Development of an effective predictive screening tool for prostate cancer using the ClarityDX machine learning platform.

The current prostate cancer (PCa) screen test, prostate-specific antigen (PSA), has a high sensitivity for PCa but low specificity for high-risk, clinically significant PCa (csPCa), resulting in overdiagnosis and overtreatment of non-csPCa. Early identification of csPCa while avoiding unnecessary biopsies in men with non-csPCa is challenging. We built an optimized machine learning platform (ClarityDX) and showed its utility in generating models predicting csPCa. Integrating the ClarityDX platform with blood-based biomarkers for clinically significant PCa and clinical biomarker data from a 3448-patient cohort, we developed a test to stratify patients' risk of csPCa; called ClarityDX Prostate. When predicting high risk cancer in the validation cohort, ClarityDX Prostate showed 95% sensitivity, 35% specificity, 54% positive predictive value, and 91% negative predictive value, at a ≥ 25% threshold. Using ClarityDX Prostate at this threshold could avoid up to 35% of unnecessary prostate biopsies. ClarityDX Prostate showed higher accuracy for predicting the risk of csPCa than PSA alone and the tested model-based risk calculators. Using this test as a reflex test in men with elevated PSA levels may help patients and their healthcare providers decide if a prostate biopsy is necessary.

Formation of Previously Undescribed Δ7-Phytosterol Oxidation Products and Tocopherylquinone Adducts in Pumpkin Seed Oil during Roasting, Screw-Pressing, and Simulated Culinary Processing at Elevated Temperatures.

The influence of pumpkin seed roasting conditions (110-140 °C) and screw-pressing on the formation of previously undescribed Δ7-phytosterol oxidation products and tocopherylquinone adducts with nucleophilic phosphatidylethanolamine species was investigated. The roasting process of pumpkin seed paste at a temperature above 120 °C for 30 min considerably enhanced the formation of Δ7-oxysterols. Targeted analysis [electron impact mass spectrometry (MS), 1D-nuclear magnetic resonance] led to the identification of five novel markers of pumpkin paste roasting, among which (3ß,5α,22E,24S)-stigmasta-7,22-dien-6-one-3-ol (6-oxo-α-spinasterol), stereoisomers of (3ß,5α,22E)-7,8-epoxystigmast-22-en-3-ol (7,8-epoxy-α-spinasterol), and (3ß,5α)-22,23-epoxystigmast-7-en-3-ol (7,8-epoxy-α-spinasterol) were reported in edible oils for the first time. Simulated culinary processing provided novel stereoisomers of (3ß,5α,22E)-stigmasta-7,22-dien-3,6-diol, unusual (3ß,5α,22E)-stigmasta-7,22-dien-6,15-dione-3-ol, and (5α,22E)-stigmasta-7,22-dien-3-one accompanied by minor stereoisomers of (3ß,5α)-7,8;22,23-diepoxystigmastan-3-ol. Moreover, a clear relationship between the pumpkin seed oil stability index and synergistic effect of glycerophospholipids with present tocochromanols was found. High-resolution atmospheric pressure chemical ionization-MS experiments clearly demonstrated the formation of various γ-tocopherylquinone adducts with primary amines, namely, octylamine. The mitigation strategy of potentially detrimental oxysterols from pumpkin seed oil included optimization of processing parameters while maintaining the formation of desirable sensory-active compounds.

The predictive value of the prostate health index vs. multiparametric magnetic resonance imaging for prostate cancer diagnosis in prostate biopsy.

PURPOSE: To compare the ability of Prostate Health Index (PHI) to diagnose csPCa, with that of total PSA, PSA density (PSAD) and the multiparametric magnetic resonance (mpMRI) of the prostate. METHODS: We analysed a group of 395 men planned for a prostate biopsy who underwent a mpMRI of the prostate evaluated using the PIRADS v1 criteria. All patients had their PHI measured before prostate biopsy. In patients with an mpMRI suspicious lesions, an mpMRI/ultrasound software fusion-guided biopsy was performed first, with 12 core systematic biopsy performed in all patients. A ROC analysis was performed for PCa detection for total PSA, PSAD, PIRADS score and PHI; with an AUC curve calculated for all criteria and a combination of PIRADS score and PHI. Subsequent sub-analyses included patients undergoing first and repeat biopsy. RESULTS: The AUC for predicting the presence of csPCa in all patients was 59.5 for total PSA, 69.7 for PHI, 64.9 for PSAD and 62.5 for PIRADS. In biopsy naive patients it was 61.6 for total PSA, 68.9 for PHI, 64.6 for PSAD and 63.1 for PIRADS. In patients with previous negative biopsy the AUC for total PSA, PHI, PSAD and PIRADS was 55.4, 71.2, 64.4 and 69.3, respectively. Adding of PHI to PIRADS increased significantly (p = 0.007) the accuracy for prediction of csPCa. CONCLUSION: Prostate Health Index could serve as a tool in predicting csPCa. When compared to the mpMRI, it shows comparable results. The PHI cannot, however, help us guide prostate biopsies in any way, and its main use may, therefore, be in pre-MRI or pre-biopsy triage.

