MRI and Trouillas' grading system of pituitary tumors: the usefulness of T2 signal intensity volumetric values.
Neuroradiology
; 65(11): 1567-1578, 2023 Nov.
Article
en En
| MEDLINE
| ID: mdl-37233819
ABSTRACT
PURPOSE:
To classify pituitary macroadenomas according to the Trouillas' grading system; to compare this grading system with T2 values of volumetric signal intensity to determine T2 values able to predict the final grade.METHODS:
A total of 106 patients with macroadenomas were grouped according to the grading system score combining proliferation and invasiveness criteria of Trouillas' classification. Normalized volumetric signal intensity values were extracted from coronal T2-weighted images (nT2mean, nT2Max, nT2min) and were compared with the final grading score system.RESULTS:
Thirty-three patients were in grade 1a (non-invasive, non-proliferative tumors), 17 patients in grade 1b (non-invasive, proliferative tumors), 36 patients in grade 2a (invasive, non-proliferative tumors), and 20 patients in grade 2b (invasive, proliferative tumors). No patient was in grade 3 (metastatic tumors). nT2Max and nT2min were the best quantitative values to discriminate invasive from non-invasive grades; in invasive grades, nT2Max intensity values were higher, and nT2min intensity values were lower than in non-invasive grades. Receiver operating characteristic analysis of nT2 values showed that nT2min values had a better diagnostic performance than nT2Max values because they allowed differentiating with a moderate accuracy invasive tumors (2a or 2b grades) from both non-invasive proliferative tumors (1b) and non-invasive-non proliferative tumors (1a) (2a vs 1b AUCnT2min = 0.78, 2b vs 1b AUCnT2min = 0.72, 2a vs 1a AUCnT2min = 0.72, 2b vs 1a AUCnT2min = 0.69).CONCLUSION:
Volumetric nT2Max and nT2min values of MRI might be practical and non-invasive markers for assessing tumor invasiveness although nT2 min signal intensity values have more effects in discriminating tumor's invasive behavior.Palabras clave
Texto completo:
1
Colección:
01-internacional
Banco de datos:
MEDLINE
Asunto principal:
Neoplasias Hipofisarias
Tipo de estudio:
Prognostic_studies
Límite:
Humans
Idioma:
En
Revista:
Neuroradiology
Año:
2023
Tipo del documento:
Article
País de afiliación:
Italia