Noninvasive prediction of TP53 mutation in prostate cancer based on advanced diffusion weighted imaging.

Background: TP53 mutation is associated with poor prognosis and resistance to androgen deprivation therapy in prostate cancer (PCa). This study investigated whether stretched exponential model (SEM) and diffusion kurtosis imaging (DKI) could serve as a non-invasive predictor of TP53 mutation. Methods: This retrospective study included 84 patients with PCa who underwent radical prostatectomy. Clinical, clinicopathological, and quantitative imaging parameters were compared between the two groups using the independent-samples t test, chi-square test, or Fisher's exact test, as appropriate. Univariate and multivariate binary logistic regression analyses were performed to identify factors associated with TP53 mutation. Five logistic regression models were constructed: model 1 (based on DKI), model 2 (based on SEM), model 3 (based on mono-exponential model), model 4 (based on DKI, SEM and mono-exponential model) and model 5 (based on mean kurtosis (MK) value). Receiver operating characteristic analysis, DeLong test, Akaike information criterion (AIC), decision curve analysis (DCA), calibration curves, Hosmer-Lemeshow test, and bootstrap internal validation were used to evaluate model performance. Results: Compared with the TP53 wild-type group, the TP53 mutated group showed significantly lower apparent diffusion coefficient (ADC), distributed diffusion coefficient (DDC), and mean diffusivity (MD) values and a significantly higher MK value (all Conclusions: MK value is an independent predictor of TP53 mutation in PCa. MK value-based model demonstrated good diagnostic performance and outperformed the SEM and mono-exponential model, suggesting that MK value may serve as a promising preoperative, noninvasive tool for assessing TP53 mutation status in PCa.