Background: Assessment of tumor burden changes is essential for the management of patients with neuroendocrine gastrointestinal (GI) tumors. Chromogranin A (CgA) is a tumor marker for such tumors; however, to the authors' knowledge, there is little information on whether serial assessments can assess changes in tumor burden. In this prospective study of patients with gastrinomas, serial changes in serum CgA levels were compared with changes in levels of the specific tumor marker gastrin to determine whether they reflected changes in tumor burden.
Methods: In 72 consecutive patients, the mean CgA and gastrin levels from three determinations were measured on each visit. Changes in markers were correlated with changes in tumor burden determined by imaging. By assessing daily changes, significance changes in CgA and gastrin levels were determined.
Results: During 103 follow-up visits (mean, 9.6 months), an increased tumor size occurred in 25% of patients, no change occurred in 62% of patients, and a decrease occurred in 13% of patients. In patients who had increasing tumor size, CgA levels increased numerically in 77% of patients, gastrin levels increased in 54% of patients, and the increases were significant in 60-80% of patients. In patients who had tumor stabilization, CgA levels in 63% of patients and gastrin levels in 73% of patients did not show a significant change. Decreased tumor size postresection showed a significant decrease in CgA and gastrin levels in all patients. The sensitivity of CgA and gastrin was as follows: sensitivity for detecting an increase, 62% for CgA and 31% for gastrin; sensitivity for detecting no change, 42% for CgA and 75% for gastrin; and sensitivity for detecting a decrease in tumor size, 85% for CgA and 85% for gastrin. The specificity varied from 53% to 99% for CgA and from 49% to 93% for gastrin.
Conclusions: In patients with gastrinomas, serum CgA and gastrin levels varied considerably from day to day, and this must be taken into consideration. Both markers had low sensitivity and specificity for detecting tumor increases and stabilization. For large tumor decreases postresection, both markers had high sensitivity and specificity. The current results suggest that these markers do not have sufficient sensitivity to replace serial imaging studies for detecting important smaller changes in tumor burden in patients with gastrinomas.