Objective To investigated the relationship between magnetic resonance imaging (MRI) parameters and the molecular pathology of primary central nervous system lymphoma (PCNSL).

Methods We retrospectively analyzed 26 patients from The First Affiliated Hospital of Harbin Medical University between January 2020 and June 2023 classified into germinal center B-cell like (GCB) and non-germinal center B-cell like （non-GCB）groups based on cell origin, into Ki-67≥75% and <75% groups based on the Ki-67 index, into BCL-2⁺ and BCL-2⁻ groups based on BCL-2 expression, and into responsive and non-responsive groups based on their response to MAP+Bruton's tyrosine kinase inhibitor（BTKi）treatment. We extracted and compared first-order parameters between the groups, including mean value, standard deviation, variance, coefficient of variation, skewness, kurtosis, and entropy from baseline MRI images.

Results Four parameters (variance, kurtosis, skewness, and coefficient of variation) showed no significant differences between groups. However, three parameters (mean, standard deviation, and entropy) significantly differed between the groups based on Ki-67 and BCL-2 expression. For the Ki-67 index, the three parameters’ areas under the curve (AUC) were 0.731, 0.831, and 0.913, respectively. For BCL-2 expression, the mean and standard deviation AUCs were 0.889 and 0.938, respectively. In addition, the mean and entropy parameters significantly differed between the groups categorized by cell origin and treatment responsiveness (P<0.05). Multi-parameter joint analysis demonstrated greater identification accuracy compared to utilizing individual quantitative parameters from texture analysis.

Conclusions The mean, standard deviation, and entropy MRI parameters can help predict Ki-67 and BCL-2 expression in patients with PCNSL and have evaluative functions for treatment. They are beneficial for preoperative non-invasive assessment of tumor malignancy, providing vidence for prognosis and treatment planning.