Next, Andrew Little will present Prospective Multi-Center Study Comparing Surgical Outcomes of Microscopic Transphenoidal Surgery and Fully Endoscopic Transphenoidal Surgery Techniques for Non-Functioning Pituitary Adenomas, Results of the Transfer Study. Thank you. Good afternoon, and thank you to the AANS and the Integra Foundation for the opportunity to speak today. I'd like to first acknowledge my co-authors. This study included 15 surgeons across seven centers. I'd also like to acknowledge the biostatistical contribution of Dr. Christina Chappell. These are the disclosures. As is well known to this audience, transphenoidal surgery is the preferred treatment for most patients with symptomatic non-functioning pituitary adenomas because of the excellent tumor control and good clinical outcomes. The two most common types of transphenoidal surgery performed now are the microscopic approach, utilizing a nasal speculum to maintain the corridor, and the operating microscope for visualization. In the endoscopic approach, an endoscope is used for visualization, and typically a three- or four-handed technique. Both approaches have advantages and disadvantages, primarily related to surgical freedom and visualization. This slide here depicts the percent of patients undergoing transphenoidal surgery by either technique for the past eight years. You'll see that in 2013 was an important inflection point, at which point more patients began to undergo surgery by the endoscopic approach compared to the microscopic approach. However, this trend hasn't been supported by high-quality clinical evidence that one technique is either equivalent or superior to the other. This slide is a snapshot of some data, some published series, showing comparative results. However, they're limited by their retrospective nature. The data is class three, in some cases expert opinion. In many cases, the cohorts are not contemporaneous, and it's the surgeon reviewing his or her own data and the associated biases with that. So the hypothesis of this study was the following. Because we think that endoscopic approaches offer better visualization, that it might permit improved gross total resection rates and overall better patient outcomes, and this is the hypothesis we attempted to address in this study. The study is a prospective study. It's multicenter. It's non-randomized. We included patients who are adults with non-functioning adenomas who are able to obtain contrasted MRI scans. We excluded the typical exclusions, such as rasty cleft cysts and pituitary apoplexy. The primary endpoint of this study was dichotomized gross total resection versus subtotal resection on MRI scans performed within three months of surgery by blinded reviewers, centrally adjudicated. We also examined a number of secondary endpoints. Surgeons and sites were eligible if the surgeon participated in more than 30 transplant operations per year and had performed more than 200 in their career, and the surgeon must practice in a multidisciplinary pituitary center environment. This is a concert flow diagram showing the patient allocations. Seven centers participated, with 15 surgeons, between 2015 and 2017. We enrolled 260 patients. Seventy-five patients were available for analysis of the primary endpoint in the microscopic group, and 166 were available for analysis in the endoscopic group. Basic patient demographics and tumor characteristics were similar between groups with respect to tumor invasiveness and tumor presentation. However, there's a very significant difference with respect to surgeon experience between the cohorts. So on average, the microscopic surgeons were significantly more experienced, both in terms of years of service and number of cases performed, compared to the endoscopic surgeons. As I mentioned, the primary endpoint was gross total resection. In the microscopic group, 80% of patients underwent gross total resection. In the endoscopic group, 83.7% had gross total resection, with a p-value of .58. We also examined several quality benchmarks. The median length of stay in both groups was two days, with no significant difference. Unplanned hospital readmission rates were also similar between groups. However, operative time was significantly shorter in the microscopic cohort by over an hour, and that is wheels-in to wheels-out time. We examined surgical complications and compared them between groups. There were no significant differences with respect to complications, with the exception of acute diabetes insipidus. So the rate of acute diabetes insipidus following surgery was significantly improved in the endoscopic group compared to the microscopic group. As part of the protocol, we examined several subgroups to specifically query whether certain subgroups of patients might benefit from one technique versus the other. We found no difference in the rates of gross total resection for large tumors greater than three centimeters. We found no difference with respect to gross total resection rates if the surgeon had intended on a gross total resection. We found no difference in virgin cases. We found a trend towards improved rates of gross total resection in the invasive tumors as graded by NOSP grade 3-4 criteria. To explore independent predictors of gross total resection, we conducted a series of univariate analyses and then chose the most interesting variables for this multivariate regression model. We found that the key independent predictors were what the goal was at the time of surgery by the surgeon, invasiveness of the tumor, and tumor size. Endoscopic technique also was an independent predictor of gross total resection with an odds ratio of 10. This was a non-randomized study. We tried to address potential confounders in the group by performing this propensity matched sample. Each group had 77 patients in this particular sample, and we matched for the confounders as you see here. The rate of gross total resection in the endoscopic propensity score matched sample was 90% versus 78% in the microscopic cohort. The strength of the study, its perspective, its multi-center, the primary and secondary