Transition Zone Prostate-specific Antigen Density Could Better Guide the Rebiopsy Strategy in Men With Prostate Inflammation at Initial Biopsy

Deep Learning Whole-Gland and Zonal Prostate Segmentation on a Public MRI Dataset

BACKGROUND

Prostate volume, as determined by magnetic resonance imaging (MRI), is a useful biomarker both for distinguishing between benign and malignant pathology and can be used either alone or combined with other parameters such as prostate-specific antigen.

PURPOSE

This study compared different deep learning methods for whole-gland and zonal prostate segmentation.

STUDY TYPE

Retrospective.

POPULATION

A total of 204 patients (train/test = 99/105) from the PROSTATEx public dataset.

FIELD STRENGTH/SEQUENCE

A 3 T, TSE T₂ -weighted.

ASSESSMENT

Four operators performed manual segmentation of the whole-gland, central zone + anterior stroma + transition zone (TZ), and peripheral zone (PZ). U-net, efficient neural network (ENet), and efficient residual factorized ConvNet (ERFNet) were trained and tuned on the training data through 5-fold cross-validation to segment the whole gland and TZ separately, while PZ automated masks were obtained by the subtraction of the first two.

STATISTICAL TESTS

Networks were evaluated on the test set using various accuracy metrics, including the Dice similarity coefficient (DSC). Model DSC was compared in both the training and test sets using the analysis of variance test (ANOVA) and post hoc tests. Parameter number, disk size, training, and inference times determined network computational complexity and were also used to assess the model performance differences. A P < 0.05 was selected to indicate the statistical significance.

RESULTS

The best DSC (P < 0.05) in the test set was achieved by ENet: 91% ± 4% for the whole gland, 87% ± 5% for the TZ, and 71% ± 8% for the PZ. U-net and ERFNet obtained, respectively, 88% ± 6% and 87% ± 6% for the whole gland, 86% ± 7% and 84% ± 7% for the TZ, and 70% ± 8% and 65 ± 8% for the PZ. Training and inference time were lowest for ENet.

DATA CONCLUSION

Deep learning networks can accurately segment the prostate using T₂ -weighted images.

EVIDENCE LEVEL

4 TECHNICAL EFFICACY: Stage 2.

Zonal adjusted PSA density improves prostate cancer detection rates compared with PSA in Taiwanese males with PSA < 20 ng/ml

BACKGROUND

The current study aimed to compare the efficacy of transition zone PSA density (TZPSAD) with traditional PSA and PSA density (PSAD), for the diagnosis of prostate cancer (PCa) in Taiwanese males.

METHODS

Men with PSA between 4.0 and 20.0 ng/ml who underwent a transrectal ultrasound (TRUS) guided prostate biopsy between the studied period were retrospectively identified. The demographic data, PSAD and TZPSAD were calculated in all patients. Receiver operating characteristic (ROC) curves were used to analyze the accuracy of a positive PCa diagnosis.

RESULTS

The area under the ROC (AUC) was 0.615, 0.748 and 0.746 for PSA, PSAD and TZPSAD, respectively. The best cut-off of value for TZPSAD in predicting PCa in men with a PSA of 4.0-10.0 ng/ml was 0.367 ng/ml/ml with a sensitivity of 50% and a specificity of 77.5%. In men with a PSA of 10.1-20.0 ng/ml, the best cut-off value was 0.454 ng/ml, with a sensitivity of 74.8% and specificity of 70.9%.

CONCLUSION

The use of TZPSAD can improve the efficiency and specificity of PSA for the diagnosis of PCa in Taiwanese men with PSA 4.0-20.0 ng/ml. TZPSAD efficiency was similar to PSAD but TZPSAD had better cancer specificity.

Prostatic chronic inflammation and prostate cancer risk at baseline random biopsy: Analysis of predictors

Objective

To evaluate predictors of prostatic chronic inflammation (PCI) and prostate cancer (PCa) in patients undergoing transperineal baseline random prostatic needle biopsies (BNB).

Patient and methods

According to BNB outcomes, patients were divided into four groups: cases without PCI or PCa (Control group), cases with PCI only (PCI group), cases with PCa and PCI (PCa+PCI group) and cases with PCa only (PCa group). A multinomial logistic regression model was used to evaluate the association of clinical factors with BNB outcomes. Additionally, clinical factors associated with the risk of PCa in the overall population were investigated using a multivariable logistic regression model (univariate and multivariate analysis).

