A Hierarchical Machine Learning Model to Discover Gleason Grade Group-specific Biomarkers in Prostate Cancer

1) Background: One of the deadliest cancers that affect men worldwide and North American men is prostate cancer. This disease motivates parts of the cells in the prostate to lose control of their growth and division. 2) Methods: We are proposing a machine learning method used to analyze gene expressions of prostate tumors with different Gleason scores, and to identify potential genetic biomarkers for each group. A publicly-available RNA-Seq dataset of a cohort of 104 prostate cancer patients have been retrieved from the National Center for Biotechnology Information's (NCBI) Gene Expression Omnibus (GEO) repository. We categorize patients by their Gleason scores into different groups to create a hierarchy of disease progression. A hierarchical model with standard classifiers in different Gleason groups (hereinafter called nodes) to identify and predict nodes based on their mRNA or gene expressions. At each node, patient samples are analyzed via class imbalance and hybrid feature selection techniques to build the prediction model. The outcome of each node is a set of genes that can separate the Gleason group from the remaining groups. To validate the proposed method, the set of identified genes are used to classify a second dataset of 499 prostate cancer patients that have been collected from cBioportal.. 3) Results: Two genes have been found to be potential biomarkers of specific Gleason groups; PIAS3 has been identifed for Gleason score 4+3=7, while UBE2 could be a poteintial biomarker for Gleason score 6. Other proposed genes that were not found in the literature might be potential biomarkers. 4) Conclusion: The latest literature supports that the genes predicted by the proposed method are strongly correlated with prostate cancer progression and tumour development processes. Furthermore, pathway analysis shows that both PIAS3 and UBE2 share the same protein interaction pathway, the JAK/STAT signaling process.

Download Full-text

A Hierarchical Machine Learning Model to Discover Gleason Grade Group-specific Biomarkers in Prostate Cancer

10.20944/preprints201911.0298.v2 ◽

2019 ◽

Author(s):

Osama Hamzeh ◽

Abedalrhman Alkhateeb ◽

Julia Zheng ◽

Srinath Kandalam ◽

Crystal Lueng ◽

...

Keyword(s):

Gene Expression ◽

Prostate Cancer ◽

Machine Learning ◽

Cancer Patients ◽

Gleason Score ◽

Class Imbalance ◽

Gleason Grade ◽

Grade Group ◽

Next Generation Sequencing Technology ◽

Potential Biomarker

1) Background: One of the most common cancer that affects men worldwide and North American men is prostate cancer. Gleason score is a pathological grading system to examine the potential aggressiveness of the disease in the prostate tissue. The advancement in computing and next-generation sequencing technology now allow us to study the genomic profiles of patients in association with their different Gleason score more accurately and effectively. 2) Methods: In this study, we used a novel machine learning method to analyze gene expression of prostate tumors with different Gleason scores, and identify potential genetic biomarkers for each Gleason group. We obtained a publicly-available RNA-Seq dataset of a cohort of 104 prostate cancer patients from the National Center for Biotechnology Information’s (NCBI) Gene Expression Omnibus (GEO) repository, and categorized patients based on their Gleason scores to create a hierarchy of disease progression. A hierarchical model with standard classifiers in different Gleason groups, also known as nodes, was developed to identify and predict nodes based on their mRNA or gene expression. In each node, patient samples were analyzed via class imbalance and hybrid feature selection techniques to build the prediction model. The outcome from analysis of each node is a set of genes that can differentiate each Gleason group from the remaining groups. To validate the proposed method, the set of identified genes are used to classify a second dataset of 499 prostate cancer patients collected from cBioportal [1]. 3) Results: The overall accuracy of applying this novel method to the first dataset was 93.3%, and further validated to 87% accuracy using the second dataset. This method also identified genes that were not previously reported as potential biomarkers for specific Gleason groups. In particular, PIAS3 was identified as a potential biomarker for Gleason score 4+3=7, and UBE2V2 for Gleason score 6. 4) Insight: Previous reports show that the genes predicted by this newly proposed method strongly correlate with prostate cancer development and progression. Furthermore, pathway analysis shows that both PIAS3 and UBE2V2 share similar protein interaction pathways, the JAK/STAT signaling process.

