OBJECTIVE: Fluorodeoxyglucose (FDG)-positron emission tomography
(PET) is an accurate non-invasive imaging test for differentiating
benign from malignant solitary pulmonary nodules (SPNs). We
aimed to assess its diagnostic accuracy for differentiating benign from
malignant SPNs in a tuberculosis (TB)-endemic area.
METHODS: Thirty patients, 22 men and 8 women, mean age 60
years, underwent dual time point FDG-PET/computed tomography
(CT) imaging, followed by histological examination of the SPN.
Maximum standard uptake values (SUVmax) with the greatest
uptake in the lesion were calculated for two time points (SUV1
and SUV2), and the percentage change over time per lesion was
calculated (%DSUV). Routine histological findings served as the
RESULTS: Histological examination showed that 14 lesions were
malignant and 16 benign, 12 of which were TB. SUVmax for benign
and malignant lesions were 11.02 (standard deviation (SD) 6.6) v.
10.86 (SD 8.9); however, when tuberculomas were excluded from
the analysis, a significant difference in mean SUV1max values
between benign and malignant lesions was observed (p=0.0059).
Using an SUVmax cut-off value of 2.5, a sensitivity of 85.7% and
a specificity of 25% was obtained. Omitting the TB patients from
analysis resulted in a sensitivity of 85.7% and a specificity of 100%.
Mean %DSUV of benign lesions did not differ significantly from
mean %DSUV of malignant lesions (17.1% (SD 16.3%) v. 19.4%
(SD 23.7%)). Using a cut-off of %DSUV >10% as indicative of
malignancy, a sensitivity of 85.7% and a specificity of 50% was
obtained. Omitting the TB patients from the analysis yielded a
sensitivity of 85.7% and a specificity of 75%.
CONCLUSION: Our findings suggest that FDG-PET cannot
distinguish malignancy from tuberculoma and therefore cannot
reliably be used to reduce futile biopsy/thoracotomy.