Errors, limitations, and pitfalls in the diagnosis of central and peripheral nervous system lesions in intraoperative cytology and frozen sections

Abstract

Context: Intraoperative cytology and frozen section play an important role in the diagnosis of neurosurgical specimens. There are limitations in both these procedures but understanding the errors and pitfalls may help in increasing the diagnostic yield.

Aims: To find the diagnostic accuracy of intraoperative cytology and frozen section for central and peripheral nervous system (PNS) lesions and analyze the errors, pitfalls, and limitations in these procedures.

Settings and design: Eighty cases were included in this prospective study in a span of 1.5 years.

Materials and methods: The crush preparations and the frozen sections were stained with hematoxylin and eosin method. The diagnosis of crush smears and the frozen sections were compared with the diagnosis in the paraffin section, which was considered as the gold standard.

Statistical analyses used: Diagnostic accuracy, sensitivity, and specificity.

Results: The diagnostic accuracy of crush smears was 91.25% with a sensitivity of 95.5% and specificity of 100%. In the frozen sections, the overall diagnostic accuracy was 95%, sensitivity was 96.8%, and specificity was 100%. The categories of pitfalls noted in this study were categorization of spindle cell lesions, differentiation of oligodendroglioma from astrocytoma in frozen sections, differentiation of coagulative tumor necrosis of glioblastoma multiforme (GBM) from the caseous necrosis of tuberculosis, grading of gliomas in frozen section, and differentiation of the normal granular cells of the cerebellum from the lymphocytes in cytological smears.

Conclusions: Crush smear and frozen section are complimentary procedures. When both are used together, the diagnostic yield is substantially increased.

Keywords: Central nervous system (CNS) tumors; crush smear; frozen section; intraoperative diagnosis; pitfalls.

Figure 1

(a) Crush smear showing spindle cells of fibroblastic meningioma arranged in fascicles without meningothelial whorls, which was misinterpreted as schwannoma (H and E, ×200) (b) Crush smear of the same case showing the fascicular arrangement of spindle cells (H and E, ×400) (c) Frozen section of fibroblastic meningioma misinterpreted as schwannoma showing the spindle cell nature of the lesion due to thick sections (H and E, ×200)

Figure 2

(a) Crush smear showing coagulative necrosis of glioblastoma multiforme, which was misinterpreted as caseous necrosis of tuberculoma (H and E, ×200) (b) Frozen section showing only coagulative necrosis of glioblastoma multiforme with no malignant cells, which was misinterpreted as tuberculoma (H and E, ×200)