Gleason Grading Of Prostate Biopsies Research Articles

Abstract IA022: AI in bladder cancer pathology and research: What can we expect and what not?

Abstract Artificial intelligence (AI) applications are rapidly evolving, disruptive technologies that recently attracted tremendous public awareness including euphoria, bull runs at stock exchanges but also criticism and skepticism. Pathology, a crucial clinical diagnostic discipline, faces significant challenges despite its critical role and broad scope. A notable issue is the declining interest among medical graduates in pursuing further training in Surgical Pathology and Molecular Pathology. This disinterest is compounded by a trend towards part-time work and the aging demographic in Western industrial nations, which increases the demand for precise cancer diagnoses, molecular pathology and advanced oncological treatments. As the need for skilled pathologists grows, the availability of these professionals is steadily decreasing. Furthermore, political policies in heavily regulated healthcare systems (e.g., Germany) are exacerbating the shortage by limiting residency and specialist training positions despite the growing need. Adding to the reluctance among medical graduates to consider pathology as discipline is the looming “threat”” of replacing the field with Artificial Intelligence (AI) technologies. However, historical precedents suggest that necessity often spurs significant innovation. AI's role in pathology is expanding, reshaping the field with new technological advancements. AI, broadly encompassing various computational algorithms designed to perform tasks typically requiring human intelligence such as learning, reasoning, and pattern recognition, is now being applied extensively across different domains including medical diagnostics. Recent implementations in pathology include AI-driven algorithms that automate processes such as the Gleason grading of prostate biopsies, lymph node metastasis detection, detection of molecular bladder cancer subtypes or grading of kidney cancer showing high accuracy levels, speed and reproducibility in delivering results. These tools, while they do not replace pathologists, significantly reduce their workload by automating routine tasks, thus freeing up time for more complex diagnostic activities or simply dealing with the increasing workload. Despite the exciting possibilities, the use of AI in predicting molecular alterations, such as FGFR3 mutations in bladder cancer, faces limitations due to insufficient sensitivity and specificity, indicating that such technologies are not yet suitable for screening purposes. In conclusion, AI is increasingly vital in addressing the challenges of modern pathology, particularly in compensating for the shortage of pathologists. As technology advances, its integration into clinical practice offers promising solutions to enhance diagnostic accuracy and efficiency in uropathology, paving the way for more personalized and effective patient care. Citation Format: Markus Eckstein. AI in bladder cancer pathology and research: What can we expect and what not? [abstract]. In: Proceedings of the AACR Special Conference on Bladder Cancer: Transforming the Field; 2024 May 17-20; Charlotte, NC. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(10_Suppl):Abstract nr IA022.

A deep learning network for Gleason grading of prostate biopsies using EfficientNet.

The most crucial part in the diagnosis of cancer is severity grading. Gleason's score is a widely used grading system for prostate cancer. Manual examination of the microscopic images and grading them is tiresome and consumes a lot of time. Hence to automate the Gleason grading process, a novel deep learning network is proposed in this work. In this work, a deep learning network for Gleason grading of prostate cancer is proposed based on EfficientNet architecture. It applies a compound scaling method to balance the dimensions of the underlying network. Also, an additional attention branch is added to EfficientNet-B7 for precise feature weighting. To the best of our knowledge, this is the first work that integrates an additional attention branch with EfficientNet architecture for Gleason grading. The proposed models were trained using H&E-stained samples from prostate cancer Tissue Microarrays (TMAs) in the Harvard Dataverse dataset. The proposed network was able to outperform the existing methods and it achieved an Kappa score of 0.5775.

Immunotherapy to prevent progression on active surveillance study (IPASS): A phase II, randomized, double-blind, controlled trial of PROSTVAC in prostate cancer patients who are candidates for active surveillance.

