Low sensitivity of chromogranin A in the diagnosis of insulinoma: a single-center study

Abstract

Background: Insulinoma is a neuroendocrine tumor (NET), with its main clinical manifestation being the hypoglycemic syndrome. The symptoms of hypoglycemia are nonspecific, and therefore, in most cases, the diagnosis is made untimely. The Russian clinical guidelines for the diagnosis and treatment of NET suggest as a diagnostic test that the universal circulating marker of all NET, chromogranin A (CgA) be determined. However, the literature data on the sensitivity of CgA in the diagnosis of insulinoma are contradictory. Aim: To evaluate the diagnostic effectiveness of the CgA test in the diagnosis of insulinoma. Materials and methods: This was a hospital-based single-center, cross-sectional comparative (first step) and prospective non-comparative (second step) study conducted from 2016 to 2022. During the first part of the study, we determined serum CgA in 120 patients with suspected non-diabetic hypoglycemia and compared its levels in the patients with and without confirmed insulinoma (n = 87 and n = 33, respectively). During the second study step, CgA was measured in the insulinoma patients at 6 [4.0; 7.0] months after surgery. The CgA levels at baseline and post-surgery were analyzed in 74 patients (those with recurring non-diabetic hypoglycemia were excluded from the analysis). Results: In the study subjects without insulinoma, the median CgA level was 0.7 [0.5; 1.1] (range, 0.1 to 2.0) nmol/l and the difference (with Bonferroni adjustment) from its levels in the patients with insulinoma before surgery was non-significant (1.0 [0.7; 1.4], range, 0.1 to 8.5 nmol/l, р = 0.045). The CgA concentration in the insulinoma patients after surgery was 0.9 [0.7; 1.2], range, 0 to 1.9 nmol/l and significantly differed from that at baseline (1.0 [0.7; 1.4], range, 0.1 to 8.5 nmol/l, p = 0.012, Wilcoxon test). In the patients with insulinoma before surgery the CgA levels exceeding the generally accepted reference range ( 2 nmol/l) was found in 11.5% (n = 10), with its median level of 2.5 [2.3; 4.1], range 2.3 to 8.5 nmol/l. There were no significant associations between the CgA levels and localization, tumor numbers, their size and malignization grade, insulin and proinsulin values, and duration of fasting. The sensitivity and specificity of the CgA test were 12% [95% confidence interval [CI]: 6%–20%] and 100% [95% CI: 97%–100%], respectively. The prognostic value of a positive result (PVPR) was 100% [95% CI: 69%–100%] and the prognostic value of a negative result (PVNR) 30% [95% CI: 28%–32%]. Conclusion: As a diagnostic test, CgA has high specificity and prognostic value of a positive result. However, the uncertainty of the prognostic value of a negative result is unacceptably high. A special study is required to clarify the operational characteristics of CgA in insulinoma.