Accuracy of positron emission pornography for diagnosis of pulmonary nodules and mass lesions - A meta-analysis

Context Focal pulmonary lesions are commonly encountered in clinical practi ce, and positron emission tomography (PET) with the glucose analog 18-fluor odeoxyglucose (FDG) may be an accurate test for identifying malignant lesio ns. Objective To estimate the diagnostic accuracy of FDG-PET for malignant foca l pulmonary lesions. Data Sources Studies published between January 1966 and September 2000 in t he MEDLINE and CANCERLIT databases; reference lists of identified studies; abstracts from recent conference proceedings; and direct contact with inves tigators. Study Selection Studies that examined FDG-PET or FDG with a modified gamma camera in coincidence mode for diagnosis of focal pulmonary lesions; enroll ed at least 10 participants with pulmonary nodules or masses, including at least 10 participants with malignant lesions; and presented sufficient data to permit calculation of sensitivity and specificity were included in the anaylsis. Data Extraction Two reviewers independently assessed study quality and abst racted data regarding prevalence of malignancy and sensitivity and specific ity of the imaging test. Disagreements were resolved by discussion. Data Synthesis We used a meta-analytic method to construct summary receiver operating characteristic curves. Forty studies met inclusion criteria. Stu dy methodological quality was fair. Sample sizes were small and blinding wa s often incomplete. For 1474 focal pulmonary lesions of any size, the maxim um joint sensitivity and specificity (the upper left point on the receiver operating characteristic curve at which sensitivity and specificity are equ al) of FDG-PET was 91.2% (95% confidence interval, 89.1%-92.9%). In current practice, FDG-PET operates at a point on the summary receiver operating ch aracteristic cu rye that corresponds approximately to a sensitivity and spe cificity of 96.8% and 77.8%, respectively. There was no difference in diagn ostic accuracy for pulmonary nodules compared with lesions of any size (P=. 43), for semiquantitative methods of image interpretation compared with qua litative methods (P=.52), or for FDG-PET compared with FDG imaging with a m odified gamma camera in coincidence mode (P=.19). Conclusions Positron emission tomography with 18-fluorodeoxyglucose is an a ccurate noninvasive imaging test for diagnosis of pulmonary nodules and lar ger mass lesions, although few data exist for nodules smaller than 1 cm in diameter. In current practice, FDG-PET has high sensitivity and intermediat e specificity for malignancy.