249 Background: Immunotherapy could potentially prevent disease progression for early-stage prostate cancer. In this randomized Phase 2 clinical trial, we evaluated the clinical effects of PROSTVAC, a vaccinia/fowlpox viral vector-based immunotherapy that contains PSA and three T-cell costimulatory molecules, in patients with localized prostate cancer. Methods:154 patients with clinically localized, low- or favorable intermediate-risk prostate cancer active surveillance were randomized (2:1) to receive 7 doses of subcutaneous PROSTVAC or placebo (empty fowlpox vector) over 140 days. Post-intervention prostate biopsy was performed 7-14 days after the last dose. Participants were followed for 6 months post-treatment. The primary outcome was change from baseline to post-vaccination in CD4 and CD8 T cell infiltration in biopsy tumor tissue. Secondary outcomes included changes in prostate biopsy Gleason grade (Grade Group) and serum PSA. Results: There were no differences in CD4 and CD8 densities (count of cells/mm2) in post-treatment biopsy tumor tissue between groups ( p = 0.63 and p = 0.75, respectively). Compared to placebo, patients who received PROSTVAC were less likely to demonstrate upgrading at follow-up biopsy, but this difference did not attain significance (22% vs. 40%, p= 0.08). There was no difference in the change of PSA from baseline to 6 months post-treatment between arms ( p= 0.30). Conclusions: In this first-of-kind trial of immunotherapy for localized prostate cancer, PROSTVAC was well tolerated but did not elicit significant prostate tissue T-cell responses compared to placebo. The favorable post-treatment biopsy grade findings in PROSTVAC patients merit further evaluation and longer-term clinical follow-up. Clinical trial information: NCT02326805.

Evaluation of the Use of Combined Artificial Intelligence and Pathologist Assessment to Review and Grade Prostate Biopsies

Expert-level artificial intelligence (AI) algorithms for prostate biopsy grading have recently been developed. However, the potential impact of integrating such algorithms into pathologist workflows remains largely unexplored. To evaluate an expert-level AI-based assistive tool when used by pathologists for the grading of prostate biopsies. This diagnostic study used a fully crossed multiple-reader, multiple-case design to evaluate an AI-based assistive tool for prostate biopsy grading. Retrospective grading of prostate core needle biopsies from 2 independent medical laboratories in the US was performed between October 2019 and January 2020. A total of 20 general pathologists reviewed 240 prostate core needle biopsies from 240 patients. Each pathologist was randomized to 1 of 2 study cohorts. The 2 cohorts reviewed every case in the opposite modality (with AI assistance vs without AI assistance) to each other, with the modality switching after every 10 cases. After a minimum 4-week washout period for each batch, the pathologists reviewed the cases for a second time using the opposite modality. The pathologist-provided grade group for each biopsy was compared with the majority opinion of urologic pathology subspecialists. An AI-based assistive tool for Gleason grading of prostate biopsies. Agreement between pathologists and subspecialists with and without the use of an AI-based assistive tool for the grading of all prostate biopsies and Gleason grade group 1 biopsies. Biopsies from 240 patients (median age, 67 years; range, 39-91 years) with a median prostate-specific antigen level of 6.5 ng/mL (range, 0.6-97.0 ng/mL) were included in the analyses. Artificial intelligence-assisted review by pathologists was associated with a 5.6% increase (95% CI, 3.2%-7.9%; P < .001) in agreement with subspecialists (from 69.7% for unassisted reviews to 75.3% for assisted reviews) across all biopsies and a 6.2% increase (95% CI, 2.7%-9.8%; P = .001) in agreement with subspecialists (from 72.3% for unassisted reviews to 78.5% for assisted reviews) for grade group 1 biopsies. A secondary analysis indicated that AI assistance was also associated with improvements in tumor detection, mean review time, mean self-reported confidence, and interpathologist agreement. In this study, the use of an AI-based assistive tool for the review of prostate biopsies was associated with improvements in the quality, efficiency, and consistency of cancer detection and grading.

Artificial intelligence assistance significantly improves Gleason grading of prostate biopsies by pathologists

The Gleason score is the most important prognostic marker for prostate cancer patients, but it suffers from significant observer variability. Artificial intelligence (AI) systems based on deep learning can achieve pathologist-level performance at Gleason grading. However, the performance of such systems can degrade in the presence of artifacts, foreign tissue, or other anomalies. Pathologists integrating their expertise with feedback from an AI system could result in a synergy that outperforms both the individual pathologist and the system. Despite the hype around AI assistance, existing literature on this topic within the pathology domain is limited. We investigated the value of AI assistance for grading prostate biopsies. A panel of 14 observers graded 160 biopsies with and without AI assistance. Using AI, the agreement of the panel with an expert reference standard increased significantly (quadratically weighted Cohen’s kappa, 0.799 vs. 0.872; p = 0.019). On an external validation set of 87 cases, the panel showed a significant increase in agreement with a panel of international experts in prostate pathology (quadratically weighted Cohen’s kappa, 0.733 vs. 0.786; p = 0.003). In both experiments, on a group-level, AI-assisted pathologists outperformed the unassisted pathologists and the standalone AI system. Our results show the potential of AI systems for Gleason grading, but more importantly, show the benefits of pathologist-AI synergy.

