Background: The aim of this study was to compare three different techniques for transsphenoidal pituitary surgery: (1) sublabial transseptal approach with microscopic resection, (2) transnasal transseptal approach with endoscopic resection, and (3) endoscopic approach with endoscopic resection.
Methods: We performed a retrospective review of 50 pituitary surgeries performed by the same neurosurgeon. Demographic, radiographic, and clinical data were collected.
Results: Fifteen patients underwent sublabial approach with microscopic tumor resection, 21 patients underwent the transnasal approach with endoscopic resection, and 14 patients underwent the completely endoscopic technique. There were a total of 20 complications in the sublabial group, 13 transnasal complications, and 6 endoscopic complications. Cerebrospinal fluid leak incidence was 53% in the sublabial approaches, 47% transnasal, and 28% in the endoscopic patients. Diabetes insipidus was encountered in 33% of sublabial approaches, 5% of transnasal approaches, and 7% of endoscopic approaches. Lumbar drains were required in 40% of sublabial approaches, 38% of transnasal approaches, and 7% of endoscopic approaches. Nasal packing was used in 100% of sublabial and transnasal approaches and 0% of endoscopic approaches. Mean recurrence rate and follow-up was sublabial in 6.6% (50 months), transnasal in 9.5% (11 months), and endoscopic in 0% (7 months). Average hospital stay for sublabial approaches, transnasal approaches, and endoscopic approaches was 8.3, 6.2, and 3.4 days, respectively (p < 0.05).
Conclusion: Transsphenoidal pituitary surgery has evolved over the past several decades, because advances in technology have been the catalyst for minimally invasive surgeries. Less invasive approaches, such as the transnasal approach with endoscopic resection of tumor and the completely endoscopic .technique have less morbidity and a shorter hospital stay than traditional sublabial approaches. Continued follow-up is needed to confirm long-term benefits and similar recurrence rates.