Download Full-text

A Hierarchical Machine Learning Model to Discover Gleason Grade-Specific Biomarkers in Prostate Cancer

Diagnostics ◽

10.3390/diagnostics9040219 ◽

2019 ◽

Vol 9 (4) ◽

pp. 219 ◽

Cited By ~ 3

Author(s):

Osama Hamzeh ◽

Abedalrhman Alkhateeb ◽

Julia Zhuoran Zheng ◽

Srinath Kandalam ◽

Crystal Leung ◽

...

Keyword(s):

Gene Expression ◽

Prostate Cancer ◽

Machine Learning ◽

Cancer Patients ◽

Gleason Score ◽

Class Imbalance ◽

Gene Expression Omnibus ◽

Gleason Grade ◽

Next Generation Sequencing Technology ◽

Potential Biomarker

(1) Background:One of the most common cancers that affect North American men and men worldwide is prostate cancer. The Gleason score is a pathological grading system to examine the potential aggressiveness of the disease in the prostate tissue. Advancements in computing and next-generation sequencing technology now allow us to study the genomic profiles of patients in association with their different Gleason scores more accurately and effectively. (2) Methods: In this study, we used a novel machine learning method to analyse gene expression of prostate tumours with different Gleason scores, and identify potential genetic biomarkers for each Gleason group. We obtained a publicly-available RNA-Seq dataset of a cohort of 104 prostate cancer patients from the National Center for Biotechnology Information’s (NCBI) Gene Expression Omnibus (GEO) repository, and categorised patients based on their Gleason scores to create a hierarchy of disease progression. A hierarchical model with standard classifiers in different Gleason groups, also known as nodes, was developed to identify and predict nodes based on their mRNA or gene expression. In each node, patient samples were analysed via class imbalance and hybrid feature selection techniques to build the prediction model. The outcome from analysis of each node was a set of genes that could differentiate each Gleason group from the remaining groups. To validate the proposed method, the set of identified genes were used to classify a second dataset of 499 prostate cancer patients collected from cBioportal. (3) Results: The overall accuracy of applying this novel method to the first dataset was 93.3%; the method was further validated to have 87% accuracy using the second dataset. This method also identified genes that were not previously reported as potential biomarkers for specific Gleason groups. In particular, PIAS3 was identified as a potential biomarker for Gleason score 4 + 3 = 7, and UBE2V2 for Gleason score 6. (4) Insight: Previous reports show that the genes predicted by this newly proposed method strongly correlate with prostate cancer development and progression. Furthermore, pathway analysis shows that both PIAS3 and UBE2V2 share similar protein interaction pathways, the JAK/STAT signaling process.

Download Full-text

Androgen Receptor Expression of Prostate Cancer Correlates with Gleason Score and Perineural Invasion in West Sumatera, Indonesia

International Journal Of Medical Science And Clinical Invention ◽

10.18535/ijmsci/v7i11.010 ◽

2020 ◽

Vol 7 (11) ◽

pp. 5125-5129

Author(s):

Anandia Putriyuni ◽

Meta Zulyati Oktora

Keyword(s):

