In:
Journal of Thoracic Imaging, Ovid Technologies (Wolters Kluwer Health), Vol. 38, No. 5 ( 2023-09), p. 278-285
Abstract:
Pulmonary hamartomas (HAs) and neuroendocrine neoplasms (NENs) are often impossible to discriminate using high-resolution computed tomography (CT) as they share morphologic features. This challenge makes differential diagnosis crucial as HAs are invariably benign, whereas NENs must be considered malignant, thus requiring them to be evaluated for surgical excision. Our aim was, therefore, to develop a simple method to discriminate between pulmonary “fat-poor” HAs and NENs using contrast-enhanced CT (CECT). Materials and Methods: Between September 2015 and December 2021, 95 patients with a histologically proven diagnosis of lung NENs (74) and HAs (21) and who underwent a preoperative CECT scan were initially identified through a review of our pathologic and radiologic databases. Among these, 55 cases (18 HAs and 37 NENs), which have been studied with biphasic CECT, were ultimately selected and reviewed by 3 radiologists with different levels of experience, analyzing their morphologic and enhancement features. The enhancement analysis was performed by placing a region of interest within the lesion in noncontrast (NCp), postcontrast (PCp, 55 to 65 s after intravenous contrast injection), and delayed phases (Dp, 180 to 300 s). A subgroup of 35 patients who underwent 18FDG-PET/CT was evaluated in a secondary analysis. Results: HU values were significantly different between NENs and HAs in the PCp ( P 〈 0.001). NCp and Dp attenuation values did not show significant differences in the 2 groups. Differences in values of HUs in PCp and Dp allowed to discriminate between NENs and HAs. Conclusion: Wash-out analysis, ΔHU (PCp-Dp), can perfectly discriminate pulmonary “fat-poor” HAs from NENs.
Type of Medium:
Online Resource
ISSN:
0883-5993
DOI:
10.1097/RTI.0000000000000712
Language:
English
Publisher:
Ovid Technologies (Wolters Kluwer Health)
Publication Date:
2023
detail.hit.zdb_id:
2048799-X
Bookmarklink