Deep-learning approaches for Gleason grading of prostate biopsies.

Deep-learning approaches for Gleason grading of prostate biopsies.

Predictors of Gleason Score Upgrading after Radical Prostatectomy in Low-Risk Prostate Cancer

Purpose: The Gleason score is an important predictor of outcome that is used in conjunction with clinical stage and prostate-specific antigen to guide clinical decision making. The prostate biopsy Gleason grade frequently differs from the radical prostatectomy grade. The aim of this study was to determine the risk factors of Gleason upgrading in patients with low-risk prostate cancer after radical prostatectomy. Materials and Methods: We retrospectively analyzed the medical records of 146 patients who underwent radical prostatectomy between 1998 and 2008 in two hospitals of Jeonbuk province in Korea. Pathological Gleason score upgrading was defined as an increase in the Gleason score from ≤6 to ≥7 between the biopsy and radical prostatectomy specimen. Pretreatment clinical and pathological parameters were used to identify predictors of pathological upgrading. Results: Of the total 146 patients, 51 (34.9%) were upgraded postoperatively. Small prostate volume (p=0.008), abnormality on the digital rectal examination, and positive surgical margin (p=0.001) were significantly and positively associated with upgrading after radical prostatectomy. A total of 17 of 65 patients with low-risk prostate cancer (26.2%) were upgraded postoperatively. Small prostate volume (<30 ml) was significantly (p= 0.026) and positively associated with upgrading after radical prostatectomy in patients with low-risk prostate cancer. Conclusions: Overall, 26% of patients with low-risk disease were upgraded postoperatively. Small prostate volume was associated with an increased risk for pathological upgrading after radical prostatectomy. These conclusions should be kept in mind when making treatment decisions for men with low-risk prostate cancer. (Korean J Urol 2009;50:1182-1187) 󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏󰠏

Comparing the Gleason Prostate Biopsy and Gleason Prostatectomy Grading System: The Lahey Clinic Medical Center Experience and an International Meta-Analysis

BackgroundThe accuracy of the prostate biopsy Gleason grade to predict the prostatectomy Gleason grade varies tremendously in the literature. ObjectivesDetermine the accuracy and distribution of the prostate biopsy Gleason grade and prostatectomy Gleason grade at LCMC (Lahey Clinic Medical Center) and worldwide. Design, Setting, and ParticipantsParticipants included 2890 patients who had not received preoperative hormones, and for whom preoperative and postoperative Gleason sums were available. Participants underwent radical prostatectomy at LCMC, an academic referral center, from 1982–2007. Studies for the meta-analysis were selected from Medline: 1994–2007. Search criteria included keywords “Gleason,” “biopsy,” and “prostatectomy,” ≥200 patients, and whether the biopsy and prostatectomy Gleason scores categorized into the predefined Gleason grades. The meta-analysis included 15 studies and the LCMC database for 14,839 total patients. MeasurementsGleason scores 2–6, 7, and 8–10 were converted to low, moderate, and high grade, respectively. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value were calculated. The kappa statistic and chi-square were used to compare biopsy and prostatectomy grades. Results and LimitationsThe percentage of patients in whom the prostatectomy grade was accurately predicted, upgraded, and downgraded was 58%, 36%, and 5% at LCMC and 63%, 30%, and 7% in the meta-analysis, respectively. The PPV for low-, moderate-, and high-grade cancer was 54%, 70%, and 60% for LCMC and 62%, 70%, and 50% for the meta-analysis, respectively.The sensitivity decreased with increasing Gleason grade (low, moderate, and high) for LCMC (91%, 38%, 28%) and the meta-analysis (90%, 40%, 33%), respectively. The distribution of low-, moderate-, and high-grade cancer on biopsy (69%, 25%, and 6%) and prostatectomy specimen (47%, 44%, and 9%) demonstrated only “fair” agreement (kappa, 0.37). ConclusionsPatients and practitioners need to be cognizant of significant upgrading for low-grade disease and the downgrading for high-grade disease.