Prostate Cancer ◽

Androgen Receptor ◽

Gleason Score ◽

Cancer Progression ◽

Perineural Invasion ◽

Receptor Expression ◽

Grade Group ◽

Who Grade ◽

Who Grade Group ◽

Histopathological Grading

Prostate cancer is the second most common and the fifth leading cause of death by cancer in men worldwide now. The failure of androgen deprivation therapy (ADT) for prostate cancer caused by activated androgen receptor (AR) signaling pathways mostly found. The role of AR in growth and progression of prostate cancer is still unclear. Analysis of AR expression in prostate cancer has never been done in West Sumatera. This study aims to determine AR expression of prostate cancer and correlate with Gleason score and perineural invasion. A total of 56 prostate cancer from department of anatomical pathology in West Sumatera. Hematoxylin and eosin (HE) stained slides and paraffin blocks were retrieved. Slides of all cases were evaluated to review Gleason score, histopathological grading, WHO grade group based on ISUP 2014/WHO 2016 and perineural invasion. Androgen receptor immunohistochemistry (IHC) was applied on all cases. High AR expression was the mostly found (51,79%). The mostly prostate cancer is Gleason score 9 (44,64%), histopathological grading poorly differentiated/undifferentiated (76,78%), WHO grade group 5 (48,21%). Perineural invasion was noted in 39,29%. There was significant statistical correlation between AR expression and Gleason score, but no significant correlation with perineural invasion. AR expression is the important marker of prostate cancer progression.

Download Full-text

Natural history of prostate cancer on active surveillance: stratification by MRI using the PRECISE recommendations in a UK cohort

European Radiology ◽

10.1007/s00330-020-07256-z ◽

2020 ◽

Cited By ~ 2

Author(s):

Francesco Giganti ◽

Armando Stabile ◽

Vasilis Stavrinides ◽

Elizabeth Osinibi ◽

Adam Retter ◽

...

Keyword(s):

Prostate Cancer ◽

Active Surveillance ◽

Gleason Score ◽

Rank Test ◽

Gleason Grade ◽

Clinical Progression ◽

Radiological Progression ◽

Grade Group ◽

Psa Density

Abstract Objectives The PRECISE recommendations for magnetic resonance imaging (MRI) in patients on active surveillance (AS) for prostate cancer (PCa) include repeated measurement of each lesion, and attribution of a PRECISE radiological progression score for the likelihood of clinically significant change over time. We aimed to compare the PRECISE score with clinical progression in patients who are managed using an MRI-led AS protocol. Methods A total of 553 patients on AS for low- and intermediate-risk PCa (up to Gleason score 3 + 4) who had two or more MRI scans performed between December 2005 and January 2020 were included. Overall, 2161 scans were retrospectively re-reported by a dedicated radiologist to give a PI-RADS v2 score for each scan and assess the PRECISE score for each follow-up scan. Clinical progression was defined by histological progression to ≥ Gleason score 4 + 3 (Gleason Grade Group 3) and/or initiation of active treatment. Progression-free survival was assessed using Kaplan-Meier curves and log-rank test was used to assess differences between curves. Results Overall, 165/553 (30%) patients experienced the primary outcome of clinical progression (median follow-up, 74.5 months; interquartile ranges, 53–98). Of all patients, 313/553 (57%) did not show radiological progression on MRI (PRECISE 1–3), of which 296/313 (95%) had also no clinical progression. Of the remaining 240/553 patients (43%) with radiological progression on MRI (PRECISE 4–5), 146/240 (61%) experienced clinical progression (p < 0.0001). Patients with radiological progression on MRI (PRECISE 4-5) showed a trend to an increase in PSA density. Conclusions Patients without radiological progression on MRI (PRECISE 1-3) during AS had a very low likelihood of clinical progression and many could avoid routine re-biopsy. Key Points • Patients without radiological progression on MRI (PRECISE 1–3) during AS had a very low likelihood of clinical progression and many could avoid routine re-biopsy. • Clinical progression was almost always detectable in patients with radiological progression on MRI (PRECISE 4–5) during AS. • Patients with radiological progression on MRI (PRECISE 4–5) during AS showed a trend to an increase in PSA density.

Download Full-text

The Combined Percentage of Gleason Patterns 4 and 5 Is the Best Predictor of Cancer Progression After Radical Prostatectomy

Journal of Clinical Oncology ◽

10.1200/jco.2005.03.018 ◽

2005 ◽

Vol 23 (13) ◽

pp. 2911-2917 ◽

Cited By ~ 75

Author(s):

Liang Cheng ◽

Michael O. Koch ◽

Beth E. Juliar ◽

Joanne K. Daggy ◽

Richard S. Foster ◽

...

