Evaluating Sellar Healing following Various Reconstruction Techniques in Transsphenoidal Surgery: An Imaging and Endoscopic Review

Background: Many institutions utilize a graded algorithm to dictate sellar reconstruction following transsphenoidal pituitary surgery based on the defect encountered. At our institution, one of three repair techniques are commonly used for pituitary surgery: a regenerated oxidized cellulose onlay, a synthetic dural substitute inlay with dural sealant glue, or synthetic dural substitute inlay with nasoseptal flap (NSF) onlay and dural sealant glue. The objective of this study is to assess postoperative healing of the sphenoid and sella following reconstruction with these three techniques based on magnetic resonance imaging (MRI) and endoscopic features, and to compare these modalities.

Methods: Twenty consecutive patients who underwent sellar reconstruction with each of the above-mentioned techniques were included (regenerated oxidized cellulose onlay = group ROC, synthetic dural substitute inlay/dural sealant glue = group DS, and synthetic dural substitute inlay/NSF onlay = group F). Three-month postoperative MRIs were reviewed and compared with preoperative MRIs and 3-month postoperative nasal endoscopy when available. Each MRI was reviewed by a neuroradiologist and graded on a binary scale based on the presence of imaging features of sinusitis and mucosal health including an air-fluid level, the uniformity of mucosal thickness, the presence of enhancing soft tissue, and the presence of layering of the sella. Likewise, each endoscopy video was reviewed by two otolaryngologists and evaluated for the presence of crusting, sinusitis, and degree of remucosalization. Postoperative morbidity was evaluated. Data were analyzed using Chi-squared and ANOVA; a result was considered significant when p < 0.05. Imaging and endoscopic assessments were compared.

Results: There were three exclusions from group F due to incomplete imaging records. MRI evaluation showed that all patients appeared to have complete sphenoid remucosalization, including the sella. Sphenoid mucosal thickening, asymmetric soft-tissue thickening, and sphenoid air-fluid levels were all most frequently identified in group F; however, these features were not statistically significant (70% in group F vs. 45 and 45% in groups ROC and DS, p = 0.25; 100 vs. 5% and 5%, p < 0.001; 12 vs. 10% and 5%; p = 0.34). Soft-tissue enhancement was identified in one patient in each group (p = 0.99). Layering of the sella was seen frequently in groups DS and F (100 and 71%, respectively) but was not observed in group ROC (p < 0.001). On nasal endoscopy, the sella and sphenoid were predominantly remucosalized in 100% of cases in all groups. Group F patients were more likely to have crusting, although this was not statistically significant (50 vs. 20% and 14%, p = 0.22). Group F patients were more likely to require postoperative antibiotics for a clinical sinusitis (OR = 6.0). There were no complications of cerebrospinal fluid leak in any group. There was no correlation between positive endoscopic findings and positive imaging features (χ ² = 2.32, p = 0.99).

Conclusion: Imaging features of sinusitis and asymmetric sphenoid mucosal healing are more frequently identified on postoperative MRI than on nasal endoscopy. All techniques heal equally well both radiographically and via direct visualization, although the use of a NSF increases morbidity. More conservative reconstruction methods when safe may reduce the postoperative morbidity while not sacrificing remucosalization and sinonasal health.

Die Autoren geben an, dass kein Interessenkonflikt besteht.