A novel evaluation method for Ki-67 immunostaining in paraffin-embedded tissues

The Ki-67 labeling index is traditionally used to investigate tumor aggressiveness. However, no diagnostic or prognostic value has been associated to the heterogeneous pattern of nuclear positivity. The aims of this study were to develop a classification for the patterns of Ki-67-positive nuclei; to search scientific evidence for the Ki-67 expression and location throughout the cell cycle; and to develop a protocol to apply the classification of patterns of Ki-67-positive nuclei in squamous epithelium with different proliferative activities. Based on empirical observation of paraffin sections submitted to immunohistochemistry for the determination of Ki-67 labeling index and literature review about Ki-67 expression, we created a classification of the patterns of nuclear positivity (NP1, NP2, NP3, NP4, and mitosis). A semi-automatic protocol was developed to identify and quantify the Ki-67 immunostaining patterns in target tissues. Two observers evaluated 7000 nuclei twice to test the intraobserver reliability, and six evaluated 1000 nuclei to the interobserver evaluation. The results showed that the immunohistochemical patterns of Ki-67 are similar in the tumoral and non-tumoral epithelium and were classified without difficulty. There was a high intraobserver reliability (Spearman correlation coefficient > 0.9) and moderate interobserver agreement ( k = 0.523). Statistical analysis showed that non-malignant epithelial specimens presented a higher number of NP1 (geographic tongue = 83.8 ± 21.8; no lesion = 107.6 ± 52.7; and mild dysplasia = 86.6 ± 25.8) when compared to carcinoma in Situ (46.8 ± 34.8) and invasive carcinoma (72.6 ± 37.9). The statistical evaluation showed significant difference ( p < 0.05). Thus, we propose a new way to evaluate Ki-67, where the pattern of its expression may be associated with the dynamics of the cell cycle. Future proof of this association will validate the use of the classification for its possible impact on cancer prognosis and guidance on personalized therapy.