Insulinoma-associated protein 1 is a novel diagnostic marker of small cell lung cancer in bronchial brushing and cell block cytology from pleural effusions: Validity and reliability with cutoff value.

BACKGROUND:In recent years, insulinoma-associated protein 1 (INSM1) has been shown to be a key regulator of neuroendocrine development, and it has been evaluated for diagnostic use in some organs. METHODS:To evaluate the relationship between INSM1 and synaptophysin, and to confirm the cutoff value using receiver operating characteristic curve (ROC) analysis, the authors performed INSM1 immunocytochemistry using cell block (CB) samples from 53 cases of bronchial brushings and 32 cases of pleural effusions (29 small cell lung cancer [SCLC] specimens and 56 non-small cell lung cancer specimens). The marker expression ratio was calculated by counting the positive tumor cells, and the tumor proportion score (TPS) was applied. RESULTS:INSM1 was expressed in all SCLC specimens, but generally was expressed in <10% of tumor cells in adenocarcinomas. In bronchial brushing samples of SCLC, the INSM1 TPS was 37.5% (staining of >50%), 25.0% (staining of 25%-50%), 29.5% (staining of 10%-25%), and 8.3% (staining of 1%-10%). There were 3 cases (12.5%) with no detectable synaptophysin expression, although the correlation between nuclear INSM1 expression and cytoplasmic synaptophysin expression was statistically significant (P < .001). Receiver operating characteristic curve analysis indicated that 8.68% was the best cutoff value for INSM1, and the sensitivity and specificity between SCLC and non-small cell lung cancer for expression of INSM1 were 95.8% and 100.0%, respectively, in bronchial brushing samples at that cutoff value. CONCLUSIONS:INSM1 is a novel diagnostic marker for SCLC, and is useful in both bronchial brushing and pleural effusion cytology specimens. Because INSM1 generally is expressed in <10% of tumor cells in adenocarcinomas, determining an accurate cutoff value for INSM1 is important in the diagnosis of SCLC.

KEYWORDS: {{ getKeywords(articleDetailText.words) }}