May 03, 2025
Since its founding in 1864, Mayo Clinic has been at the forefront of pituitary tumor treatment, giving hope to thousands of patients by combining cutting-edge expertise with compassionate team-based care. Since 2013, over 1,200 patients have been enrolled in the innovative Mayo Adenoma of the Pituitary Enterprise Registry (MAPER), built by Yuki Shinya, M.D., Ph.D., research fellow at Mayo Clinic, and Jamie J. Van Gompel, M.D., Neurologic Surgery at Mayo Clinic in Rochester, Minnesota, in collaboration with colleagues in Endocrinology and Otolaryngology. This extensive database has facilitated the development of prediction algorithms for treatment outcomes and helped more accurately develop tailored treatment plans to optimize patient outcomes.
A study published in the Journal of Neurosurgery analyzed 150 patients with Cushing disease, who were operated on primarily by two neurosurgeons between 2013 and 2023, and revealed that 72% achieved complete remission after the initial surgery. Intervention-free survival rates were 83% at three years and 78% at five years. Dr. Shinya notes: "We developed a supervised tree-based machine learning model that achieved a 91% accuracy rate in predicting treatment outcomes and identifying key predictive factors as tumor size, Knosp-Steiner grade, patient's age and body mass index. The model's high sensitivity (87%) and specificity (89%) enable personalized treatment planning, supporting Mayo Clinic's transition to active incorporation of AI-driven clinical care."
Dr. Van Gompel continues: "Our most recent study comprised of 100 eligible patients with growth hormone-secreting adenomas operated between 2013 and 2023 primarily by two neurosurgeons showed an 81% accuracy rate in predicting treatment outcomes using machine learning models. The study found a 67% intervention-free rate at five years; however, 32% of patients required additional intervention following initial surgery. The key predictive factors for better intervention-free survival were tumor size < 9 millimeters, complete tumor removal, patient's age < 65 years, and lower Knosp-Steiner grade according to our supervised machine learning model." The results of this study were published in a 2025 issue of the Journal of Neurological Surgery Part B: Skull Base and Neurosurgical Focus.
The results of these two published studies underpin Mayo Clinic's comprehensive approach to personalized care for pituitary tumors with the utilization of AI technology. This work leverages the recently built MAPER database to advance the goal of transforming clinical documentation into actionable treatment prediction. By improving and facilitating the clinical decision-making process, it aims to guide expectations of both patients and healthcare professionals, ultimately enhancing patient outcomes across the spectrum of pituitary disorders.
Dr. Shinya concludes, "Looking ahead, our team is launching a transformative AI initiative to automate pituitary tumor data extraction. This system will integrate radiomics features and enable real-time clinical updates, aiming to improve prognostic accuracy. Built on our proven machine learning models in Cushing disease and acromegaly, this framework will strengthen Mayo Clinic's position as a leader in AI-driven neuro-oncology, fostering multi-institutional collaboration through standardized protocols and shared implementation strategies."
For more information
Shinya Y, et al. Machine learning-based model to predict long-term tumor control and additional interventions following pituitary surgery for Cushing's disease. Journal of Neurosurgery. 2025. In press.
Shinya Y, et al. Machine learning-based model to predict long-term tumor control and additional interventions following transsphenoidal surgery for acromegaly patients. Journal of Neurological Surgery Part B: Skull Base. 2025;86:S1.
Shinya Y, et al. A supervised machine learning approach for predicting the need for post-surgical intervention in acromegaly. Neurosurgical Focus. 2025;13. In press.
Refer a patient to Mayo Clinic.