Diagnostic utility of Gallium-68-somatostatin receptor PET/CT in ectopic ACTH-secreting tumors: a systematic literature review and single-center clinical experience

Abstract

Purpose: Tumors causing ectopic Cushing’s syndrome (ECS) are often not visible with conventional imaging. Gallium-68-DOTATATE, DOTATOC, and DOTANOC positron emission tomography/computed tomography (68Ga-SSTR PET/CT) reportedly exhibits greater sensitivity in identifying an ECS source, however, evidence is limited to mainly case reports and a few small retrospective studies. Previous systematic ECS imaging review has shown 68Ga-SSTR PET/CT sensitivity is similar to CT (81.8%) in histologically-proven cases and is 100% in covert-cases, however, the number of patients was small and no occult cases were reported. Methods: We performed a systematic literature review of 68Ga-SSTR PET/CT use in ECS patients. We also report 6 consecutive patients with confirmed active and occult ECS who underwent 68Ga-DOTATATE PET/CT and were followed at our institution between 2014 and 2019. Results: We identified 33 articles (23 case-reports, 4 case-series, 5 retrospective studies and 1 prospective study) detailing 68Ga-SSTR PET/CT in 69 ECS patients. Overall 68Ga-SSTR PET/CT sensitivity was 64.0%, while in histologically confirmed cases (67 lesions), sensitivity was 76.1%. There were two false-positives cases, both in the adrenal glands. In covert cases, 68Ga-SSTR PET/CT identified 50% of lesions. There were ten occult cases where all imaging failed to identify an adrenocorticotropic hormone source; source remains unknown. In our case series, 68Ga-DOTATATE PET/CT showed decreased uptake in pancreatic neuroendocrine tumor in one patient and did not help identify an ECS source in 5 patients. Conclusion: Both this systematic literature review, the largest to date, and our single- center experience demonstrate a lower than previously reported 68Ga-SSTR PET/CT sensitivity for ECS, especially in occult lesions. We suggest that the data on 68Ga-SSTR PET/CT in ECS is subject to publication bias, and false-negatives are likely underreported; it’s diagnostic value for ECS needs further study.