SENSITIVITY AND SPECIFICITY OF INSM1 IN DIFFERENTIATING NEUROENDOCRINE CARCINOMA OF THE HEAD AND NECK FROM NONKERATINIZING NASOSPHARYNGEAL CARCINOMA AND P16-POSITIVE OROPHARYNGEAL SQUAMOUS CELL CARCINOMA

Background: Neuroendocrine carcinomas (NECs) rarely arise in the head and neck region. Their diagnosis presents challenges due to morphological overlap with other entities, particularly nonkeratinizing nasopharyngeal carcinomas (NK-NPC) and p16-positive oropharyngeal squamous cell carcinomas (p16-positive OPSCC), both of which are prevalent in Thailand. Insulinoma-associated protein 1 (INSM1) is a relatively new marker that has demonstrated favorable sensitivity and specificity in various organs. However, despite its promising potential, there is a paucity of studies investigating its utility in the head and neck region compared to other anatomical sites, especially in Thailand. Objectives: This study aimed to evaluate the diagnostic performance of INSM1 in distinguishing NECs of the head and neck region from NK-NPC and p16-positive OPSCC by comparing its sensitivity and specificity with classic neuroendocrine markers, chromogranin A (CGA) and synaptophysin (SYN). Methods: This retrospective cohort study analyzed 14 samples of NEC and 109 samples, comprising 93 NK-NPC and 16 p16-positive OPSCC cases. Immunohistochemical (IHC) staining for INSM1, CGA, and SYN was performed on all cases. Receiver Operating Characteristic (ROC) curve analysis was utilized to determine the optimal cutoff point for INSM1 positivity, maximizing both sensitivity and specificity. Results: INSM1 demonstrated an overall sensitivity of 92.9% for head and neck NECs, comparable to SYN (100.0%, p = 0.001) but significantly higher than CGA (78.6%, p = 0.006). All three markers (INSM1, CGA, and SYN) achieved 100.0% specificity in differentiating NECs from NK-NPC and p16-positive OPSCC. ROC analysis determined an optimal cutoff of 75% tumor cell positivity for INSM1, with a Youden’s index of 0.93 and an Area Under the Curve (AUC) of 0.952, indicating excellent diagnostic accuracy. Notably, one case of Epstein-Barr virus (EBV)-positive NK-NPC exhibited INSM1 positivity in 40% of tumor cells with moderate to strong intensity. Conclusion: INSM1 exhibits good sensitivity and excellent specificity for head and neck NECs, comparable to or surpassing those of CGA and SYN, respectively. While its high specificity is valuable, the observed positivity in a subset of NK-NPC cases, even below the optimal cutoff, suggests that INSM1 should not be used as a standalone diagnostic marker for NECs. Caution is advised when interpreting INSM1 staining in less than 75% of tumor cells, as this may reduce the reliability of a positive finding. A comprehensive panel that includes classic neuroendocrine markers and, where appropriate, EBER in situ hybridization remains crucial for accurate diagnosis.