Keyword(s):

Prostate Cancer ◽

Radical Prostatectomy ◽

Cancer Patients ◽

Gleason Score ◽

Cancer Progression ◽

Tumor Volume ◽

High Grade ◽

Cox Regression Analysis ◽

Total Tumor Volume ◽

Increased Risk

Purpose Clinical outcome is variable in prostate cancer patients treated with radical prostatectomy. The Gleason histologic grade of prostatic adenocarcinoma is one of the strongest predictors of biologic aggressiveness of prostate cancer. We evaluated the significance of the relative proportion of high-grade cancer (Gleason patterns 4 and/or 5) in predicting cancer progression in prostate cancer patients treated with radical prostatectomy. Patients and Methods Radical prostatectomy specimens from 364 consecutive prostate cancer patients were totally embedded and whole mounted. Various clinical and pathologic characteristics were analyzed. All pathologic data, including Gleason grading variables, were collected prospectively. Results A multiple-factor analysis was performed that included the combined percentage of Gleason patterns 4 and 5, Gleason score, tumor stage, surgical margin status, preoperative prostate-specific antigen (PSA), extraprostatic extension, and total tumor volume. Using Cox regression analysis with bootstrap resampling for predictor selection, we identified the combined percentage of Gleason patterns 4 and 5 (P < .0001) and total tumor volume (P = .009) as significant predictors of PSA recurrence. Conclusion The combined percentage of Gleason patterns 4 and 5 is one of the most powerful predictors of patient outcome, and appears superior to conventional Gleason score in identifying patients at increased risk of disease progression. On the basis of our results, we recommend that the combined percentage of Gleason patterns 4 and 5 be evaluated in radical prostatectomy specimens. The amount of high-grade cancer in a prostatectomy specimen should be taken into account in therapeutic decision making and assessment of patient prognosis.

Download Full-text

The Utilization of Gleason Grade as the Primary Criterion for Ordering Nuclear Bone Scan in Newly Diagnosed Prostate Cancer Patients

The Scientific World JOURNAL ◽

10.1100/tsw.2009.113 ◽

2009 ◽

Vol 9 ◽

pp. 1040-1045 ◽

Cited By ~ 9

Author(s):

Chad W. M. Ritenour ◽

John T. Abbott ◽

Michael Goodman ◽

Naomi Alazraki ◽

Fray F. Marshall ◽

...

Keyword(s):

Prostate Cancer ◽

Cancer Patients ◽

Gleason Score ◽

Bone Scan ◽

Stage Migration ◽

Gleason Grade ◽

Newly Diagnosed ◽

High Gleason Score ◽

Bone Scans ◽

The Relationship

Utilization of nuclear bone scans for staging newly diagnosed prostate cancer has decreased dramatically due to PSA-driven stage migration. The current criteria for performing bone scans are based on limited historical data. This study evaluates serum PSA and Gleason grade in predicting positive scans in a contemporary large series of newly diagnosed prostate cancer patients. Eight hundred consecutive cases of newly diagnosed prostate cancer over a 64-month period underwent a staging nuclear scan. All subjects had histologically confirmed cancer. The relationship between PSA, Gleason grade, and bone scan was examined by calculating series of crude, stratified, and adjusted odds ratios with corresponding 95% confidence intervals. Four percent (32/800) of all bone scans were positive. This proportion was significantly lower in patients with Gleason score ≤7 (1.9%) vs. Gleason score ≥8 (18.8%,p< 0.001). Among patients with Gleason score ≤7, the rate of positive bones scans was 70-fold higher when the PSA was >30 ng/ml compared to ≤30 ng/ml (p< 0.001). For Gleason score ≥8, the rate was significantly higher (27.9 vs. 0%) when PSA was >10 ng/ml compared to ≤10 ng/ml (p= 0.002). The combination of Gleason score and PSA enhances predictability of bone scans in newly diagnosed prostate cancer patients. The PSA threshold for ordering bone scans should be adjusted according to Gleason score. For patients with Gleason scores ≤7, we recommend a bone scan if the PSA is >30 ng/ml. However, for patients with a high Gleason score (8–10), we recommend a bone scan if the PSA is >10 ng/ml.

