Detectability of prostate cancer in different parts of the gland with 3-Tesla multiparametric magnetic resonance imaging: correlation with whole-mount histopathology

A comprehensive comparison between mpMRI of the prostate, MR-US fusion biopsy and whole mount histopathology

OBJECTIVES

To compare multiparametric magnetic resonance imaging (mpMRI) findings, US-MR fusion prostate biopsy results and whole-mount thin-section histopathology after radical prostatectomy.

PATIENTS AND METHODS

Overall 259 patients, who had undergone mpMRI with lesions reported as PI-RADS 3-5, underwent a MR-US fusion biopsy between 2018 and 2020. Overall 186 biopsies yielded prostate cancer and 104 patients subsequently underwent endoscopic extraperitoneal radical prostatectomy. Histopathology of biopsies was compared to the final histopathology in whole mount thin sections after radical prostatectomy by means of descriptive statistics, and further, the lesions from mpMRT were compared to whole mount histology.

RESULTS

Prostate cancer was diagnosed in 186 (71.8%) of 259 patients (median age 69.2 y, range 42-82 y, median PSA 7.8 ng/ml, range 2.1-31.3 ng/ml). Of those, 95 (51,1%) underwent radical endoscopic prostatectomy, and 80 (43%) chose radiotherapy or active surveillance. In 52/95 (54,7%) with RPE additional lesions were found in the final histological whole mount sections not described at mpMRI. 22/95 (23,2%) of RPE patients had ≥ 1 additional Gleason score ≥ 7 lesions, 23 /259 (8,4%) of biopsies, respectively. The Gleason score after surgery was upgraded in 37/95 (38,9%) and downgraded in 18/95 (18,9%) patients.

CONCLUSION

If we compare all 259 performed biopsies with the final histological whole mount sections which showed additional lesions with Gleason ≥ 7 (23,2%), it can be assumed that up to 10% of clinical significant carcinomas are missed during primary assessment via mpMRI. The majority of additional findings after RP were intermediate/high risk tumors. Upgrades from low-risk to intermediate or high-risk occurred.

Deep Learning in Prostate Cancer Diagnosis Using Multiparametric Magnetic Resonance Imaging With Whole-Mount Histopathology Referenced Delineations

Background: Multiparametric magnetic resonance imaging (mpMRI) plays an important role in the diagnosis of prostate cancer (PCa) in the current clinical setting. However, the performance of mpMRI usually varies based on the experience of the radiologists at different levels; thus, the demand for MRI interpretation warrants further analysis. In this study, we developed a deep learning (DL) model to improve PCa diagnostic ability using mpMRI and whole-mount histopathology data. Methods: A total of 739 patients, including 466 with PCa and 273 without PCa, were enrolled from January 2017 to December 2019. The mpMRI (T2 weighted imaging, diffusion weighted imaging, and apparent diffusion coefficient sequences) data were randomly divided into training (n = 659) and validation datasets (n = 80). According to the whole-mount histopathology, a DL model, including independent segmentation and classification networks, was developed to extract the gland and PCa area for PCa diagnosis. The area under the curve (AUC) were used to evaluate the performance of the prostate classification networks. The proposed DL model was subsequently used in clinical practice (independent test dataset; n = 200), and the PCa detective/diagnostic performance between the DL model and different level radiologists was evaluated based on the sensitivity, specificity, precision, and accuracy. Results: The AUC of the prostate classification network was 0.871 in the validation dataset, and it reached 0.797 using the DL model in the test dataset. Furthermore, the sensitivity, specificity, precision, and accuracy of the DL model for diagnosing PCa in the test dataset were 0.710, 0.690, 0.696, and 0.700, respectively. For the junior radiologist without and with DL model assistance, these values were 0.590, 0.700, 0.663, and 0.645 versus 0.790, 0.720, 0.738, and 0.755, respectively. For the senior radiologist, the values were 0.690, 0.770, 0.750, and 0.730 vs. 0.810, 0.840, 0.835, and 0.825, respectively. The diagnosis made with DL model assistance for radiologists were significantly higher than those without assistance (P < 0.05). Conclusion: The diagnostic performance of DL model is higher than that of junior radiologists and can improve PCa diagnostic accuracy in both junior and senior radiologists.

