Intro: Critical to the presurgical planning process for the resection of suspected pituitary macroadenomas is defining tumor involvement with local anatomy including invasion of the cavernous sinus, compression of internal carotid arteries, and displacement of the optic chiasm. Magnetic resonance (MR) serves as the principle imaging modality for this purpose. Currently, standard pituitary gland MRI protocols include combinations of T1, T2, T1 C+ dynamic, T1 C+ delayed, DWI/ADC, and FLAIR studies. Absent from this protocol, however, is one of the established go-to sequences for defining cranial nerves and cavernous sinuses: constructive interference in steady state (CISS). Equivalent sequences on other scanners, such as GE, include FIESTSA-C (Fast Imaging Employing Steady-state Acquisition Cycled Phases). Herein, we describe the superior image quality with equivalent protocol time achieved with the introduction of CISS as the preferred MR sequence in radiographic studies for pituitary macroadenoma presurgical planning.
Methods: In 2016, our skull base program transitioned imaging preferences from pituitary protocol MRI to CISS protocol for pituitary macroadenomas. This included standard pre- and post-contrast imaging and post contrast CISS sequences. Those used for presurgical planning were examined for protocol type. The pituitary protocol was contrasted against CISS based on protocol length and imaging quality with respect to defining tumor, local anatomy, and tumor involvement with the local anatomy (specifically optic chiasm, cavernous sinuses, and carotid artery encasement).
Results: Between 2016 and 2018, 10 of 104 patients who received pre-surgical MRI for pituitary macroadenoma were evaluated with CISS protocol instead of pituitary protocol. In all CISS protocol patients, the carotid arteries, optic chiasm, and cavernous sinuses were clearly demarcated in 3 planes for surgical planning. Normal pituitary gland and infundibulum were visualized well in nearly all patients.
Patients imaged with pituitary protocol with timed contrast in the coronal plane had well visualized carotid arteries, but with less clear understanding of the tumor relationship with optic chiasm and cavernous invasion in multiple planes. Further, the timed contrast did not reveal information about the tumor that impacted surgical planning for macroadenomas.
CISS protocol image acquisition time was 36 minutes, compared to the pituitary protocol of 30 minutes.
Discussion: We describe the imaging superiority of 3d CISS imaging in surgical planning in these patients with little addition of image acquisition time. Pituitary protocol of timed contrast in the coronal plane did not reveal helpful information in surgical planning for macroadenomas. The low number of patients receiving this protocol as well as historical imaging practices reveals a reflexive protocol for pituitary patients that is difficult to overcome.
Given the documented superiority of CISS to other sequences in defining neurovascular compression as well as defining cerebrospinal fluid-filled cavernous spaces, superiority to standard pituitary protocol for presurgical planning was not unexpected. Future efforts will focus on analyzing and quantifying surgical outcomes of pituitary macroadenoma resections that involved CISS in presurgical planning vs. those that did not.