Objective:

To explore the predictive factors of pathological grade of brain stem glioma and its relationship with amide proton transfer (APT) imaging

in order to provide reference for early diagnosis and recognition of high-grade glioma patients and subsequent prediction model construction.

Methods:

A total of 79 brain stem glioma patients under

went surgical treatment and APT imaging were retrospectively chosen in the period from January 2019 to January 2023 in the 909th Hospital and grouped according to histopathological grading into high-grade group (38 cases) and low-grade group (41 cases). The independent predictors of pathological grade of brain stem glioma were evaluated by univariate factor and multivariate factor method. Predictive efficacies of prognostic factors for pathological grading of brainstem gliomas were analyzed.

Results:

The results of univariate analysis showed that cystic lesions

maximum lesion diameter

mean and maximum APT imaging signal intensity may be related to the pathological grade of brain stem glioma (

P

＜0.05). The results of multi-factor analysis showed that cystic lesions

maximum lesion diameter ≥2 cm

average and maximum APT imaging signal intensity were independent predictors of the pathological grade of brain stem glioma (

P

＜0.05). Cystic lesions

maximum lesion diameter ≥2 cm

mean value of APT imaging signal intensity

maximum value of APT imaging signal intensity and logistic model prediction probability were used to predict the pathological grade of brain stem glioma. The optimal cut- off values were 0.50

0.50

2.95%

4.11% and 37.85%

respectively. The Yoden index was 38.19%

44.42%

51.73%

42.17% and 65.28%

respectively.

Conclusion:

Cystic lesion

maximum lesion diameter and APT imaging signal intensity can be independent predictors of brain stem glioma pathological grade

and mean APT imaging signal intensity has better predictive efficacy.