Expression of insulinoma-associated protein 1 in non–small cell lung cancers: a diagnostic pitfall for neuroendocrine tumors

• Published in: Human pathology 2021-09, Vol.115, p.104-111
• Main Authors: Wang, Minhua, Abi-Raad, Rita, Baldassarri, Rebecca, Adeniran, Adebowale J., Cai, Guoping
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Biomarkers, Tumor - analysis; Biomarkers, Tumor - metabolism; Biopsy; Carcinoma, Non-Small-Cell Lung - diagnosis; Cloning; Diagnosis, Differential; Female; Humans; INSM1; Lung cancer; Lung Neoplasms - diagnosis; Male; Middle Aged; Neuroendocrine tumor; Neuroendocrine tumors; Neuroendocrine Tumors - diagnosis; Non-neuroendocrine tumor; Non–small cell cancer; Pathology; Repressor Proteins - analysis; Repressor Proteins - biosynthesis
• ISSN: 0046-8177; 1532-8392
• DOI: 10.1016/j.humpath.2021.06.006

Insulinoma-associated protein 1 (INSM1) has been reported as a highly sensitive and specific marker of neuroendocrine tumors. INSM1 expression has also been reported, although uncommonly, in non-neuroendocrine tumors. This study aimed to elucidate potential nonspecific INSM1 expression in non–small cell non-neuroendocrine lung cancers (NSCNELCs), especially in squamous cell carcinomas (SqCCs) with basaloid features to avoid diagnostic pitfalls. Tissue microarrays (TMAs) were constructed for 324 NSCNELCs, including 196 adenocarcinomas (AdCs), 86 SqCCs, and 42 other NSCNELCs. In addition, 38 whole-tissue sections of SqCCs with basaloid features were examined. INSM1 immunostain was semiquantitively evaluated based on the percentage of nuclear staining in tumor cells, categorized as negative, focal (10% tumor cells). Among 324 TMAs, 6.2% (20/324) were positive for INSM1, 4.9% (16/324) were focal, and 88.9% (289/34) were negative. Of 196 AdCs, 5.1% (10/196) were positive for INSM1, 4.7% (9/196) were focal, and 90.3% (177/196) were negative. Of 86 SqCCs, 9.3% (8/86) were positive for INSM1, 5.8% (5/86) were focal, and 84.9% (73/86) were negative. Of the remaining 42 NSCNELCs, 4.8% (2/42) were positive for INSM1, 4.8% (2/42) were focal, and 90.4% (38/44) were negative. Among 38 cases of whole-tissue sections of SqCCs with basaloid features, 15.8% (6/38) were positive for INSM1, 18.4% (7/38) were focal, and 65.8% (25/38) were negative. Our study demonstrates that INSM1 is expressed in a significant subset of NSCNELCs, suggesting caution in interpreting INSM1 staining, especially with limited samples. INSM1 should not be used as a stand-alone neuroendocrine marker in differentiating primary lung tumors. •Insulinoma-associated protein 1 (INSM1) is expressed in a significant subset of non–small cell non-neuroendocrine lung carcinomas.•Caution is needed in interpreting INSM1 staining, especially with limited samples, such as biopsies and cytology cell blocks.•INSM1 is best not to be used as a stand-alone neuroendocrine marker in classifying primary lung tumors.•The use of a combination of INSM1 and another neuroendocrine marker as well as other lineage markers is advised.