Refine
Document Type
- Article (10) (remove)
Language
- English (10)
Has Fulltext
- yes (10)
Is part of the Bibliography
- no (10)
Keywords
- Prostate cancer (10) (remove)
Institute
- Medizin (9)
- Informatik und Mathematik (2)
- Georg-Speyer-Haus (1)
The Gleason grading system remains the most powerful prognostic predictor for patients with prostate cancer since the 1960s. Its application requires highly-trained pathologists, is tedious and yet suffers from limited inter-pathologist reproducibility, especially for the intermediate Gleason score 7. Automated annotation procedures constitute a viable solution to remedy these limitations. In this study, we present a deep learning approach for automated Gleason grading of prostate cancer tissue microarrays with Hematoxylin and Eosin (H&E) staining. Our system was trained using detailed Gleason annotations on a discovery cohort of 641 patients and was then evaluated on an independent test cohort of 245 patients annotated by two pathologists. On the test cohort, the inter-annotator agreements between the model and each pathologist, quantified via Cohen’s quadratic kappa statistic, were 0.75 and 0.71 respectively, comparable with the inter-pathologist agreement (kappa = 0.71). Furthermore, the model’s Gleason score assignments achieved pathology expert-level stratification of patients into prognostically distinct groups, on the basis of disease-specific survival data available for the test cohort. Overall, our study shows promising results regarding the applicability of deep learning-based solutions towards more objective and reproducible prostate cancer grading, especially for cases with heterogeneous Gleason patterns.
Objectives: To analyze the performance of radiological assessment categories and quantitative computational analysis of apparent diffusion coefficient (ADC) maps using variant machine learning algorithms to differentiate clinically significant versus insignificant prostate cancer (PCa). Methods: Retrospectively, 73 patients were included in the study. The patients (mean age, 66.3 ± 7.6 years) were examined with multiparametric MRI (mpMRI) prior to radical prostatectomy (n = 33) or targeted biopsy (n = 40). The index lesion was annotated in MRI ADC and the equivalent histologic slides according to the highest Gleason Grade Group (GrG). Volumes of interest (VOIs) were determined for each lesion and normal-appearing peripheral zone. VOIs were processed by radiomic analysis. For the classification of lesions according to their clinical significance (GrG ≥ 3), principal component (PC) analysis, univariate analysis (UA) with consecutive support vector machines, neural networks, and random forest analysis were performed. Results: PC analysis discriminated between benign and malignant prostate tissue. PC evaluation yielded no stratification of PCa lesions according to their clinical significance, but UA revealed differences in clinical assessment categories and radiomic features. We trained three classification models with fifteen feature subsets. We identified a subset of shape features which improved the diagnostic accuracy of the clinical assessment categories (maximum increase in diagnostic accuracy ΔAUC = + 0.05, p < 0.001) while also identifying combinations of features and models which reduced overall accuracy. Conclusions: The impact of radiomic features to differentiate PCa lesions according to their clinical significance remains controversial. It depends on feature selection and the employed machine learning algorithms. It can result in improvement or reduction of diagnostic performance.
Background: Prostate cancer is a major health concern in aging men. Paralleling an aging society, prostate cancer prevalence increases emphasizing the need for efcient diagnostic algorithms.
Methods: Retrospectively, 106 prostate tissue samples from 48 patients (mean age,
66 ± 6.6 years) were included in the study. Patients sufered from prostate cancer (n = 38) or benign prostatic hyperplasia (n = 10) and were treated with radical prostatectomy or Holmium laser enucleation of the prostate, respectively. We constructed tissue microarrays (TMAs) comprising representative malignant (n = 38) and benign (n = 68) tissue cores. TMAs were processed to histological slides, stained, digitized and assessed for the applicability of machine learning strategies and open–source tools in diagnosis of prostate cancer. We applied the software QuPath to extract features for shape, stain intensity, and texture of TMA cores for three stainings, H&E, ERG, and PIN-4. Three machine learning algorithms, neural network (NN), support vector machines (SVM), and random forest (RF), were trained and cross-validated with 100 Monte Carlo random splits into 70% training set and 30% test set. We determined AUC values for single color channels, with and without optimization of hyperparameters by exhaustive grid search. We applied recursive feature elimination to feature sets of multiple color transforms.
