Context: (123)I-mIBG scintigraphy has been in clinical use for more than 20 yr for diagnostic assessment of patients with neural crest and neuroendocrine tumors. Prospective validation of the performance characteristics of this method has recently been published.
Objective: A meta-analysis was performed to obtain best estimates of performance characteristics of (123)I-mIBG imaging for the two most common applications, evaluation of patients with neuroblastoma and pheochromocytoma.
Data sources: Articles published between 1980 and 2007 were identified from searches of multiple computer databases, including MEDLINE, BIOSIS, EMBASE, and SciSearch.
Study selection: Primary inclusion criteria were: acceptable reference standard(s) for confirming subjects with disease (histopathology and/or a combination of imaging and catecholamine results); reference standards applied to all subjects who received (123)I-mIBG; and data on a minimum of 16 patients confirmed to have or not have the disease(s) under consideration. Two physician reviewers independently evaluated all articles against the inclusion/exclusion criteria. Twenty-two of 100 articles reviewed were included in the final analysis.
Data extraction: The two reviewers extracted the data from eligible articles using a standardized form, capturing both study quality and efficacy information. Disagreements were resolved by consensus.
Data synthesis: Sensitivity of (123)I-mIBG scans for detection of neuroblastoma was 97% [95% confidence interval (CI), 95 to 99%]; data were insufficient to estimate specificity. For pheochromocytoma, with application of the random-effects model, sensitivity and specificity were 94% (95% CI, 91-97%) and 92% (95% CI, 87-98%), respectively.
Conclusion: Based upon the literature, (123)I-mIBG scintigraphy has sensitivity and specificity greater than 90% for detection of neuroblastoma and pheochromocytoma.