ESPE Abstracts (2018) 89 RFC7.6

Assessment of Pituitary Stalk Anatomy by T2 DRIVE without Gadolinium in Pituitary Diseases

Flavia Napolia, Elisabetta Godanoa, Giovanni Moranab, Natascia Di Iorgia, Angela Pistorioc, Anna Elsa Maria Allegria, Roberto Gastaldia, Annalisa Calcagnoa, Giuseppa Pattia, Annalisa Galliziaa, Sara Notarnicolaa, Marta Giaccardia, Serena Nolia, Mariasavina Severinob, Domenico Tortorab, Andrea Rossib & Mohamad Maghniea


aDepartment of Pediatrics, Istituto Giannina Gaslini, University of Genoa, Genova, Italy; bPediatric Neuroradiology Unit, Istituto Giannina Gaslini, Genova, Italy; cEpidemiology and Biostatistics Unit, Istituto Giannina Gaslini, Genova, Italy


Objectives: To evaluate the potential diagnostic role and sensitivity of T2-weighted DRIVE sequence in pituitary stalk (PS) identification and measurements in patients with hypothalamic-pituitary disorders. The degree of agreement and reliability between standard pre- and post-contrast T1-weighted images and T2-DRIVE will be tested in a large group of patients with pituitary dysfunction.

Design: We searched for pituitary MRI reports using ‘T2-DRIVE’ in our Institutional database between 2006 and 2015. Among 135 eligible patients, 102 showed eutopic posterior pituitary(PP)gland and 33 showed ‘ectopic’PP(EPP).

Methods: In patients with eutopic PP, two readers measured the PS size in the sagittal plane, drawing a line perpendicular to the axis of the major stem at three levels: proximal, midpoint, and distal on pre- and post-contrast T1-weighted and on T2-DRIVE images. The pituitary stalk was assessed on pre-contrast T1 and T2-DRIVE sequences in those with EPP. Cohen’s kappa coefficient was then used to evaluate the chance-correct concordance for the case between two different sequences that are expressed in the form of categorical data.

Results: The agreement between the measurements of the two readers showed that the ICC in the T2-DRIVE sequence was 0.96 at the proximal level of the PS, 0.99 at the midpoint level and 0.97 at the distal level. In pre-contrast T1-weighted sequence, the ICC was 0.89 (proximal part), 0.85 (midpoint), and 0.76 (distal part). Finally, on the post-contrast T1, the ICC was 0.88 at the proximal, 0.87 at the midpoint and 0.79 at the distal PS levels. A significant difference between the ICC on the T2-DRIVE and the pre-and post-contrast T1-weighted sequences was demonstrated. The percentage of PS identified by T2-DRIVE in EPP patients was 72.7% compared to 30.3% of T1 pre-contrast sequences. A significant association was found between the visibility of PS on T2-DRIVE and the height of AP.

Conclusion: T2-DRIVE sequence is precise and reliable for the evaluation of PS size and the recognition of PS abnormalities; the use of gadolinium does not add significant information. T2-DRIVE images allow for a better diagnosis of pituitary gland and PS disorders. A sagittal T2-DRIVE sequence without gadolinium takes less than 3 minutes to acquire, and its inclusion into routine sellar MRI protocols is recommended as a valid alternative to post-contrast imaging which - also in view of safety issues - may be avoided in subjects with pituitary disorders without evident sellar/suprasellar mass lesions.