Several studies have compared outcomes and complications between endoscopic and microscopic endonasal surgery in treating pituitary tumor. Endoscopic approach was reported to have technical superiority of panoramic visualization.16 Various studies reported several advantages, including superior volume of exposure, improved view of the anatomic structures, less mucosal trauma, shorter hospital stay, decreased blood transfusions, decreased patient discomfort, and improved visual outcomes.17–21
Our studies showed that surgical approach did not have significant relation with extent of resection. The result is not in accordance with previous studies that reported endoscopic transsphenoidal surgery significantly increased incidence of gross total removal compared to microscopic surgery.22–24 However, more studies showed that the endoscopic transsphenoidal approach did not significantly correlate with extent of resection,21,25–28 similar with our result, though a meta-analysis of 18 papers showed endoscopic approach was superior.29 Our research also showed that none of the patients who underwent endoscopic surgery had gross total removal, while five of twenty seven patients of microscopic surgery had total tumor resection. Besides higher grade of Knosp in endoscopic patients, this phenomenon could also be explained by the lack of experience in handling endoscopy instruments. Endoscopic transsphenoidal surgery at our institution was first performed in 2020, compared to years of experience of our neurosurgical team in microscopic transsphenoidal surgery. A study by Zaidi et al. showed that microscopic surgeries performed by more experienced neurosurgeons produced more gross total tumor resection than endoscopic surgery. However, the result was not statistically significant.28
Besides surgical technique, tumor size and extension should be considered when setting total gross removal as a goal of surgery. Tumors with higher Knosp grade and larger size significantly had lower rate of complete resection, according to a study by Karppinen et al.25 Those findings are consistent with the result of our research; tumor extension to cavernous sinus was inversely related to gross total resection, while suprasellar extension did not significantly correlate with the extent of resection. The findings are understandable given that the transsphenoidal approach could not reach the lateral side of the sella, and aggressive resection of tumor at cavernous sinus could lead to massive intraoperative bleeding and cranial nerve palsies. It should be noted that endoscopic patients in this study had higher grade of Knosp, but the resection rate was not significantly lower than the microscopic patients’. Suprasellar extension is not a barrier to total resection because it can be resected through an extended endoscopic transsphenoidal approach.30
The transsphenoidal approach has offered minimal invasiveness and comparable gross total removal rate than the transcranial approach; however, it has significantly higher rates of CSF leakage complication.31 Persistent CSF leaking is the major cause of morbidity following transsphenoidal surgery for pituitary tumor.32 Previous studies' comparison of leakage rates between endoscopic and microscopic transsphenoidal surgery showed different results. In one retrospective study, CSF leakage was significantly higher in endoscopic patients.33 However, many other papers concluded that surgical approach did not significantly affect the incidence of CSF leakage,21,22,24–28,34,35 including a meta-analysis of fifteen studies.13 Our seven months of experience in endoscopic transsphenoidal surgery showed that the post-operative leakage rate was not significantly different from the microscopic approach. It also appeared that tumor extension to suprasellar region or cavernous sinus did not affect the incidence of postoperative leakage. Although the CSF leakage rate was not different, endoscopic surgery is preferable to identify and repair leakage due to its enhanced illumination and visualization. One study reported a high CSF leakage repair success rate with precise confirmation and sufficient exposure of the leakage site using endoscopic transsphenoidal technique.36
The rate of DI after surgery was various in previous reports. Some studies reported no significant difference in DI rate between endoscopic and microscopic surgery,22,24–27,34 while reports by Zaidi et al. and Razak et al. favoured endoscopic over microscopic approach.28,35 These contrast with a study by Azad et al. that favoured a microscopic approach,33 consistent with our finding. Unfortunately, our research did not track whether the patients had temporary or permanent DI. With its enhanced visualization, the surgeons may perform more aggressive resections that could lead to stalk manipulation; this phenomenon may explain our result that showed higher incidence of DI in endoscopic patients. However, though the operators tended to be aggressive with an endoscope, the rate of postoperative CSF leakage was not different between the two approaches; this could be elaborated that the intraoperative leakage correction was easier and treated more precisely using endoscopy assistance. This led to successful closure, so clinical CSF leakage did not appear postoperatively.
Several studies reported that endoscopic approach was favourable for shorter hospital stays.13,34 A meta-analysis by Rotenberg et al. showed that endoscopic patients with fewer hospital stays had less overall surgical duration. The authors suggested the endoscopic surgery was less painful or had less complicated hospital course.37 However, another meta-analysis by Goudakos et al. reported that, although the result showed significantly shorter hospital stays for endoscopic patients, the difference of operative time and blood loss was not significant.38 We did not include hospital stay as an outcome variable because it is our hospital protocol for patients who underwent surgery more than 4 hours to receive a minimum five-day course of antibiotics before discharge.
In our study, the endoscopic approach's blood loss and operative length were higher and longer than the microscopic one, although not statistically significant. It appeared that the operator’s skill is an essential factor determining the result. A study by Guo-Dong et al. suggested that duration of surgery affected intraoperative bleeding. Some factors could cause the operative time of endoscopic surgery longer, such as preparation of both nostrils, manipulations that caused bleeding which obscured the lens and required saline irrigation, and inappropriate irrigation.34 Control of bleeding in the endoscopic approach is challenging, which could prolong surgical time. With years of experience in transsphenoidal surgery using a microscope, it takes time for our neurosurgeons to become accustomed to handling endoscopic instruments effectively. However, the surgical length in our centre gradually declined since the introduction of endoscope to the next seven months, as shown in Figure 2, reflecting our neurosurgical team learning curve.
This study reports our institutional experience in pituitary tumors treatment and evaluates our learning process in handling more sophisticated instruments. We have only used the endoscopy less than a year and already experienced the advantages, such as no need for fluoroscopy intraoperatively and the availability of performing pituitary surgery concomitantly with other microscopic surgery, as our center only have one operating microscope. Several surgical endoscopy workshops and training had been organized to expand our knowledge and sharpen our skills. We believe this report should be continued in the following years after the number of endoscopy surgery is large enough along with improvement of our skills. Our study also showed the outcomes and complications rate between microscopic and endoscopic approach were comparable.
There are some limitations to this study. First, the design of this report is retrospective. Second, we think the number of endoscopic patients in our institution is still small. Third, both microscopic and endoscopic surgeries were performed by different surgeons, so this could introduce outcome bias. Fourth, we did not include clinical outcomes as variables because too few endoscopic patients are available for long-term evaluation.