Correlating peptidylglycine alpha-amidating monooxygenase (PAM) expression with clinicopathologic variables in gastrointestinal (GI) neuroendocrine tumors (NETs).

e16243 Background: Peptidylglycine alpha-amidating monooxygenase (PAM) is a hypoxia-responsive protein which is responsible for peptide alpha-amidation, important in neuroendocrine hormone maturation. A prior study of well-differentiated neuroendocrine tumors of multiple sites, both GI and non-GI, has shown that PAM expression by immunohistochemistry correlates with important clinical and pathologic parameters. We assessed PAM expression by immunohistochemistry using two separate scoring systems in patients with well-differentiated neuroendocrine tumors (NETs) exclusively of the GI tract and assessed its correlation with clinical and pathologic variables. Methods: A tissue microarray was constructed from 58 patients with NETs of the GI tract. Two separate scoring systems were used, the immunoreactive score (IRS) and one as described by Horton et al (2020). The IRS score was derived by scoring for intensity (range 0-3) and percent of cells staining (range 0-4), multiplying these to give an overall IRS between 0 and 12. These results were dichotomized into 0-1 (negative) and 2-12 (positive). The score as described by Horton et al is a 4-tiered system: 0, no staining; 1, very faint and scarce staining; 2, staining easily visible in > 50% of tissue; 3, moderately dark and intense staining across 50% of tissue; 4, very intense dark staining across > 50% of tissue. These were then divided into negative (0), low (1) and high (2-4) scores. Clinical and pathologic variables were collected from the clinical records and correlated with PAM staining. Results: Patients had NETs involving the stomach (n = 1), duodenum (n = 3), ileum (n = 18), small bowel (n = 12), cecum (n = 2), and pancreas (n = 13). Twenty-one patients (36.2%) had 1 metastatic site, 12 patients (20.7%) had > = 2 metastatic sites and 25 patients (43.1%) lacked metastases. IRS and Horton scores were compared and were correlated (p = 0.007). By IRS, 9 patients (31%) had negative staining, and 20 patients (69%) had positive PAM staining. PAM staining did not correlate with functional status of the tumor (p = 1), synchronicity of disease (p = 1), number of metastatic sites (p = 0.69), Octreotide receptor positivity (p = 0.72) or blood chromogranin A (p = 0.58) and urine 5 HIAA levels (p = 0.52). PAM scoring also did not correlate with Ki-67 of primary tumors (p = 0.57). Conclusions: In our GI NET cohort, PAM staining did not correlate with any of the assessed clinical variables or with Ki-67 of primary tumors. A previous study of NETs involving more body sites found that PAM staining did inversely correlate with Ki-67 scoring. This difference may be due to our assessment of solely gastrointestinal NETs.