F18-FDG PET-CT analyses of small peripheral adenocarcinoma of the lung

Abstract

Background

Radiological discrimination of histologic subtypes of small peripheral adenocarcinoma of the lung is clinically important. Although there are many articles in which CT findings were used for this purpose, there are only a few reports on the capability of FDG PET-CT findings for histologic classification of this tumor.

Purpose

To investigate the correlation between visual assessment or maximum standard uptake values (SUV_max) on F18-FDG PET-CT and histology grading of small peripheral adenocarcinoma of the lung.

Material and Methods

Proportions of positive PET-CT diagnoses and SUV_max were retrospectively reviewed on 96 solitary pulmonary nodules of ≤2 cm in 90 consecutive patients. Tumors were classified into four groups according to Noguchi's classification (group 1 [n = 10], atypical adenomatous hyperplasia and type A tumors; group 2 [n = 12], type B tumors; group 3 [n = 42], type C tumors; group 4 [n = 32], types D, E, and F tumors). Proportions of positive PET-CT diagnoses and mean SUV_max of lesions among four groups were compared using trend tests to examine if there is a significant linear correlation with the progression of the histology grading of tumors. Then, an optimal threshold of SUV_max was proposed to best discriminate tumors of poor (groups 3 and 4) from good (groups 1 and 2) prognosis.

Results

There was a significant linear trend for both visual assessment (P < 0.01) and SUV^max (P < 0.01). A SUV_max of 0.42 showed the highest accuracy of 84% with 95% sensitivity and 50% specificity for predicting tumors of poor prognosis. A SUV_max of 2.05 showed 100% specificity with 49% sensitivity, and 60% accuracy. Positive visual diagnoses showed accuracy of 83% with 90% sensitivity and 59% specificity.

Conclusion

Visual assessment and SUV_max on PET-CT correlated well with the histology grading of small peripheral adenocarcinoma of the lung.