The Value of Postoperative 3T MR Imaging Following Endoscopic Endonasal Sellar/Suprasellar Surgery

Introduction: There is no consensus regarding optimal timing of postoperative MRI for sellar/suprasellar lesions, and most surgeons wait 2-3 months to allow regression of blood and packing material. Currently, there is no literature evaluating immediate postoperative 3T MR imaging for these lesions. We hypothesized that 3-tesla MRI of the sellar/pituitary gland performed on the first postoperative day (POD #1) is just as effective at detecting residual disease and/or reconstruction materials as the MRI performed at 3 months following endoscopic endonasal resection of sellar/suprasellar lesions.

Methods: Between February 2009 and October 2010, we evaluated 102 consecutive patients who underwent endoscopic endonasal surgery for sellar/suprasellar lesions. Expanded endoscopic procedures were excluded from this series. Sixty-four patients met the inclusion criteria with POD #1 and at 3 months MR imaging following surgery. The MRIs were evaluated by two sets of observers. Observer agreement was analyzed using kappa index. The following parameters were assessed: (1) enhancement pattern of the pituitary gland, (2) pituitary stalk and its deviation, (3) nodular enhancement (residual tumor) or linear enhancement (non-tumoral), and (4) residual reconstruction/packing materials. Any differences in interpretation between POD #1 and 3-month MRIs were identified.

Results: Fifty-nine of 64 patients harbored pituitary adenomas; 4 had Rathke’s cleft cysts, and 1 had a schwannoma. Fifty-three of the 59 adenomas were macroadenomas with 12 demonstrating cavernous sinus invasion. Gross total resection of the tumors residing within the sella and/or suprasellar space was achieved in 48 out of 52 (92%) patients. The pituitary gland, position of stalk, and nasoseptal flap could be identified on both POD #1 and 3-month MR imaging in all patients. Five patients showed nodular enhancement on POD #1 MR imaging, suggestive of residual tumor, confirmed by MRI at 3 months in 4 of 5 patients. The fifth patient’s 3-month MRI did not demonstrate nodular enhancement. Similarly, two patients had no nodular enhancement on POD #1 MRI but clear nodular enhancement on 3-month MRI. The sensitivity and specificity for residual tumor detection by MRI on POD #1 were 97% and 98%, respectively. Linear enhancement seen in 12 patients on POD #1 MRI disappeared in 6 at 3 months and remained stable in remaining 6. Seven patients’ 3-month MRI demonstrated a change in shape and/or stalk position when compared with POD #1 MRI. The packing materials were distinguishable from the pituitary gland and the stalk in 54 patients on POD #1. The reconstruction/packing material showed complete resolution in 22 patients, partial resolution in 33, and no resolution in 9. The kappa index was 0.81, suggesting high interobserver agreement.

Conclusion: MR imaging performed POD #1 following endoscopic endonasal resection of sellar/suprasellar lesions provides accurate and reliable information regarding the presence of residual tumor compared with reconstruction and packing materials. This information will allow treating physicians to advance the patient’s plan of care more expeditiously, while leaving the option of performing MR imaging at 3 months if the results of POD #1 imaging are unclear. Immediate postoperative MRI should become a standard of care after endoscopic resection of all pituitary adenomas and other sellar lesions.