyethylene glycol for only a short period of
`time. Because tire solution bathes the cells exclusively
`from the surface, the much smaller alcohol molecules
`diffuse rapidly into the inner parts of the epithelial
`fragment, causing shrinkage. Thus, the shrinking ef
`fect of the fixative in the smear technique predomi
`nates to such an extent that the nuclei are smaller
`instead of larger, resulting in an AREA smear/AREA
`section ratio > 1. Interestingly, the ratio was lower for
`the four negative cases (1.5) than for the five positive
`cases (2.2), demonstrating that the net result of shrink
`
`ing plus swelling plus stretching differs between be
`nign cells and neoplastic cells.
`How does cutting nuclei
`in thin slices (plastic-
`embedded sections) compare to projecting them
`(smear)
`for
`the dimensionless form parameters
`FORM ELL and FORM PE? For egg-shaped nuclei
`(Figure 7) with a long axis A and a short axis B, one
`can compute11 that, theoretically on average, FORM
`ELL is 0.870 for cutting and 0.858 for projecting when
`B/A = 0.8. For B/A - 0.5, one obtains values of 0.690
`and 0.605, respectively. Thus, the observed values for
`plastic-embedded sections are anticipated to be more
`or less equal (but somewhat larger) than those for
`smears (see Table IV). For FORM PE, the same trends
`prevail, but the theoretical and practical values lie
`closer to 1. Only in the two cases with irregular nuclei
`(case 7 [invasive squamous-cell carcinoma] and case 9
`[adenocarcinoma]) were much lower values for
`FORM PE found in plastic sections (0.43), with mean
`values of 0.83 for case 7 and 0.63 for case 9.
`The nuclear images in plastic-embedded sections
`are morphometrically superior for many reasons. In
`addition, mitotic figures are readily identified (see
`Figure 3) and their location in the cell strata of the
`fragment can be ascertained. Nucleoli are readily
`noted (Figure 4) and can be automatically measured if
`special stains are used. In these thin sections, the
`
`

`

`Volume 35
`Number 1
`Jcmuary-Febmanj 1991
`
`observed number of nucleoli may be smaller than the
`actual number; stereologic methods can be used to
`calculate the true number.9 Polynucleolarity is a key
`factor in oncologic research. A current area of interest
`is the etiologic meaning of increases in the number of
`nucleoli, which may even surpass the number of
`nucleolar organizer zones in the nucleus.1 In addition,
`we found that the chromatin pattern of the nuclei in
`the plastic-embedded sections can be easily quan
`tified.17
`It is evident that the epithelial fragments contain
`important diagnostic information. Therefore, we sug
`gest that these fragments should be included in the
`analysis of samples by automated cytology: automa
`tion should not be focused exclusively on the analysis
`of the "easier" single cells and small cellular aggre
`gates.
`'
`To prepare a plastic-embedded section of epithelial
`fragments dislodged by the bristles of the Cytobrush,
`a new sample is necessary. Thus, the patient has to be
`invited for a repeat investigation. The preparation of a
`plastic-embedded section is much more complicated
`than the preparation of a simple smear. Therefore, the
`use of plastic embedding in laboratories involved in
`routine cytologic examinations should be reserved for
`the solution of problem cases. Only a small fraction of
`the women sampled with the Cytobrush (around
`0.1%) needed this special method in our laboratory. In
`this relatively small group of 77 problem cases, histo
`logic and cytologic follow-up showed only 35 to be
`without changes of the cervical epithelium. The re
`maining 42 cases included 18 cervical intraepithelial
`squamous neoplasias, 1 adenocarcinoma hi situ, 1
`squamous-cell carcinoma and 2 adenocarcinomas of
`the endocervical epithelium. Two of the three inva
`sive carcmomas and two of the eight in situ car
`cinomas were of glandular origin. This finding is in
`keeping with our experience that we are detecting
`more glandular neoplasms since the introduction of
`the spatula-Cytobmsh sampling method in our prac
`tice.4
`In summary, the "toothpick effect" of dislodging of
`neoplastic epithelial fragments by the Cytobrush can
`be fully exploited by embedding Cytobrush samples
`in plastic and cutting thin sections thereof. These
`sections are superior for both diagnostic and quantita
`tive microscopy.