Upgrading and Downgrading of Prostate Needle Biopsy Specimens: Risk Factors and Clinical Implications

The prostate biopsy Gleason grade frequently differs from the radical prostatectomy (RP) grade. Given the critical role that needle biopsy plays in treatment decisions, we sought to determine the risk factors for upgrading and downgrading the prostate biopsy specimen. We determined the significant predictors of upgrading (worse RP grade than biopsy grade) and downgrading (better RP grade than biopsy grade) among 1113 men treated with RP from 1996 to 2005 within the Shared Equal Access Regional Cancer Hospital (SEARCH) database who had undergone at least sextant biopsy. The Gleason sum was examined as a categorical variable of 2 to 6, 3+4, and 4+3 or greater. Overall, the disease of 299 men (27%) was upgraded and 123 (11%) was downgraded, and 691 men (62%) had identical biopsy and pathologic Gleason sum groups. Upgrading was associated with adverse pathologic features (P < or = 0.001) and the risk of biochemical progression (P = 0.001). Downgrading was associated with more favorable pathologic features (P < or = 0.01) and a decreased risk of progression (P = 0.04). On multivariate analysis, greater prostate-specific antigen levels (P < 0.001), more biopsy cores with cancer (P = 0.001), and obesity (P = 0.003) were all significantly and positively associated with upgrading. In contrast, biopsy Gleason sum 3+4 (P = 0.001) and obtaining eight or more biopsy cores (P = 0.01) were associated with a lower likelihood of upgrading. Men whose disease was upgraded were at a greater risk of adverse pathologic features and biochemical progression. Men with "high-risk" cancer (greater prostate-specific antigen levels, more positive cores, and obese) were more likely to have their disease category upgraded, and obtaining more biopsy cores reduced the likelihood of upgrading.

Prediction of Indolent Prostate Cancer: Validation and Updating of a Prognostic Nomogram

Screening with serum prostate specific antigen testing leads to the detection of many prostate cancers early in their natural history. Statistical models have been proposed to predict indolent cancer. We validated and updated model predictions for a screening setting. We selected 247 patients with clinical stage T1C or T2A from the European Randomized Study on Screening for Prostate Cancer who were treated with radical prostatectomy. We validated a nomogram that had previously been developed in a clinical setting. Predictive characteristics were serum prostate specific antigen, ultrasound prostate volume, clinical stage, prostate biopsy Gleason grade, and total length of cancer and noncancer tissue in biopsy cores. Indolent cancer was defined as pathologically organ confined cancer 0.5 cc or less in volume without poorly differentiated elements. Logistic regression was used to update the previous model and examine the contribution of other potential predictors. Overall 121 of 247 patients (49%) had indolent cancer, while the average predicted probability was around 20% (p <0.001). Effects of individual variables were similar to those found before and discriminative ability was adequate (AUC 0.76). An updated model was constructed, which merely recalibrated the nomogram and did not apply additional predictors. Prostate cancers identified in a screening setting have a substantially higher likelihood of being indolent than those predicted by a previously proposed nomogram. However, an updated model can support patients and clinicians when the various treatment options for prostate cancer are considered.

Risk assessment for biochemical recurrence prior to radical prostatectomy: Significant enhancement contributed by human glandular kallikrein 2 (hK2) and free prostate specific antigen (PSA) in men with moderate PSA‐elevation in serum

Most models to predict biochemical recurrence (BCR) of prostate cancer use pretreatment serum prostate-specific antigen (PSA), clinical stage and prostate biopsy Gleason grade. We investigated whether human glandular kallikrein 2 (hK2) and free prostate-specific antigen (fPSA) measured in pretreatment serum enhance prediction. We retrospectively measured total PSA (tPSA), fPSA and hK2 in preoperative serum samples from 461 men with localized prostate cancer treated with radical prostatectomy between 1999 and 2001. We developed a regression model to predict BCR using preoperative tPSA, clinical stage and biopsy Gleason grade. We then compared the predictive accuracy of this "base" model with a model with fPSA and hK2 as additional predictors. BCR was observed in 90 patients (20%), including 48 patients with a pretreatment tPSA < or = 14 ng/ml (13%), and 28 patients (10%) with a pretreatment tPSA < or = 10 ng/ml. Overall, the predictive accuracy of the base model (bootstrap-corrected concordance index of 0.813) was not improved after the addition of fPSA or hK2 (0.818). However, for men with moderate tPSA-elevation (tPSA < or = 10 ng/ml), addition of fPSA and hK2 data increased predictive accuracy (from a base model concordance index of 0.756-0.815, p = 0.005). The improvement in accuracy was not sensitive to the threshold for "moderately elevated" PSA. For patients with a moderate tPSA-elevation (tPSA < or = 10 ng/ml), which closely corresponds to concurrent disease demographics, BCR-prediction was enhanced when fPSA and hK2 were added to the conventional model. Measurements of fPSA and hK2 improve on our ability to counsel patients prior to treatment as to their risk of BCR.