Results: Mean AUC was above 0.80. PIN-4 stainings yielded higher AUC than H&E and
ERG. For PIN-4 with the color transform saturation, NN, RF, and SVM revealed AUC of 0.93 ± 0.04, 0.91 ± 0.06, and 0.92 ± 0.05, respectively. Optimization of hyperparameters improved the AUC only slightly by 0.01. For H&E, feature selection resulted in no increase of AUC but to an increase of 0.02–0.06 for ERG and PIN-4.
Conclusions: Automated pipelines may be able to discriminate with high accuracy between malignant and benign tissue. We found PIN-4 staining best suited for classifcation. Further bioinformatic analysis of larger data sets would be crucial to evaluate the reliability of automated classifcation methods for clinical practice and to evaluate potential discrimination of aggressiveness of cancer to pave the way to automatic precision medicine.
Background: To study the expression pattern, localisation and potential clinical significance of aquaporin water channels (AQP) both in prostate cancer (PC) cell lines and in benign and malignant human prostate tissue.
Methods: The AQP transcript and protein expression of HPrEC, LNCaP, DU-145 and PC3 cell lines was investigated using reverse transcriptase polymerase chain reaction (RT-PCR) and immunofluorescence (IF) microscopy labelling. Immunohistochemistry (IHC) was performed to assess AQP protein expression in surgical specimens of benign prostatic hyperplasia as well as in PC. Tissue mRNA expression of AQPs was quantified by single-step reverse transcriptase quantitative polymerase chain reaction (qPCR). Relative gene expression was determined using the 40-ΔCT method and correlated to clinicopathological parameters.
Results: Transcripts of AQP 1, 3, 4, 7, 8, 10 and 11 were expressed in all four cell lines, while AQP 9 transcripts were not detected in malignant cell lines. IF microscopy confirmed AQP 3, 4, 5, 7 and 9 protein expression. IHC revealed highly heterogeneous AQP 3 protein expression in PC specimens, with a marked decrease in expression in tumours of increasing malignancy. Loss of AQP 9 was shown in PC specimens. mRNA expression of AQP3 was found to be negatively correlated to PSA levels (ρ = − 0.354; p = 0.013), D’Amico risk stratification (ρ = − 0.336; p = 0.012), ISUP grade (ρ = − 0.321; p = 0.017) and Gleason score (ρ = − 0.342; p = 0.011).
Conclusions: This is the first study to systematically characterize human prostate cell lines, benign prostatic hyperplasia and PC in relation to all 13 members of the AQP family. Our results indicate the differential expression of several AQPs in benign and malignant prostate tissue. A significant correlation was observed between AQP 3 expression and tumour grade, with progressive loss in more malignant tumours. Taken together, AQPs may play a role in the progression of PC and AQP expression patterns may serve as a prognostic marker.
Background: Molecular markers for prostate cancer (PCa) are required to improve the early definition of patient outcomes. Atypically large extracellular vesicles (EVs), referred as "Large Oncosomes" (LO), have been identified in highly migratory and invasive PCa cells. We recently developed and characterized the DU145R80 subline, selected from parental DU145 cells as resistant to inhibitors of mevalonate pathway. DU145R80 showed different proteomic profile compared to parental DU145 cells, along with altered cytoskeleton dynamics and a more aggressive phenotype.
Methods: Immunofluorescence staining and western blotting were used to identify blebbing and EVs protein cargo. EVs, purified by gradient ultra-centrifugations, were analyzed by tunable resistive pulse sensing and multi-parametric flow cytometry approach coupled with high-resolution imaging technologies. LO functional effects were tested in vitro by adhesion and invasion assays and in vivo xenograft model in nude mice. Xenograft and patient tumor tissues were analyzed by immunohistochemistry.
Results: We found spontaneous blebbing and increased shedding of LO from DU145R80 compared to DU145 cells. LO from DU145R80, compared to those from DU145, carried increased amounts of key-molecules involved in PCa progression including integrin alpha V (αV-integrin). By incubating DU145 cells with DU145R80-derived LO we demonstrated that αV-integrin on LO surface was functionally involved in the increased adhesion and invasion of recipient cells, via AKT. Indeed either the pre-incubation of LO with an αV-integrin blocking antibody, or a specific AKT inhibition in recipient cells are able to revert the LO-induced functional effects. Moreover, DU145R80-derived LO also increased DU145 tumor engraftment in a mice model. Finally, we identified αV-integrin positive LO-like structures in tumor xenografts as well as in PCa patient tissues. Increased αV-integrin tumor expression correlated with high Gleason score and lymph node status.
