A comparative study of cell blocks versus cytospins and Conventional smears in effusions and cyst fluids cytology at Kenyatta National Hospital
Abstract
Excess fluids of the serous membranes are often the first clinical symptom of
malignant tumors or of their metastatic manifestation. In one third of malignant effusions
cytologygives the first indication of malignancy. Accurately diagnosing cells as being either
malignant,benign 'reactive mesothelial cells' or adenocarcinoma in serous fluids is a
common diagnostic challenge. To improve diagnostic accuracy, most laboratories use two or
more preparation methods.
Objective: The objective of this study is to compare cell block preparations versus cytospins
and conventional smears in cytodiagnosis of effusions and cyst fluids.
Design: This is a descriptive cross-sectional study.
Subjects: Body fluids received in the KNH cytology laborotary that met the inclusion criteria
forthe duration of the study.
Setting: KNH Histology/Cytology laboratory.
Methods: Fluid specimens received in the laboratory were centrifuged and two smears were
prepared from exfoliated cells from the deposit. The remaining sample was divided into two
equal volumes, one volume was processed using Shandon cytospin and the second volume
was processed as a cell block.
Ethical clearance was obtained from Kenyatta National Hospital/ University of Nairobi
Ethics and Research Committee {KNH/UON-ERC) before carrying out the study. Consent to
use the residual fluid samples was sought from the KNH Head of Department Laboratory
Medicine through the cytology laboratory in charge. Analysis was done using SI'SS version
17.0 for Windows (SPSS Inc.). The test statistics for testing for the level of significance were
McNemar test and Chi-square.
Results
In this study cell blocks had 16% samples with definite malignancy, cytospins had 18% and
conventional smears 15%. In the suspicious category cell blocks had 6f%, cytospins 5%.
conventional smears had 8%. Benign category had cell blocks with 64%, cytospins 65% and
conventional smears had 69% while in the insufficient category the highest was in cell blocks
with 13% followed by cytospins with 1 I% and conventional smears had only 7°/u. When the
cell blocks were interpreted together with the cytospins and conventional smears the final
cytodiagnosis was, malignant cases 19%, suspicious for malignancy 7''/0, benign 71% and
insufficient were 2°1l1.
Conclusion
Cytospins are superior to both conventional smears and cell blocks in couccnuutiug cells.
Cell blocks demonstrated the best architectural and cellular arrangement.
Cell blocks when interpreted in association with cytospins and conventional smears
increased the diagnostic accuracy in this study by 8.2%.
Cytodiagnosis is shown to be dependent on cellularity, architectural and cellular arrangement
and clinical information.
Recommendations
Cytospin method of processing body fluids should be adopted as a routine procedure.
A minimum volume of 20 mls is recommended when multiple cytological techniques will be
done for fluids.
Cell blocks can be prepared in selected eases after review of the routine smears for better
characterization of the tumor and where special stains and immunohistochemistry is required.