Risk assessment of multi-factorial complications after transrectal ultrasound-guided prostate biopsy: a single institutional retrospective cohort study

BACKGROUND

Transrectal ultrasound-guided prostate biopsy (TRUSPB) is widely used to diagnose prostate cancer (PCa). The aim of this study was to evaluate the risk of multi-factorial complications (febrile genitourinary tract infection (GUTI), rectal bleeding, and urinary retention) after TRUSPB.

METHODS

N = 2053 patients were Japanese patients undergoing transrectal or transperineal TRUSPB for suspicious of PCa. To assess risk of febrile GUTI adequately, the patients were divided into four groups: low-risk patients before starting a rectal culture, low-risk patients after starting a rectal culture, high-risk patients, and patients undergoing transperineal TRUSPB. Furthermore, to identify risk of rectal bleeding and urinary retention, patients were divided into transrectal and transperineal group.

RESULTS

Febrile GUTI significantly decreased owing to risk classification. The frequency of rectal bleeding was 1.43% (transrectal: 25/1742), while it did not happen in transperineal group. The patients with rectal bleeding had a significantly lower body mass index (BMI) (P < 0.01). The frequency of urinary retention was 5.57% (transrectal: 97/1742), while it did not happen in transperineal group. The patients with urinary retention had a significantly higher prostate-specific antigen (PSA) (P = 0.01) in transrectal group.

CONCLUSIONS

Risk classification, rectal swab culture, and selected antimicrobial prophylaxis for transrectal TRUSPB were extremely effective to reduce the risk of febrile GUTI. Furthermore, lower BMI and higher PSA were novel clinical predictors for rectal bleeding and urinary retention, respectively. When urologists perform transrectal TRUSPB to their patients, they can correctly understand and explain each complication risk to their patients based on these novel risk factors.

The diagnostic accuracy of multiparametric MRI for detection and localization of prostate cancer depends on the affected region

Objectives

To determine the diagnostic accuracy of 3T multiparametric magnetic resonance imaging (mpMRI) for detecting and locating prostate cancer (PCa) on Dickinson's 27-sector map, using histopathology specimens from radical prostatectomy (RP) as the reference standard.

Patients and methods

The authors studied a continuous series of 140 patients who underwent RP over three consecutive years. Prior to RP, all patients had mpMRI for detection and localization of PCa and further assessment by biopsy. To minimize the potential of disease progression, 25 patients were excluded because the interval between mpMRI and RP exceeded 6 months, which left 115 patients eligible for analysis. The mpMRI findings were reported using the Prostate Imaging-Reporting and Data System (PI-RADS) v2, considering PI-RADS ≥ 3 to indicate PCa. The histopathology findings from RP specimens were graded using the Gleason scoring system, considering Gleason ≥ 6 to indicate PCa. The location of the tumors was mapped on Dickinson's 27-sector map for both mpMRI and histopathology and compared by rigid sector-by-sector matching.

Results

The cohort of 115 patients eligible for analysis was aged 66.5 ± 6.0 years at RP. Of the 3105 sectors analyzed, there were 412 true positives (13%), 28 false positives (1%), 68 false negatives (2%), and 2597 true negatives (84%). Across the 27 sectors of the prostate, mpMRI sensitivity ranged from 50% to 100% and specificity from 96% to 100%, while PPV ranged from 50% to 100%, and NPV from 91% to 100%. For the anterior prostate, mpMRI had a sensitivity of 80% (CI, 71%-86%), specificity of 99% (CI, 99%-100%), PPV of 91% (CI, 83%-95%), and NPV of 99% (CI, 98%-99%). For the posterior prostate, mpMRI had a sensitivity of 88% (CI, 84%-91%), specificity of 98% (CI, 97%-99%), PPV of 94% (CI, 92%-96%), and NPV of 96% (CI, 94%-97%). Overall, mpMRI had a sensitivity of 86%, specificity of 99%, PPV of 94%, and NPV of 97%.

Conclusions

The accuracy of mpMRI in detecting and locating prostate tumors depends on the affected region, but its high NPV across all sectors suggests that negative findings may not need corroboration by other techniques.

Application of hierarchical clustering to multi-parametric MR in prostate: Differentiation of tumor and normal tissue with high accuracy

PURPOSE

Hierarchical clustering (HC), an unsupervised machine learning (ML) technique, was applied to multi-parametric MR (mp-MR) for prostate cancer (PCa). The aim of this study is to demonstrate HC can diagnose PCa in a straightforward interpretable way, in contrast to deep learning (DL) techniques.