Conclusions: Overall, this study is the first to demonstrate the critical role of αV-integrin positive LO in PCa aggressive features, adding new insights in biological function of these large EVs and suggesting their potential use as PCa prognostic markers.
Probably, patients with de novo (synchronous) and recurrent (metachronous) oligometastatic hormone-sensitive prostate cancer have different oncologic outcomes. Thus, we are challenged with different scenarios in clinical practice, where different treatment options may apply. In the last years, several prospective studies have focused on the treatment of patients with de novo oligometastatic hormone-sensitive prostate cancer. Not only the addition of systemic therapeutic treatments, such as chemotherapy with docetaxel, abiraterone, enzalutamide, and apalutamide, next to androgen deprivation therapy, demonstrated to improve outcomes in these patients but also local therapy of the primary has been demonstrated to improve outcomes of low-volume metastatic disease. Next to radiotherapy, also radical prostatectomy has been reported as a feasible and safe treatment option. Additional metastasis-directed therapy in de novo metastatic disease is currently examined by four trials. In the recurrent metastatic setting, less data are available, and it remains uncertain if patients can be treated in the same way as synchronous oligometastatic disease. Metastasis-directed therapy has demonstrated to prolong outcomes, while data on survival are still missing.
Background: The most recent overall survival (OS) and adverse event (AE) data have not been compared for the three guideline-recommended high-risk non-metastatic castration-resistant prostate cancer (nmCRPC) treatment alternatives.
Methods: We performed a systematic review and network meta-analysis focusing on OS and AE according to the most recent apalutamide, enzalutamide, and darolutamide reports. We systematically examined and compared apalutamide vs. enzalutamide vs. darolutamide efficacy and toxicity, relative to ADT according to PRISMA. We relied on PubMed search for most recent reports addressing prospective randomized trials with proven predefined OS benefit, relative to ADT: SPARTAN, PROSPER, and ARAMIS. OS represented the primary outcome and AEs represented secondary outcomes.
Results: Overall, data originated from 4117 observations made within the three trials that were analyzed. Regarding OS benefit relative to ADT, darolutamide ranked first, followed by enzalutamide and apalutamide, in that order. In the subgroup of PSA-doubling time (PSA-DT) ≤ 6 months patients, enzalutamide ranked first, followed by darolutamide and apalutamide in that order. Conversely, in the subgroup of PSA-DT 6–10 months patients, darolutamide ranked first, followed by apalutamide and enzalutamide, in that order. Regarding grade 3+ AEs, darolutamide was most favorable, followed by enzalutamide and apalutamide, in that order.
Conclusion: The current network meta-analysis suggests the highest OS efficacy and lowest grade 3+ toxicity for darolutamide. However, in the PSA-DT ≤ 6 months subgroup, the highest efficacy was recorded for enzalutamide. It is noteworthy that study design, study population, and follow-up duration represent some of the potentially critical differences that distinguish between the three studies and remained statistically unaccounted for using the network meta-analysis methodology. Those differences should be strongly considered in the interpretation of the current and any network meta-analyses.
Background: Measurement of prostate-specific antigen (PSA) advanced the diagnostic and prognostic potential for prostate cancer (PCa). However, due to PSA’s lack of specificity, novel biomarkers are needed to improve risk assessment and ensure optimal personalized therapy. A set of protein molecules as potential biomarkers was therefore evaluated in serum of PCa patients.
Methods: Serum samples from patients undergoing radical prostatectomy (RPE) for biopsy-proven PCa without neoadjuvant treatment were compared to serum samples from healthy subjects. Preliminary screening of 119 proteins in 10 PCa patients and 10 controls was carried out by the Proteome Profiler Antibody Array. Those markers showing distinct differences between patients and controls were then further evaluated by ELISA in the serum of 165 PCa patients and 19 controls. Uni- and multivariate as well as correlation analysis were performed to test the capability of these molecules to detect disease and predict pathological outcome.
