← Back to Previous Page

 

S033: THREE-DIMENSIONAL TOPOGRAPHY OF PITUITARY MICROADENOMAS: A RETROSPECTIVE STUDY ON TUMOR DISTRIBUTION
Sandhya R Palit, MD¹; Yuki Shinya, MD, PhD^1,2; Dana Erickson, MD³; Justine S Herndon, PAC³; Irina Bancos, MD³; Jamie J Van Gompel, MD¹; ¹Department of Neurologic Surgery, Mayo Clinic, Rochester, Minnesota, USA; ²Department of Neurosurgery, University of Virginia, Charlottesville, Virginia, USA; ³Division of Endocrinology, Diabetes, and Nutrition, Mayo Clinic, Minnesota, USA

Introduction: The precise intra-glandular location of functioning pituitary microadenomas and its relationship to tumor characteristics and outcomes remains a topic of interest. While extensive research exists on pituitary adenomas, a systematic analysis of tumor distribution and outcomes based on a granular, topographical model is an important area of inquiry. To address this gap, this study investigated the intra-glandular topographical distribution of 134 functioning pituitary microadenomas by dividing the gland across three distinct planes. 

Methods: This retrospective study analyzed a cohort of 134 functioning pituitary microadenoma patients based on a review of their magnetic resonance imaging (MRI) scans. Tumors were radiologically graded by their location across three planes: horizontal (left, middle, right), anteroposterior, and superior/inferior by a neurosurgeon. Descriptive statistics were used to determine the frequency and distribution of tumors within each category. Chi-square tests of independence were then performed to identify statistically significant associations between the topographical variables. 

Results: Analysis of our 134-patient cohort revealed a non-uniform topographical distribution of pituitary microadenomas. No statistically significant association was found between a tumor's topographical location across the horizontal, anteroposterior, and superior/inferior planes and its hormonal subtype (p = 0.842, p = 0.775, p = 0.186, respectively). The analysis of surgical outcomes, however, revealed one notable finding: tumors located in the middle of the horizontal plane were associated with a statistically significant increase in the rate of intraoperative cerebrospinal fluid (CSF) leaks (p = 0.005), with a rate of 36.8% compared to 10.3% for left-sided and 8.5% for right-sided tumors.  

To further investigate remission and recurrence, time-to-event and binary outcome models were employed. A Cox proportional hazards model, which analyzed time to recurrence, found no statistically significant relationship between the tumor's location and the time until a subsequent intervention. In contrast, a logistic regression model, which examined the odds of achieving remission, revealed a significant three-way interaction between the horizontal, anteroposterior, and superior/inferior planes (p = 0.0277) (Figure 1). This indicates that while no single plane predicted remission, the specific combination of all three positional variables was a statistically significant predictor of a patient's remission status. This finding is supported by a detailed breakdown of remission rates by 3D location, which showed a wide variance in outcomes, ranging from a 100% remission rate for tumors in the Middle-Posterior-Inferior (M-P-I) quadrant to a 0% remission rate for those in the Middle-Posterior-Superior (M-P-S) quadrant. 

Conclusion: This study demonstrates that a tumor’s precise, three-dimensional location is a key predictor of surgical outcome, a relationship not captured by two-dimensional models. 

 

← Back to Previous Page