Download Full-text

Prediction of Gleason Grade Group of Prostate Cancer on Multiparametric MRI using Deep Machine Learning Models

International Journal of Radiation Oncology*Biology*Physics ◽

10.1016/j.ijrobp.2020.02.484 ◽

2020 ◽

Vol 108 (2) ◽

pp. E9-E10

Author(s):

Weiwei Zong ◽

Joon Lee ◽

Milan Pantelic ◽

Ning Wen

Keyword(s):

Prostate Cancer ◽

Machine Learning ◽

Multiparametric Mri ◽

Gleason Grade ◽

Learning Models ◽

Grade Group ◽

Machine Learning Models

Download Full-text

Predicting Prostate Cancer Upgrading of Biopsy Gleason Grade Group at Radical Prostatectomy Using Machine Learning-assisted Decision-support Models

10.21203/rs.3.rs-78116/v1 ◽

2020 ◽

Author(s):

Hailang Liu ◽

Kun Tang ◽

Ejun Peng ◽

Liang Wang ◽

Ding Xia ◽

...

Keyword(s):

Prostate Cancer ◽

Machine Learning ◽

Logistic Regression ◽

Radical Prostatectomy ◽

Predictive Value ◽

Support Vector ◽

Gleason Grade ◽

Good Discrimination ◽

Grade Group ◽

Testing Dataset

Abstract Background: This study aimed to develop a machine learning (ML)-assisted model capable of accurately predicting the probability of biopsy Gleason grade group upgrading before making treatment decisions.Methods: We retrospectively collected data from prostate cancer (PCa) patients who underwent systematic biopsy and radical prostatectomy from January 2015 to December 2019 at Tongji Hospital of Tongji Medical College, Huazhong University of Science and Technology. The study cohort was divided into training and testing datasets in a 70:30 ratio for further analysis. Four ML-assisted models were developed from 16 clinical features using logistic regression (LR), logistic regression optimized by least absolute shrinkage and selection operator (Lasso) regularization (Lasso-LR), random forest (RF) and support vector machine (SVM). The area under the curve (AUC) was applied to determine the model with the highest discrimination. Calibration plots were used to investigate the extent of over- or underestimation of predicted probabilities relative to the observed probabilities in models. Results: In total, 530 PCa patients were included, with 371 patients in the training dataset and 159 patients in the testing dataset. The Lasso-LR model showed good discrimination with an AUC, accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of 0.776, 0.712, 0.679, 0.745, 0.730 and 0.695, respectively, followed by SVM (AUC 0.740, 95% confidence interval [CI]: 0.690–0.790), LR (AUC 0.725, 95% CI: 0.674–0.776) and RF (AUC 0.666, 95% CI: 0.618–0.714). Validation of the model showed that the Lasso-LR model had the best discriminative power (AUC 0.735, 95% CI: 0.656–0.813), followed by SVM (AUC 0.723, 95% CI: 0.644–0.802), LR (AUC 0.697, 95% CI: 0.615–0.778) and RF (AUC 0.607, 95% CI: 0.531–0.684) in the testing dataset. Both the Lasso-LR and SVM models were well-calibrated. Conclusion: The Lasso-LR model had good discrimination in the prediction of patients at high risk of harboring incorrect Gleason grade group assignment, and the use of this model may be greatly beneficial to urologists in treatment planning, patient selection, and the decision-making process for PCa patients.

Download Full-text