Results

Overall, 945 patients were evaluated and grouped as follows: Control group, 308 patients (32.6%); PCI group, 160 (16.9%); PCa+PCI group, 45 (4.8%); and PCa group, 432 (45.7%). Amongst these, PCa was independently predicted by age (odds ratio [OR] 1.081), prostate specific-antigen level (PSA; OR 1.159), transition zone volume (TZV; OR 0.916), and abnormal digital rectal examination (DRE; OR 1.962). PCa and PCI (4.8%) were independently predicted by age (OR 1.081), PSA level (OR 1.122) and TZV (OR 0.954). In the group without PCa, the PSA level was the only factor associated with the risk of PCI when compared to the control group (OR 1.051, P = 0.042). Among patients with PCa, independent factors associated with the risk of only PCa compared to cases with PCA+PCI were TZV (OR 0.972) and number of positive cores (OR 1.149). In the overall population, PCI was the strongest predictor of a decreased risk of PCa (multivariate model, OR 0.212; P < 0.001)

Conclusions

At BNB, PCI was associated with both a decreased risk of PCa and less aggressive tumour biology amongst patients with PCa. The presence of PCI on biopsy cores should be reported because of its implications in clinical practice.

Abbreviations

BGG: biopsy Gleason Group; BPC: biopsy positive (cancer) cores; BMI: body mass index; FGF-2: fibroblast growth factor 2; IL: interleukin; ISUP: International Society of Urologic Pathology; NIH: National Institutes of Health; OR: odds ratio; PCa: prostate cancer; PCI: prostatic chronic inflammation; TGF: transforming growth factor; TPV: total prostate volume; TZV: transition zone volume

Impact of PSA density of transition zone as a potential parameter in reducing the number of unnecessary prostate biopsies in patients with psa levels between 2.6 and 10.0 ng/mL

Purpose:

To assess the accuracy of prostate-specific antigen (PSA) adjusted for the transition zone volume (PSATZ) in predicting prostate cancer by comparing the ability of several PSA parameters in predicting prostate cancer in men with intermediate PSA levels of 2.6 – 10.0 ng/mL and its ability to reduce unnecessary biopsies.

Materials and Methods:

This study included 656 patients referred for prostate biopsy who had a serum PSA of 2.6 – 10.0 ng/mL. Total prostate and transition zone volumes were measured by transrectal ultrasound using the prolate ellipsoid method. The clinical values of PSA, free-to-total (F/T) ratio, PSA density (PSAD) and PSATZ for the detection of prostate cancer were calculated and statistical comparisons between biopsy-positive (cancer) and biopsy-negative (benign) were conducted.

Results:

Cancer was detected in 172 patients (26.2%). Mean PSA, PSATZ, PSAD and F/T ratio were 7.5 ng/mL, 0.68 ng/mL/cc. 0.25 ng/mL/cc and 0.14 in patients with prostate cancer and 6.29 ng/mL, 0.30 ng/mL/cc, 0.16 ng/mL/cc and 0.22 in patients with benign biopsies, respectively. ROC curves analysis demonstrated that PSATZ had a higher area under curve (0,838) than F/T ratio (0.806) (P<0.001) and PSAD (0.806) (P<0.001). With a cut-off value of 0.22 ng/mL/cc, PSATZ had 100% of sensitivity and could have prevented 24% of unnecessary biopsies.

Conclusions:

PSATZ may be useful in enhancing the specificity of serum PSA. Compared to other PSA related parameters, it was better in differentiating between prostate cancer and benign prostatic enlargement. Also, PSATZ could reduce a significant number of unnecessary biopsies.

Inflammation on Prostate Needle Biopsy is Associated with Lower Prostate Cancer Risk: A Meta-Analysis

PURPOSE

We performed a comprehensive literature review and meta-analysis to evaluate the association of inflammation on prostate needle biopsies and prostate cancer risk.

MATERIALS AND METHODS

We searched Embase®, PubMed® and Web of Science™ from January 1, 1990 to October 1, 2016 for abstracts containing the key words prostate cancer, inflammation and biopsy. Study inclusion criteria were original research, adult human subjects, cohort or case-control study design, histological inflammation on prostate needle biopsy and prostate cancer on histology. Two independent teams reviewed abstracts and extracted data from the selected manuscripts. Combined ORs and 95% CIs of any, acute and chronic inflammation were calculated using the random effects method.

RESULTS

Of the 1,030 retrieved abstracts 46 underwent full text review and 25 were included in the final analysis, comprising a total of 20,585 subjects and 6,641 patients with prostate cancer. There was significant heterogeneity among studies (I² = 84.4%, p <0.001). The presence of any inflammation was significantly associated with a lower prostate cancer risk in 25 studies (OR 0.455, 95% CI 0.337-0.573). There was no evidence of publication bias (p >0.05). When subanalyzed by inflammation type, acute inflammation in 4 studies and chronic inflammation in 15 were each associated with a lower prostate cancer risk (OR 0.681, 95% CI 0.450-0.913 and OR 0.499, 95% CI 0.334-0.665, respectively).

CONCLUSIONS

In a meta-analysis of 25 studies inflammation on prostate needle biopsy was associated with a lower prostate cancer risk. Clinically the presence of inflammation on prostate needle biopsy may lower the risk of a subsequent prostate cancer diagnosis.