Validation of FDG positron emission tomography for differentiation of unknown pulmonary lesions.

OBJECTIVE

The impact of the (2-(fluorine-18)-fluoro-2-2deoxy-D-glucose)-positron emission tomography ((18)F-FDG-PET) for discrimination of pulmonary lesions was evaluated in a single centre prospective study.

METHODS

In the study, 109 patients with pulmonary lesions of unknown origin verified by computed tomography were enrolled consecutively (April 1999--May 2000). They were subject to (18)F-FDG-PET diagnostics. (18)F-FDG-PET images were interpreted by two independent nuclear medicine physicians who were blinded to the results of other imaging procedures. In 87 patients, surgery was applied followed by histological investigation, which served as the gold standard. In 22 other patients, extensive tumour load or assumed benign dignity of the lesions prevented surgery.

RESULTS

Overall sensitivity of (18)F-FDG-PET in 87 resected patients was 0.86. Differentiation in malignant (n = 69) and benign lesions (n = 18) revealed sensitivities of 0.9 and 0.72, respectively. Sensitivity of (18)F-FDG-PET in inflammatory lesions was markedly lower (0.43) than in benign tumours (0.91). Standard uptake values were significantly increased in malignant tumours compared with benign lesions (9.9 and 1.6, respectively; P = 0.035). There was a clear correlation of sensitivity with tumour size with a failure rate of 27% in lesions < or = 1cm (n = 15), 10% (n = 20) in lesions between 1 and 2 cm and 12% (n = 45) above 2 cm. In primary bronchial carcinoma, a clear correlation of sensitivity was observed with regard to tumour grading (G1, three out of five; G2, 24 out of 27; G3, 26 out of 26; and G4, one out of one). Lymph node involvement was correctly suggested in 10 out of 19 (52.6%) patients. However, false positive lymph node enhancement was indicated in one out of 18 (5.5%) operated patients with benign lesions and eight out of 39 (20.5%) with bronchial carcinoma.

CONCLUSION

(18)F-FDG-PET at present does not serve as the gold standard for early detection of small and well-differentiated tumours. However, it contributes efficiently to the detection of malignancy in tumours >1cm, which are moderately or poorly differentiated. Positive lymph node imaging must not preclude surgery but requires histological proof. Discrimination of benign and malignant pulmonary tumours by (18)F-FDG-PET appears to be hampered in inflammatory lesions.