Preoperative prostate health index predicts adverse pathology and Gleason score upgrading after radical prostatectomy for prostate cancer.

BACKGROUND: We aimed to explore the utility of prostate specific antigen (PSA) isoform [- 2] proPSA and its derivatives for prediction of pathological outcome after radical prostatectomy (RP). METHODS: Preoperative blood samples were prospectively and consecutivelyanalyzed from 472 patients treated with RP for clinically localized prostate cancerat four medical centers. Measured parameters were PSA, free PSA (fPSA), fPSA/PSA ratio, [- 2] proPSA (p2PSA), p2PSA/fPSA ratio and Prostate Health Index (PHI)(p2PSA/fPSA)*√PSA]. Logistic regression models were fitted to determine the accuracy of markers for prediction of pathological Gleason score (GS) ≥7, Gleason score upgrading, extracapsular extension of the tumor (pT3) and the presence of positive surgical margin (PSM). The accuracy of predictive models was compared using area under the receiver operating curve (AUC). RESULTS: Of 472 patients undergoing RP, 339 (72%) were found to have pathologic GS ≥ 7, out of them 178 (53%) experienced an upgrade from their preoperative GS = 6. The findings of pT3 and PSM were present in 132 (28%) and 133 (28%) cases, respectively. At univariable analysis of all the preoperative parameters, PHI was the most accurate predictor of pathological GS ≥7 (OR 1.02, 95% CI 1.01-1.03, p<0.001), GS upgrading (OR 1.02, 95% CI 1.01-1.03, p<0.003), pT3 disease (OR 1.01, 95% CI 1.00-1.02, p<0.007) and the presence of PSM (OR 1.01, 95% CI 1.00-1.02, p<0.002). Adding of PHI into the base multivariable model increased significantly the accuracy for prediction of pathological GS by 4.4% to AUC = 66.6 (p = 0.015) and GS upgrading by 5.0% to AUC = 65.9 (p = 0.025), respectively. CONCLUSIONS: Preoperative PHI levels may contribute significantly to prediction of prostate cancer aggressiveness and expansion of the tumor detected at final pathology.

Cancer detection rates and inter-examiner variability of MRI/TRUS fusion targeted biopsy and systematic transrectal biopsy.

BACKGROUND: Software-based MRI/TRUS fusion biopsy depends on the coordination of several steps, and inter-examiner differences could influence the results. The aim of this bicentric prospective study was to compare the detection rates of MRI/TRUS fusion targeted biopsy (TG) and systematic biopsy (SB), and the detection rates of examiners with different levels of previous experience in prostate biopsy. METHODS: A total of 419 patients underwent MRI based on a suspicion of prostate cancer with elevated PSA levels. MRI was positive in 395 patients (221 in the first biopsy group [FB] and 174 in the repeated biopsy group [RB]). A subsequent TG, followed by a SB, was performed on these patients by four different examiners. RESULTS: In the detection of clinically significant prostate cancer, a significant difference was found for TG+SB against SB in the RB group (35.1% vs. 25.3%, P=0.047). In the detection of clinically insignificant prostate cancer, the SB had a significantly higher detection rate than TG in both subgroups (FB: 11.9% vs. 4.7%, P=0.008; RB: 13.8% vs. 6.9%, P=0.034). A significant difference was found between the four examiners in the FB for TG (P=0.028), SB (P=0.036), and TG+SB (P=0.017). CONCLUSION: MRI/TRUS TG in combination with SB had significantly higher detection rates than SB in the RB group only. Differences in detection rates between examiners were dependent on the level of previous experience with TRUS guided biopsy.

Use of Prostate Specific Antigen Density Combined with Multiparametric Magnetic Resonance Imaging Improves Triage for Prostate Biopsy.

BACKGROUND: Multi-parametric magnetic resonance imaging (mpMRI)-directed biopsy for prostate cancer (PC) diagnosis improves the detection of clinically significant prostate cancer (CSPC) and decreases the rate of over-diagnosis of insignificant disease. The aim of this study was to investigate the value of mpMRI combined with prostate specific antigen density (PSAD) in the decision making related to the biopsy. METHODS: mpMRI and mpMRI/transrectal ultrasound fusion targeted biopsies with subsequent systematic biopsies were performed in 397 patients (223 biopsy-naïve and 174 with a previous biopsy). Detection rates of (CSPC) and insignificant PC were stratified using the PIRADS score, and the number of avoidable biopsies and missed (CSPC) were plotted against PSAD values of 0.1-0.5 ng/mL2. RESULTS: PIRADS <3 and PSAD <0.2 ng/mL2 were the safest criteria for not performing a biopsy. When applied to the biopsy-naïve group, 21.5% (48/223) of the biopsies could have been avoided and 3.7% (3/82) of CSPC would have been missed. In the repeat biopsy group, 12.6% (22/174) of biopsies could have been avoided and 6.9% (4/58) of (CSPC) would have been missed. CONCLUSIONS: A combination of mpMRI and PSAD might reduce the number of biopsies performed with the cost of missing <4% of CSPC.