Qualitative and semiquantitative parameters in solitary pulmonary nodule: a multicenter 18F-FDG-PET/CT study

Background-Aim: The aim of this study is to evaluate the reliability of 18F-FDG-PET/CT qualitative and semiquantitative parameters, analysed either individually or as a whole, in patients with solitary pulmonary nodule (SPN), in order to predict the risk of malignancy. Methods: A retrospective evaluation of 18F-FDG-PET/CT of 146 patients (49 females, 97 males with a mean age of 68.23 (± 8.32) years) positive for SNP was performed, according to their pre-test probability of malignancy determined by the presence of risk factors such as smoking, occupational exposure, previous cancer and pulmonary disease (COPD, pulmonary fibrosis, emphysema). PET/CT examinations were acquired according to international acquisition procedure guidelines using a Discovery 710 PET/CT system (GE Healthcare). Cytopathological examination of samples, considered to be the gold standard for lung cancer diagnosis, was performed in order to assess malignancy of the lesions. Qualitative parameters with 3 points scoring system (high, medium or absent uptake), dimension and site of SPN, as well as semiquantitative parameters (TLG, MTV, SUVmax, SUVmean) were taken in account. Results: Among the 146 patient who underwent FDG-PET/CT, SPN was found to be positive for malignancy in 89 cases (60.95%) at cytopathological examination. Uptake was absent in 20 cases, moderate in 44 and high in 82. The mean dimension at CT of SPN was 18.28 (± 8.32) mm, 58 were localized in the inferior lobe, 74 in the superior lobe and 14 in the medium lobe. No statistical significance was found between the tracer uptake, the dimension or the site of the SPN and the risk of malignancy. The pre-test probability of malignancy doesn’t correlate significantly with any semiquantitative parameters taken in account, but a direct correlation was observed between the diameter of the SPN at CT and the pre-test probability of malignancy. Average values for semiquantitative parameters were found: SUVmax 6.53 (± 4.63), SUVmean 3.508 (± 2.234), TLG 19.88 (± 33.45), MTV 3.29 (± 2.99). Optimal cut-offs for semiquantitative parameters, in terms of predictive values, have been found: a cut-off of 3.625 of SUVmax has 86.6% sensitivity and 69.1% specificity; SUVmean’s optimal cut-off is 2.51, with 79.1% sensitivity and 76.3% specificity; an MTV cut-off of 2.55 shows 74.7% sensitivity and 70.9% specificity while a TLG’s cut-off of 11.8 shows 66% sensitivity and 83.6% specificity. Furthermore, it has been observed how TLG values directly correlates with SUVmax, SUVmean and MTV: indeed, by means of Spearman correlations, it was seen that an increase in TLG consequently overlaps with an increase in SUVmax, SUVmean and MTV. Conclusions: The qualitative parameters, size and location of SPN do not show statistical significance for predicting the risk of malignancy. No correlations have been found between the pre-test probability of malignancy and the semiquantitative parameters. Instead semiquantitative parameters can have an important predictive role in patients with SPN at 18F-FDG-PET/CT. Cut-offs for SUVmax (3.625), SUVmean (2.51), MTV (2.55) and TLG (11.8) show good sensitivity and specificity. TLG has shown to be directly correlated to SUVmax and SUVmean and could play an important predictive role in patients with SPN at FDG-PET/CT as a predictive parameter for malignancy.