endpoints were perspectively defined. The treatment groups were contemporaneous, and the imaging was centrally adjudicated. It's a large data set that I think gives a nice sampling of surgery centers across the United States. There are some criticisms or limitations that deserve discussion. One criticism this study has received is, well, these results are not new or surprising. However, I think it's important to make sure that as we trend towards more endoscopic surgery that the results are at least as equivalent to a microscopic approach. There were three surgical deaths in the series with a rate of just greater than 1%, which is similar to other retrospectively reported data and also similar to a self-reported survey published in 1997. The study is not randomized. We tried to address the potential confounders with propensity score matching. There's also a significant difference between groups with respect to the enthusiasm which investigators wanted to participate. Endoscopic surgeons are much more enthusiastic about participating, it's much easier to recruit endoscopic surgeons. And then finally, I think tumor recurrence would have been a more clinically meaningful endpoint in this study. However, because it's a benign disease, the length of follow-up would have been prohibitive, so we chose gross total resection as a surrogate endpoint for that. Importantly, we will be reporting the anterior gland dysfunction rates as the data matures and also reporting on volumetric extent of resection. So in conclusion, microscopic and endoscopic transferable approaches yield similar extents of gross total resection in an unadjusted analysis. There are some signals in the data that suggest that the endoscopic approach may be more favorable. Complications, length of stay, readmission rates are similar between groups and techniques. The endoscopic cohort had a lower rate of postoperative diabetes insipidus, suggesting it may be possible to identify the posterior gland better with the endoscopic approach, and this has been seen in other retrospective studies. However, the microscopic approach was much faster than the endoscopic approach. In the end, both surgical techniques allowed surgeons to accomplish their goals. Thank you. Commenting on the trial design and level of evidence will be Dr. Sepi Amin-Hinjani. Good morning. So Little and colleagues compared the rate of gross total resection between open surgical and endoscopic transplanoidal surgery in a non-randomized multicenter cohort. So although initial univariate comparison failed to show a difference in the outcome, subsequent multivariate analysis indicated that the endoscopic technique is strongly associated with gross total resection, a finding that the authors further reinforced with a propensity score matched analysis, showing an odds ratio of almost 30 in favor of the endoscopic technique. Although multivariate regression modeling is a common and useful technique to adjust for potentially confounding variables in non-randomized trials, it is important to note that such analysis rests on the assumption that the variables that are examined within the model are normally distributed and independent from each other. Rather, if such variables are highly correlated, such as tumor size, NOS grade, and resection goal may be, for example, the estimates for a given variable in the model can be inaccurate or change erratically with small changes in data. This may explain the dramatic difference in odds ratio between the univariate and multivariate modeling analysis and would require some careful consideration of techniques to assess and adjust for multicollinearity. Thank you. Dr. Nelson Oisiku will discuss the manuscript. Thank you, Dr. Simmons. This is an interesting study because in background, surgeons have adopted increasingly the endoscopic approach over the microscopic approach over the last few years. The question is whether or not there are superior outcomes based on analysis to support that trend. And so in this study, the extent of tumor resection was compared after endoscopic or microscopic surgery with the primary endpoint being gross total resection on MRI. So each technique has its own advantages and disadvantages, and here's my thoughts based on the results. First of all, the study design, unfortunately, is unable to equilibrate for surgeon factor. Microscopes and endoscopes, no matter how well engineered, do not operate. Surgeons do. Secondly, the gross total resection rate, my verdict is that it's essentially a wash. It's the same. And most importantly is that, in my opinion, and I think in the opinion of most neuroradiologists, an MRI cannot be truly blinded as to the approach because a good quality MRI scan is able to discern quite clearly whether or not the endoscopic approach was done or it was a microscopic approach. And that's based on the fact that the tissue removal is completely different in one versus the other. Length of stay, mortality, and readmission rates were washed, so the verdict there is the same. Yes, it's true that there was some findings in terms of DI and hyperendotremia, but reading the manuscript, there's insufficient data to make that judgment call. Microscopic patients had a shorter operative duration, and that is, in fact, the case, but does that really matter? This is not a race. This is an operative procedure, and it's the results that really count. And finally, as noted already in the statistical analysis, there was a finding in terms of endoscopic technique predicting GTR, but there are statistical issues regarding that finding anyway. So take-home messages, there are similar outcomes noted. Both techniques allow the surgeon to achieve its objectives, and choices should really be guided by what's best in the patient's interest, and each technique has its own merits and shortcomings. And obviously, over the years, technology will improve, and we shall be the better for it. Thank you very much. Thank you.

prospective multicenter study

microscopic transphenoidal surgery

fully endoscopic transphenoidal surgery

non-functioning pituitary adenomas

gross total resection