Results: Screening showed that soluble (s)E-cadherin, E-selectin, MMP2, MMP9, TIMP1, TIMP2, Galectin and Clusterin warranted further evaluation. sE-Cadherin, TIMP1, Galectin and Clusterin were significantly over- and MMP9 under-expressed in PCa compared to controls. The concentration of sE-cadherin, MMP2 and Clusterin correlated negatively and that of MMP9 and TIMP1 positively with the Gleason Sum at prostatectomy. Only sE-cadherin significantly correlated with the highest Gleason pattern. Compared to serum PSA, sE-cadherin provided an independent and better matching predictive ability for discriminating PCas with an upgrade at RPE and aggressive tumors with a Gleason Sum ≥7.
Conclusions: sE-cadherin performed most favorably from a large panel of serum proteins in terms of diagnostic and predictive potential in curatively treatable PCa. sE-cadherin merits further investigation as a biomarker for PCa.
The aim of this study is to investigate the incidental prostate cancer (iPCa) detection rates of different embedding methods in a large, contemporary cohort of patients with bladder outlet obstruction (BOO) treated with transurethral surgery. We relied on an institutional tertiary-care database to identify BOO patients who underwent either transurethral loop resection or laser (Holmium:yttrium–aluminium garnet) enucleation of the prostate (HoLEP) between 01/2012 and 12/2019. Embedding methods differed with regard to the extent of the additional prostate tissue submitted following the first ten cassettes of primary embedding (cohort A: one [additional] cassette/10 g residual tissue vs. cohort B: complete embedding of the residual tissue). Detection rates of iPCa among the different embedding methods were compared. Subsequently, subgroup analyses by embedding protocol were repeated in HoLEP-treated patients only. In the overall cohort, the iPCa detection rate was 11% (46/420). In cohort A (n = 299), tissue embedding resulted in a median of 8 cassettes/patient (range 1–38) vs. a median of 15 (range 2–74) in cohort B (n = 121) (p < .001). The iPCa detection rate was 8% (23/299) and 19% (23/121) in cohort A vs. cohort B, respectively (p < .001). Virtual reduction of the number of tissue cassettes to ten cassettes resulted in a iPCa detection rate of 96% in both cohorts, missing one stage T1a/ISUP grade 1 carcinoma. Increasing the number of cassettes by two and eight cassettes, respectively, resulted in a detection rate of 100% in both cohorts without revealing high-grade carcinomas. Subgroup analyses in HoLEP patients confirmed these findings, demonstrated by a 100 vs. 96% iPCa detection rate following examination of the first ten cassettes, missing one case of T1a/ISUP 1. Examination of 8 additional cassettes resulted in a 100% detection rate. The extent of embedding of material obtained from transurethral prostate resection correlates with the iPCa detection rate. However, the submission of 10 cassettes appears to be a reasonable threshold to reduce resource utilization while maintaining secure cancer detection.
Background: Up- and/or downgrading rates in single intermediate-risk positive biopsy core are unknown.
Methods: We identified single intermediate-risk (Gleason grade group (GGG) 2/GGG3) positive biopsy core prostate cancer patients (≤ cT2c and PSA ≤ 20 ng/mL) within the Surveillance, Epidemiology, and End Results (SEER) database (2010–2015). Subsequently, separate uni- and multivariable logistic regression models tested for independent predictors of up- and downgrading.
Results: Of 1,328 assessable patients with single core positive intermediate-risk prostate cancer at biopsy, 972 (73%) harbored GGG2 versus 356 (27%) harbored GGG3. Median PSA (5.5 vs 5.7; p = 0.3), median age (62 vs 63 years; p = 0.07) and cT1-stage (77 vs 75%; p = 0.3) did not differ between GGG2 and GGG3 patients. Of individuals with single GGG2 positive biopsy core, 191 (20%) showed downgrading to GGG1 versus 35 (4%) upgrading to GGG4 or GGG5 at RP. Of individuals with single GGG3 positive biopsy core, 36 (10%) showed downgrading to GGG1 versus 42 (12%) significant upgrading to GGG4 or GGG5 at RP. In multivariable logistic regression models, elevated PSA (10–20 ng/mL) was an independent predictor of upgrading to GGG4/GGG5 in single GGG3 positive biopsy core patients (OR:2.89; 95%-CI: 1.31–6.11; p = 0.007).
Conclusion: In single GGG2 positive biopsy core patients, downgrading was four times more often recorded compared to upgrading. Conversely, in single GGG3 positive biopsy core patients, up- and downgrading rates were comparable and should be expected in one out of ten patients.