Ultrarapid Ki-67 immunostaining in frozen section interpretation of gliomas

Abstract

Background: Astrocytic tumours, the most common gliomas, are often classified intraoperatively using standard morphological staining. The final diagnosis and grading of gliomas on paraffin wax sections is often assisted by Ki-67 immunohistochemistry, but standard immunostaining protocols take too long to be used intraoperatively.

Aims: To investigate a new rapid Ki-67 immunohistochemical test for its use in an intraoperative setting.

Methods: The new Ki-67 immunostaining (Ultrarapid-Ki67) method on frozen sections can be carried out in 10 minutes. Thirty four pilocytic and diffuse astrocytomas were immunostained by rapid Ki-67 and results were compared with corresponding MIB-1 staining, histological grading, and prognosis.

Results: The staining protocol was practical to perform and the results were morphologically and quantitatively indistinguishable from those after immunostaining with MIB-1, an antibody recognising Ki-67 in paraffin wax embedded tissue. A comparison of Ultrarapid-Ki67 and MIB-1 immunostaining of paraffin wax sections showed almost identical quantitative correlation in astrocytic gliomas (r = 0.916; p<0.001). The Ultrarapid-Ki67 indices (percentage of positive cells) of low grade (I/II) astrocytomas ranged from 0% to 6.1%, whereas those of representative high grade (III/IV) tumours were significantly higher (range, 5.6-45%; p<0.001). The best prognostic cutoff point for Ultrarapid-Ki67 was 7.5%, which divided diffuse grade II-IV astrocytomas into significantly differing subsets (p = 0.0008).

Conclusion: Ultrarapid-Ki67 immunostaining is a useful adjunct to morphological diagnosis and grading of astrocytic tumours, and as a fast test (approximately 10 minutes for staining plus three to four minutes for scoring), it could be used in routine intraoperative diagnosis of gliomas and other neoplastic diseases.

Figure 1

Immunostaining in a frozen section of (A) low grade and (B) high grade glioma using the Ultrarapid-Ki67 kit. An additional clinical case demonstrating the usefulness of intraoperative Ki-67 is shown in (C–E). (C) Haematoxylin and eosin staining of a cortical tumour was inconclusive, leaving the differential diagnosis of primary brain tumour (preferentially meningioma, one mitosis/10 high power fields) and metastatic carcinoma. (D) A high Ki-67 labelling index ruled out low grade and benign tumours. (E) Additional rapid intraoperative pan-cytokeratin staining defined the tumour finally as a metastatic carcinoma.

Figure 2

Growth fractions (percentage of Ki-67 positive cells) by rapid frozen section and standard paraffin wax embedded section Ki-67 immunostaining in matched pairs of human astrocytoma samples. (A) Comparison between Ultrarapid-Ki67 values of frozen sections and corresponding MIB-1 values of paraffin wax embedded control sections analysed by the same observer (r  =  0.916; p < 0.001). The regression line and its 95% confidence intervals are plotted in solid and dashed lines, respectively. (B) Comparison between Ultrarapid-Ki67 index of frozen sections and selected tumour areas showing highest MIB-1 index of paraffin wax embedded tumour material analysed by computer assisted image analysis (r  =  0.790 p < 0.001). Open symbols, low grade tumour in final diagnosis; filled symbols, high grade tumour in final diagnosis.

Figure 3

Receiver operating characteristics curve showing prognostic true and false positivity rates of the Ki-67 indices. The predictive properties (%) are calculated on the basis of three year survival of the 34 patients divided into six subgroups of approximately equal size. The patient groups are placed in the curve according to decreasing proliferation: patients with the highest tumour proliferation are at the bottom left corner of the curve, and patients with the lowest tumour proliferation are at the top right corner of the curve. The most efficient cutoff point of each Ki-67 index is shown.