Advancing Precision in Prostate Cancer (PCa) Diagnosis: A Comprehensive Analysis of 68Ga-PSMA PET/CT Semi-Quantitative and Texture Parameters

Introduction: Tumors are known to be heterogeneous on both gross and cellular levels, as well as genetic and phenotypic levels, with spatial heterogeneity in cellular density, angiogenesis, and necrosis. Furthermore, this heterogeneity might affect prognosis thereby impacting the treatment.

⁶⁸Ga-PSMA, a new PET tracer recently introduced to image patients with PCa for initial diagnosis and biochemical failure post-treatment. Semi-quantitative PET parameters help in assessing the tumour burden. However, PET/CT also has limitations in characterizing subtle tissue variations and the heterogeneity of tissue or tumors. Recently, texture analysis is being applied to study the spatial heterogeneity in tumors that involve the application of various mathematical methods to analyse the relationship between the grey level intensity of pixels or voxels and their position within an image, thereby providing an objective, quantitative assessment of tumour heterogeneity. Literature suggests that the texture features could enhance the diagnostic accuracy (DA) of prostate cancer (PCa). Thus, the present study is aimed to distinguish between healthy and malignant PCa based on ⁶⁸Ga-PSMA PET/CT semi-quantitative parameters and histogram texture features.

Methods: We ambispectively reviewed prostate cancer patients who underwent ⁶⁸Ga-PSMA PET/CT to know the disease status. Patients were categorized into healthy and malignant PCa patients based on PSMA expression on ⁶⁸Ga-PSMA PET/CT. PET semi-quantitative parameters SUVmax, SUVmean, TLG, MTV, Tumor to background ratio with respect to liver (T/B liver), Tumor to background ratio with respect to parotid (T/B parotid) were calculated on both group. Haralick texture features, histogram (8 features) were obtained as a result of texture analysis performed using MATLAB (v. 2018; MathWorks, Natick, MA, USA). Normality of the data was checked with Shapiro-Wilk test. Accordingly, Man Whitney U-test tests were applied to check the significant difference in the mean of PET/CT semi-quantitative and texture parameters between healthy and malignant PCa patients. Receiver Operative Characteristic (ROC) curve analysis was done on significant parameters to estimate the cut-off values along with corresponding sensitivity and specificity. All the analysis was done using SPSS software v22.0.

Results: A total 150 patients with healthy and malignant PCa patients, mean age= 68.18±7.08 years were included in this study. Of 150 PCa patients, analyses were performed in only 119 patients (37 healthy: 82 malignant PCa patients) and due to non-availability of Gleason score. Six of 8 Histogram features were found to be statistically significant. Also, all semi-quantitative PET parameters [SUVmax, SUVmean, TLG, MTV, T/B ratio (liver), T/B ratio (parotid)] were found to be statistically significant. Concordance and discordance of PET semi-quantitative parameters and texture features was determined based on the Gleason score as reference standard. The sensitivity, specificity, diagnostic accuracy (DA), positive predictive value (PPV) of the histogram texture features such as mean, median, mode, standard deviation, variance, entropy was found to be [63.64%; 83.87%; 91.8%; 68.91%], [68.18%; 80.65%; 90.91%; 71.43%], [68.18%; 74.19%; 88.24%; 69.75%], [73.86%; 96.77%; 98.48%; 79.83%], [90.91%; 100%; 100%; 93.28%]. Thus, entropy has the highest diagnostic accuracy of 93.28%. Similarly, among PET semi-quantitative parameters, TLG showed the highest diagnostic accuracy of 89.92%. However, the sensitivity, specificity, and PPV of TLG was found to be 87.5%; 96.77%; 89.92%.

Conclusions: TLG as a semi-quantitative PET parameter and entropy as a histogram texture parameter are particularly effective in accurately differentiating between individuals with a healthy prostate and those with malignant prostate cancer.