Background Pituitary adenomas often invade the medial wall of the cavernous sinus (CS) without
extending into the cavernous sinus compartments. Surgical removal of the medial wall
of the CS, while considered important to increase remission rates and decrease tumor
recurrence, is generally not recommended secondary to the risk of vascular and cranial
nerve injury and massive blood loss.
Objective The purpose of this study is to present a step-by-step surgical technique designed
to completely remove the medial wall of the CS, and to report the surgical outcomes
in a large series of invasive pituitary adenomas.
Methods A retrospective review was conducted to identify patients who underwent an endoscopic
endonasal approach (EEA) for pituitary adenoma with selective resection of the medial
wall of the CS since its implementation in 2012. Cases with tumor invasion into cavernous
sinus compartments were excluded. Patient complications, resection, and remission
rates (using the most up-to-date criteria) were assessed.
Results The key steps of the proposed surgical technique are: (1) wide exposure of the anterior
wall of the CS and skeletonization of the internal carotid artery (ICA); (2) removal
of all sellar tumor and inspection of the medial wall; (3) opening of the anterior
wall to directly access the CS; (4) gentle mobilization of the medial wall starting
at the sellar floor; (5) identification of the cavernous ICA and inferior CS ligament;
(6) coagulation and transection of the inferior hypophyseal artery; (7) further mobilization
of the medial wall in a posterior direction to reach the posterior clinoid, and in
a superior direction up to the carotico-clinoidal ligament; and (8) medial, posterior,
and superior dural cuts.
Forty-nine patients were eligible for this study, 15 (31%) with nonfunctional adenomas
and 34 (69%) with functional adenomas, including 16 GH, 10 PRL, and 8 ACTH secreting
adenomas. Five cases (10%) corresponded to reoperations. The average size was 2.29
cm for nonfunctional and 1.29 cm for functional adenomas. Radiographically, 11 patients
(22%) were Knosp Grade I, 22 (45%) patients were Knosp Grade II, and 16 (33%) patients
were Knosp Grade III. Complete tumor resection, based on intraoperative impression
and postoperative MRI, was achieved in all cases. Mean follow-up was 13.3 months (1.8−29.8)
for nonfunctional and 25.8 months (1.7−64.1) for functional adenomas. Complete remission
in functional adenomas was seen in 33 patients (97%). No imaging recurrences were
seen during this period in nonfunctional adenomas. There were neither deaths nor ICA
injuries, and the average blood loss was 378 mL. Four patients (8%) developed a new
CN palsy that resolved completely at 3 months of follow-up in all of them, with two
requiring early postoperative clot or fat graft evacuation.
Conclusion The surgical technique described here is based on an accurate understanding of the
dural layers, CS ligaments, venous channels, and cavernous ICA trajectory and branches.
The medial wall of the CS can be removed safely and effectively, with minimal morbidity
and excellent resection and remission rates. Further follow-up is needed to determine
the long-term results of this technique.
