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.

The mechanisms by which prostate cancer shifts from an indolent castration-sensitive phenotype to lethal castration-resistant prostate cancer (CRPC) are poorly understood. Identification of clinically relevant genetic alterations leading to CRPC may reveal potential vulnerabilities for cancer therapy. Here we find that CUB domain-containing protein 1 (CDCP1), a transmembrane protein that acts as a substrate for SRC family kinases (SFKs), is overexpressed in a subset of CRPC. Notably, CDCP1 cooperates with the loss of the tumor suppressor gene PTEN to promote the emergence of metastatic prostate cancer. Mechanistically, we find that androgens suppress CDCP1 expression and that androgen deprivation in combination with loss of PTEN promotes the upregulation of CDCP1 and the subsequent activation of the SRC/MAPK pathway. Moreover, we demonstrate that anti-CDCP1 immunoliposomes (anti–CDCP1 ILs) loaded with chemotherapy suppress prostate cancer growth when administered in combination with enzalutamide. Thus, our study identifies CDCP1 as a powerful driver of prostate cancer progression and uncovers different potential therapeutic strategies for the treatment of metastatic prostate tumors.

Purpose: Recently, a significant association was shown between novel growth patterns on histopathology of prostate cancer (PCa) and prostate-specific membrane antigen (PSMA) uptake on [⁶⁸Ga]PSMA-PET. It is the aim of this study to evaluate the association between these growth patterns and ADC (mm²/1000 s) values in comparison to [⁶⁸Ga]PSMA uptake on PET/MRI. Methods: We retrospectively evaluated patients who underwent [⁶⁸Ga]PSMA PET/MRI for staging or biopsy guidance, followed by radical prostatectomy at our institution between 07/2016 and 01/2020. The dominant lesion per patient was selected based on histopathology and correlated to PET/MRI in a multidisciplinary meeting, and quantified using SUV_max for PSMA uptake and ADC_mean for diffusion restriction. PCa growth pattern was classified as expansive (EXP) or infiltrative (INF) according to its properties of forming a tumoral mass or infiltrating diffusely between benign glands by two independent pathologists. Furthermore, the corresponding WHO2016 ISUP tumor grade was evaluated. The t test was used to compare means, Pearson’s test for categorical correlation, Cohen’s kappa test for interrater agreement, and ROC curve to determine the best cutoff. Results: Sixty-two patients were included (mean PSA 11.7 ± 12.5). The interrater agreement between both pathologists was almost perfect with κ = 0.81. While 25 lesions had an EXP-growth with an ADC_mean of 0.777 ± 0.109, 37 showed an INF-growth with a significantly higher ADC_mean of 1.079 ± 0.262 (p < 0.001). We also observed a significant difference regarding PSMA SUV_max for the EXP-growth (19.2 ± 10.9) versus the INF-growth (9.4 ± 6.2, p < 0.001). Within the lesions encompassing the EXP- or the INF-growth, no significant correlation between the ISUP groups and ADC_mean could be observed (p = 0.982 and p = 0.861, respectively). Conclusion: PCa with INF-growth showed significantly lower SUV_max and higher ADC_mean values compared to PCa with EXP-growth. Within the growth groups, ADC_mean values were independent from ISUP grading.

Renal cell carcinoma (RCC) represents a heterogeneous disease, encompassing an increasing number of tumor subtypes. Post-2016, the World Health Organization (WHO) classification recognized that the spectrum of papillary renal cell carcinoma is evolving and has long surpassed the dichotomic simplistic “type 1 versus type 2” classification. The differential diagnosis of pRCC includes several new provisional/emerging entities with papillary growth. Type 2 tumors have been cleared out of several confounding entities, now regarded as independent tumors with specific clinical and molecular backgrounds. In this work we describe the prevalence and characteristics of emerging papillary tumor entities in two renal tumor cohorts (one consisting of consecutive papillary tumors from a single institute, the other consisting of consultation cases from several centers). After a review of 154 consecutive pRCC cases, 58% remained type 1 pRCC, and 34% type 2 pRCC. Papillary renal neoplasm with reversed polarity (1.3%), biphasic hyalinizing psammomatous RCC (1.3%), and biphasic squamoid/alveolar RCC (4.5%) were rare. Among 281 consultation cases, 121 (43%) tumors had a dominant papillary growth (most frequently MiT family translocation RCCs, mucinous tubular and spindle cell carcinoma and clear cell papillary RCC). Our data confirm that the spectrum of RCCs with papillary growth represents a major diagnostical challenge, frequently requiring a second expert opinion. Papillary renal neoplasm with reversed polarity, biphasic hyalinizing psammomatous RCC, and biphasic squamoid/alveolar RCC are rarely sent out for a second opinion, but correct classification and knowledge of these variants will improve our understanding of the clinical behavior of renal tumors with papillary growth.

Objectives: PSMA PET/MRI showed the potential to increase the sensitivity for extraprostatic disease (EPD) assessment over mpMRI; however, the interreader variability for EPD is still high. Therefore, we aimed to assess whether quantitative PSMA and mpMRI imaging parameters could yield a more robust EPD prediction. Methods: We retrospectively evaluated PCa patients who underwent staging mpMRI and [⁶⁸Ga]PSMA-PET, followed by radical prostatectomy at our institution between 01.02.2016 and 31.07.2019. Fifty-eight cases with PET/MRI and 15 cases with PET/CT were identified. EPD was determined on histopathology and correlated with quantitative PSMA and mpMRI parameters assessed by two readers: ADC (mm²/1000 s), longest capsular contact (LCC, mm), tumor volume (cm³), PSMA-SUV_max and volume-based parameters using a fixed threshold at SUV > 4 to delineate PSMA_total (g/ml) and PSMA_vol (cm³). The t test was used to compare means, Pearson’s test for categorical correlation, and ROC curve to determine the best cutoff. Interclass correlation (ICC) was performed for interreader agreement (95% CI). Results: Seventy-three patients were included (64.5 ± 6.0 years; PSA 14.4 ± 17.1 ng/ml), and 31 had EPD (42.5%). From mpMRI, only LCC reached significance (p = 0.005), while both volume-based PET parameters PSMA_total and PSMA_vol were significantly associated with EPD (p = 0.008 and p = 0.004, respectively). On ROC analysis, LCC, PSMA_total, and PSMA_vol reached an AUC of 0.712 (p = 0.002), 0.709 (p = 0.002), and 0.718 (p = 0.002), respectively. ICC was moderate–good for LCC 0.727 (0.565–0.828) and excellent for PSMA_total and PSMA_vol with 0.944 (0.990–0.996) and 0.985 (0.976–0.991), respectively. Conclusions: Quantitative PSMA parameters have a similar potential as mpMRI LCC to predict EPD of PCa, with a significantly higher interreader agreement.