Counseling Men With Prostate Cancer: A Nomogram for Predicting the Presence of Small, Moderately Differentiated, Confined Tumors

Men diagnosed with clinically localized prostate cancer have a number of treatment options available, including watchful waiting, radical prostatectomy and radiation therapy. With the widespread use of serum prostate specific antigen (PSA) testing, prostate cancers are being diagnosed earlier in their natural history, with many tumors being small and of little health risk to the patient, at least in the short term. To better counsel men diagnosed with prostate cancer, we developed a statistical model that accurately predicts the presence of small moderately differentiated, confined cancer based on clinical variables (serum PSA, clinical stage, prostate biopsy Gleason grade and ultrasound volume) and variables derived from the analysis of systematic biopsies. The analysis included 409 patients diagnosed by systematic needle biopsy with clinical stages T1c or T2a N0 or NX and M0 or MX prostate cancer who were treated solely with radical prostatectomy at 1 of 2 institutions. Additional biopsy features included number and percentage of biopsy cores involved with cancer and high grade cancer, in addition to total length of biopsy cores involved. Indolent cancer was defined as pathologically organ confined cancer 0.5 cc or less in volume and without poorly differentiated elements. Logistic regression was used to construct several prediction models and the resulting nomograms. Overall 80 (20%) of the patients had indolent cancer. The nomogram predicted the presence of an indolent cancer with discrimination (area under the receiver operating characteristics curves) for various models ranging from 0.64 to 0.79. Calibration of the models appeared good. Nomograms incorporating pretreatment variables (clinical stage, Gleason grade, PSA and the amount of cancer in a systematic biopsy specimen) can predict the probability that a man with prostate cancer has an indolent tumor. These nomograms have good discriminatory ability and calibration, and may benefit the patient and clinician when the various treatment options for prostate cancer are being considered.

Patterns of prostate cancer biopsy grading: trends and clinical implications.

The predictive value of Gleason grading from prostate needle biopsy was examined and the patterns of grade discordance with surgical specimens are discussed in terms of their clinical implications. Gleason scores from biopsy and matched radical prostatectomy specimens were compared in 428 consecutive patients. Patterns of concordance were examined with respect to numerical agreement as well as to whether differences result in a change in group assignment with respect to Gleason score group 2-4, 5-6, 7, and 8-10. The coefficient of agreement, kappa, and accuracy were used to measure predictive value. An exact Gleason score match was present in 41% of the cases, while 48% were undergraded and 17% overgraded. With respect to group assignment, 51% remained unchanged while 35% were undergraded and 14% overgraded. Kappa analysis yielded a value of 0.26, which represents a poor agreement beyond chance. A Gleason score of 5-6, 7, or 8 was reproduced in 52%, 53%, and 58% of cases, respectively. A systematic bias toward progressive undergrading of more well-differentiated cancers and overgrading of more poorly differentiated cancers on biopsy is suggested by the data. A pooled analysis with nine additional series (n = 2,687) confirms this conclusion. No correlation was found between the amount of tumor in the biopsy specimen and grade discordance. Biases in pathologic interpretation and sampling effects are suggested as responsible for grade discordance. Predictable differences exist between the histologic grade in prostate needle biopsies and the surgical specimen. Clinical staging of organ-confined prostate cancer should include the likelihood of histologic upgrading or downgrading when used to stratify patients for clinical trials, in comparing results among therapies based on biopsy grading and when recommending a radical therapy. Developing a methodology which reduces both sampling effects and pathologic interpretation bias would likely result in significantly improved accuracy of Gleason grading of prostate biopsies. Int. J. Cancer (Radiat. Oncol. Invest.) 90, 305-311 (2000).