Granulomatous disease: is it a nuisance or an asset during PET/computed tomography evaluation of lung cancers?

To evaluate combined PET-computed tomography (CT) criteria for differentiating between granulomatous disease (GD) and malignancy (CA) in oncologic PET-CT studies.

Sixty-two patients who were referred for fluoro-2-deoxyglucose (FDG) PET-CT evaluation of pulmonary lesion(s) without a history of concurrent infection were studied. PET-CT was performed 1.5 h after intravenous administration of 555 MBq 18F-FDG in the fasting state with oral contrast. Combined PET-CT criteria including (i) calcifications (Ca2+) within lymph nodes, (ii) Ca2+ in lung nodules, (iii) liver and/or spleen Ca2+, (iv) locations of lung lesion(s), (v) hilar FDG uptake, (vi) comparison of lung versus maximum mediastinal FDG uptake, (vii) lymph node uptake not in the most probable lymphatic drainage pathway from a particular lung lesion, and (viii) extra pulmonary abnormal FDG uptake were each assigned a numerical score (0-3) with progressively higher score and sum of scores toward the increasing likelihood of GD. These patients either had pathological confirmation by biopsy/resection or were followed radiographically for a period of 2 years (CA=13; GD=49). Discriminant analysis was performed on all the above criteria with this gold standard. Simple t-test and box plot analysis were also performed on the summation of the scores (from 0 in CA to 13 in GD).

When all eight criteria were entered into discriminant analysis, the combined PET-CT criteria classified correctly 71% of patients with a sensitivity of 65% and specificity of 92% for GD. The most significant discriminating criterion was FDG uptake in the lung lesion(s) less than maximum mediastinal uptake (P=0.01). The sum scores in GD and CA were significantly different (4.9+/-2.4 vs. 3.2+/-1.5, respectively, P=0.014). Box plots showed a clear separation at a cut-off value of around 3.5.

Results show that the set of combined PET-CT criteria are highly specific for GD, which is not necessarily a nuisance during oncologic evaluation. Knowledge of these criteria may attribute some of the abnormal PET findings to GD, which is a useful asset for quick recognition and clinical interpretation.