METHODS

HC was constructed using mp-MR including intravoxel incoherent motion, diffusion kurtosis imaging, and dynamic contrast-enhanced MRI from 40 tumor and normal tissues in peripheral zone (PZ) and 23 tumor and normal tissues in transition zone (TZ). HC model was optimized by assessing the combinations of several dissimilarity and linkage methods. Goodness of HC model was validated by internal methods.

RESULTS

Accuracy for differentiating tumor and normal tissue by optimal HC model was 96.3% in PZ and 97.8% in TZ, comparable to current clinical standards. Relationship between input (DWI and permeability parameters) and output (tumor and normal tissue cluster) was shown by heat maps, consistent with literature.

CONCLUSION

HC can accurately differentiate PCa and normal tissue, comparable to state-of-the-art diffusion based parameters. Contrary to DL techniques, HC is an operator-independent ML technique producing results that can be interpreted such that the results can be knowledgeably judged.

Multiparametric MRI for Prostate Cancer Characterization: Combined Use of Radiomics Model with PI-RADS and Clinical Parameters

Radiomics is an emerging field of image analysis with potential applications in patient risk stratification. This study developed and evaluated machine learning models using quantitative radiomic features extracted from multiparametric magnetic resonance imaging (mpMRI) to detect and classify prostate cancer (PCa). In total, 191 patients that underwent prostatic mpMRI and combined targeted and systematic fusion biopsy were retrospectively included. Segmentations of the whole prostate glands and index lesions were performed manually in apparent diffusion coefficient (ADC) maps and T2-weighted MRI. Radiomic features were extracted from regions corresponding to the whole prostate gland and index lesion. The best performing combination of feature setup and classifier was selected to compare its predictive ability of the radiologist’s evaluation (PI-RADS), mean ADC, prostate specific antigen density (PSAD) and digital rectal examination (DRE) using receiver operating characteristic (ROC) analysis. Models were evaluated using repeated 5-fold cross-validation and a separate independent test cohort. In the test cohort, an ensemble model combining a radiomics model, with models for PI-RADS, PSAD and DRE achieved high predictive AUCs for the differentiation of (i) malignant from benign prostatic lesions (AUC = 0.889) and of (ii) clinically significant (csPCa) from clinically insignificant PCa (cisPCa) (AUC = 0.844). Our combined model was numerically superior to PI-RADS for cancer detection (AUC = 0.779; p = 0.054) as well as for clinical significance prediction (AUC = 0.688; p = 0.209) and showed a significantly better performance compared to mADC for csPCa prediction (AUC = 0.571; p = 0.022). In our study, radiomics accurately characterizes prostatic index lesions and shows performance comparable to radiologists for PCa characterization. Quantitative image data represent a potential biomarker, which, when combined with PI-RADS, PSAD and DRE, predicts csPCa more accurately than mADC. Prognostic machine learning models could assist in csPCa detection and patient selection for MRI-guided biopsy.

Role of Multiparametric Magnetic Resonance Imaging in Predicting Pathologic Outcomes in Prostate Cancer

Multiparametric magnetic resonance imaging (mpMRI) and the introduction of standardized protocols for its interpretation have had a significant impact on the field of prostate cancer (PC). Multiple randomized controlled trials have shown that the sensitivity for detection of clinically significant PC is increased when mpMRI results are the basis for indication of a prostate biopsy. The added value with regards to sensitivity has been strongest for patients with persistent suspicion for PC after a prior negative biopsy. Although enhanced sensitivity of mpMRI is convincing, studies that have compared mpMRI with prostatectomy specimens prepared by whole-mount section analysis have shown a significant number of lesions that were not detected by mpMRI. In this context, the importance of an additional systematic biopsy (SB) is still being debated. While SB in combination with targeted biopsies leads to an increased detection rate, most of the tumors detected by SB only are considered clinically insignificant. Currently, multiple risk calculation tools are being developed that include not only clinical parameters but mpMRI results in addition to clinical parameters in order to improve risk stratification for PC, such as the Partin tables. In summary, mpMRI of the prostate has become a standard procedure recommended by multiple important guidelines for the diagnostic work-up of patients with suspicion of PC.