Diagnostic accuracy of 18F-fluorodeoxyglucose positron emission tomography in the follow-up of papillary or follicular thyroid cancer

Positron emission tomography with 18F-fluorodeoxyglucose is a relatively new nuclear imaging technique in oncology. We conducted a systematic review to determine the diagnostic accuracy of 18F-fluorodeoxyglucose positron emission tomography in patients suspected of recurrent papillary or follicular thyroid carcinoma. Two reviewers independently selected, extracted, and assessed data from relevant literature found in computerized databases and by reference tracking. Prospective and retrospective studies with 10 human subjects, or more, that evaluated the accuracy of ring positron emission tomography, using 18F-fluorodeoxyglucose in follicular and papillary thyroid cancer, were included. Studies on 18F-fluorodeoxyglucose imaging using γ cameras, reviews, case reports, editorials, letters, and comments were excluded. The methodological quality was assessed by applying the criteria for diagnostic tests recommended by the Cochrane Methods Group on Screening and Diagnostic Tests. A rating system was used for qualitative analysis consisting of four levels of evidence (1 = highest level; 4 = lowest level). Fourteen studies met the inclusion criteria. All studies claimed a positive role for positron emission tomography but, at evidence levels 3 or 4, precluding quantitative analysis. Methodological problems included poor validity of reference tests and a lack of blinding of test performance and interpretation. The reviewed material was heterogeneous with respect to patient variation and validation methodology. The most consistent data were found on the ability of 18F-fluorodeoxyglucose positron emission tomography to provide an anatomical substrate in patients with elevated serum Tg and negative iodine-131 scans. In conclusion, the results seem to support the potential of 18F-fluorodeoxyglucose positron emission tomography to identify and localize foci of recurrent cancer in the latter patient subset. However, implementation of positron emission tomography in a routine diagnostic algorithm requires additional evidence.