`
`References
`1. Alberts B, Bray D, Raff M, Roberts K, Watson JD: Molecular
`Biology of the Cell. New York, Garland Publishing, 1983, pp
`83-85
`
`Embedding Versus Smearing of Cytobrush Samples
`
`63
`
`2. Alons-van Kordelaar JJM, Boon ME: Diagnostic accuracy of
`squamous cervical
`lesions studied in spatula-Cytobrush
`smears. Acta Cytol 32:801-804, 1988
`•
`3. Boon ME, Alons-van Kordelaar JJM: Effect on detection of
`various cervical
`lesions using the combined spatula and
`Cytobrush sampling method. J Jpn Soc Clin Cytol 12:1030
`1033, 1987
`4. Boon ME, Alons-van Kordelaar JJM, Rietveld-Scheffrs PEM:
`Consequences of the introduction of combined spatula and
`Cytobrush sampling for cervical cytology: Improvements in
`smear quality and detection rates. Acta Cytol 30:264-270,
`1986
`
`5. Boon ME, Kok LP: An explanation for the reported variability
`of nuclear areas in air-dried Romanowsky-Giemsa-stained
`smears of follicular tumors of the thyroid. Acta Cytol 31:527
`530, 1987
`
`6. Burghardt E: Latest aspects of precancerous lesions in squa
`mous and columnar epithelium of the cervix. Int J Gynecol
`Obstet 8:573-580, 1970
`7. Gerrits PO, van Leeuwen MBM, Boon ME, Kok LP: Floating
`on a water bath and mounting glycol methacrylate and
`hydroxypropyl methacrylate sections influence final dimen
`sions. J Microsc 145:107-113, 1987
`8. Glenthoj A, Bostofte E, Rank F: Brush cytology from the
`uterine cervix. Acta Obstet Gynecol Scand 65:689-691, 1986
`9. Gramsbergen AG, Kok LP, Poortema K, Schaafsma W: Deter
`mining nucleolar multiplicity and cell number from sectional
`data. J Microsc 145:69-87, 1985
`10. Horobin RW: Understanding staining of water-miscible resin
`sections. Sheffield, Datascope Services, 1989
`11. Kok LP: 100 Problems of My Wife and Their Solution in
`Theoretical Stereology. Leiden, Coulomb Press Leyden, 1990.
`12. Kok LP, Boon ME, Ouwerkerk-Noordam E, Gerrits PO: The
`application of a microwave technique for the preparation of
`cell blocks from sputum. J Microsc 144:193-199, 1986
`13. Kung ITM, Yuen RWS: Fine needle aspiration of the thyroid:
`Distinction between colloid nodules and follicular neoplasms
`using cell blocks and 21 -gauge needles. Acta Cytol 33:53-60,
`1989
`
`14. Patten SF: Morphologic subclassification of preinvasive cervi
`cal neoplasia. In Compemdtum on Diagnostic Cytology. Fifth
`edition. Edited by GL Wied, LG Koss, JW Reagan, CM Keeb-
`ler, Chicago, Tutorials of Cytology, 1983, pp 108-117
`15. Van de KantHJG, de Rooij DG, Boon ME: Microwave stabiliz
`ation versus chemical fixation: A morphometric study in
`glycolmethacrylate- and paraffin-embedded tissue. Histo-
`chem J 22:335-340, 1990
`16. Van der Poel HG, Boon ME, Kok LP, Tolboom J, van der
`Meulen B, Ooms ECM: Can cytomorphometry replace histo-
`morphometry for grading of bladder tumours? Virchows
`Archiv [Pathol Anat] 413:249-255, 1988
`17. Van der Poel HG, Boon ME, Kok LP, van der Meulen EA, van
`Caubergh RD, de Bruijn WC, Debruyne FMJ: Morphometry,
`densitometry and chroma tin-pattern analysis for automated
`grading of plastic-embedded bladder carcinomas. Analyt
`Quant Cytol Histol (in press)
`
`