Mathematical model for predicting malignant transformation of low-grade gliomas

INTRODUCTION. Low grade gliomas (LGG) steadily undergo anaplastic transformation with progression to a higher grade of malignancy. Currently in clinical practice, there is no accurate method that can be used in predicting the risk of progression of a resected tumor that takes into account not only the multifactorial nature of glioma transformation but also the unequal influence of various predictors. AIM. To develop a mathematical model for predicting the risk of anaplastic transformation of low-grade malignant gliomas after surgical intervention. MATERIALS AND METHODS. The study group consisted of 52 patients who underwent a second surgery for advanced LGG between 2019 and 2023. Inclusion criteria: age at the time of diagnosis older than 18 years; histologically and molecularly genetically verified LGG; no tumor accumulation of contrast according to MRI before the first surgery; radiopharmaceutical accumulation index according to PET-CT with methionine less than 1.7; Ki-67 expression level less than 6 %; no vascular proliferation and endothelial swelling. Radiological characteristics included localization, tumor size, extent of resection, and contrast accumulation after progression. More than 90 parameters were analyzed to identify factors influencing the progression of LGG. The distribution of all parameters and their comparison were performed using the Mann – Whitney and Kolmogorov – Smirnov criteria, median x2. RESULTS. A model of probabilistic assessment of transformation risk with the following characteristics was built by means of logistic regression: sex; presence of seizures; tumor localization; histological data; postoperative treatment; time of progression. The role of the presented model is to obtain the characteristics of the logistic function Y for the standard equation, which makes it possible to calculate the probability of transformation depending on the factors and the degree of their influence on each other. CONCLUSION. Integration of this mathematical model into the plan of dynamic follow-up of patients with LGG will allow to adjust the strategy of observation, diagnosis and treatment.