Automated prostate gland segmentation in challenging clinical cases: comparison of three artificial intelligence methods.

OBJECTIVE: Automated methods for prostate segmentation on MRI are typically developed under ideal scanning and anatomical conditions. This study evaluates three different prostate segmentation AI algorithms in a challenging population of patients with prior treatments, variable anatomic characteristics, complex clinical history, or atypical MRI acquisition parameters. MATERIALS AND METHODS: A single institution retrospective database was queried for the following conditions at prostate MRI: prior prostate-specific oncologic treatment, transurethral resection of the prostate (TURP), abdominal perineal resection (APR), hip prosthesis (HP), diversity of prostate volumes (large ≥ 150 cc, small ≤ 25 cc), whole gland tumor burden, magnet strength, noted poor quality, and various scanners (outside/vendors). Final inclusion criteria required availability of axial T2-weighted (T2W) sequence and corresponding prostate organ segmentation from an expert radiologist. Three previously developed algorithms were evaluated: (1) deep learning (DL)-based model, (2) commercially available shape-based model, and (3) federated DL-based model. Dice Similarity Coefficient (DSC) was calculated compared to expert. DSC by model and scan factors were evaluated with Wilcox signed-rank test and linear mixed effects (LMER) model. RESULTS: 683 scans (651 patients) met inclusion criteria (mean prostate volume 60.1 cc [9.05-329 cc]). Overall DSC scores for models 1, 2, and 3 were 0.916 (0.707-0.971), 0.873 (0-0.997), and 0.894 (0.025-0.961), respectively, with DL-based models demonstrating significantly higher performance (p < 0.01). In sub-group analysis by factors, Model 1 outperformed Model 2 (all p < 0.05) and Model 3 (all p < 0.001). Performance of all models was negatively impacted by prostate volume and poor signal quality (p < 0.01). Shape-based factors influenced DL models (p < 0.001) while signal factors influenced all (p < 0.001). CONCLUSION: Factors affecting anatomical and signal conditions of the prostate gland can adversely impact both DL and non-deep learning-based segmentation models.

ISG15/USP18/STAT2 is a molecular hub regulating IFN I-mediated control of Dengue and Zika virus replication.

The establishment of a virus infection is the result of the pathogen's ability to replicate in a hostile environment generated by the host's immune system. Here, we found that ISG15 restricts Dengue and Zika viruses' replication through the stabilization of its binding partner USP18. ISG15 expression was necessary to control DV replication driven by both autocrine and paracrine type one interferon (IFN-I) signaling. Moreover, USP18 competes with NS5-mediated STAT2 degradation, a major mechanism for establishment of flavivirus infection. Strikingly, reconstitution of USP18 in ISG15-deficient cells was sufficient to restore the STAT2's stability and restrict virus growth, suggesting that the IFNAR-mediated ISG15 activity is also antiviral. Our results add a novel layer of complexity in the virus/host interaction interface and suggest that NS5 has a narrow window of opportunity to degrade STAT2, therefore suppressing host's IFN-I mediated response and promoting virus replication.

Laser enucleation of the prostate (LEP) vs. simple prostatectomy: an analysis of hospital charges and comparison of practice trends.

PURPOSE: Laser enucleation of the prostate (LEP) and simple prostatectomy (SP) are surgical treatment options for large gland Benign Prostatic Hyperplasia. While multiple studies compare clinical outcomes of these procedures, there are limited data available comparing hospital charges in the United States. Here, we present current practice trends and a hospital charge analysis on a national level using an annual insurance claims data repository. METHODS: The Healthcare Cost and Utilization Project National Inpatient Sample and Nationwide Ambulatory Surgery Sample databases for 2018 were queried. CPT and ICD-10PCS codes identified patients undergoing LEP or SP, who were then compared for practice setting, total hospital charges, and payor. Laser type for LEP and surgical approach for SP could not be differentiated. RESULTS: The median hospital charge of 5782 LEPs and 973 SPs is $26,689 and $51,250 (p < 0.001), respectively. LEP independently predicts a decreased hospital charge of $16,464 (p < 0.001) per case. Medicare is the primary payor for both procedures. More LEP procedures are completed in the outpatient setting (87.8%) vs. SPs (5.7%, p < 0.001). Median length of stay is longer for SP (LEP: 0, IQR: 0; SP: 3, IQR: 2-4; p < 0.001). In the Western region, LEP is least commonly performed (184, p < 0.001), most expensive ($43,960; p < 0.001), and has longer length of stay (2, p < 0.001). CONCLUSIONS: LEP should be considered a cost-effective alternative to SP. Regions of the U.S. that perform more LEPs have shorter length of stay and lower hospital charges associated with the procedure.

The association of sexual orientation with prostate, breast, and cervical cancer screening and diagnosis.

PURPOSE: Data on heterogeneity in cancer screening and diagnosis rates among lesbians/gays and bisexuals (LGBs) is lacking. Recent studies showed that LGBs have decreased healthcare utilization compared to heterosexual counterparts. Few studies have examined how sexual orientation impacts cancer screening and prevalence. We, therefore, investigated the association between sexual orientation and prevalent sex-specific cancer including prostate (PCa), breast (BC), and cervical (CC) cancer. METHODS: This was a cross-sectional survey-based US study, including men and women aged 18 + from the Health Information National Trends Survey (HINTS) database between 2017 and 2019. The primary endpoint was individual-reported prostate, breast, and cervical cancer screening and prevalence rates among heterosexual and LGB men and women. Multivariable logistic regression analyses assessed association of various covariates with undergoing screening and diagnosis of these cancers. RESULTS: Overall, 4,441 and 6,333 heterosexual men and women, respectively, were compared to 225 and 213 LGB men and women, respectively. LGBs were younger and less likely to be screened for PCa, BC, and CC than heterosexuals. A higher proportion of heterosexual women than lesbian and bisexual women were screened for CC with pap smears (95.36% vs. 90.48% and 86.11%, p ≤ 0.001) and BC with mammograms (80.74% vs. 63.81% and 45.37%, p ≤ 0.001). Similarly, a higher proportion of heterosexual men than gay and bisexual men were screened for PCa with PSA blood tests (41.27% vs. 30.53% and 27.58%, p ≤ 0.001). CONCLUSION: There were more heterosexuals than LGBs screened for CC, BC, and PCa. However, no association between sexual orientation and cancer diagnosis was found. Healthcare professionals should be encouraged to improve cancer screening among LGBs.

Cross-modal attention for multi-modal image registration.

In the past few years, convolutional neural networks (CNNs) have been proven powerful in extracting image features crucial for medical image registration. However, challenging applications and recent advances in computer vision suggest that CNNs are limited in their ability to understand the spatial correspondence between features, which is at the core of image registration. The issue is further exaggerated when it comes to multi-modal image registration, where the appearances of input images can differ significantly. This paper presents a novel cross-modal attention mechanism for correlating features extracted from the multi-modal input images and mapping such correlation to image registration transformation. To efficiently train the developed network, a contrastive learning-based pre-training method is also proposed to aid the network in extracting high-level features across the input modalities for the following cross-modal attention learning. We validated the proposed method on transrectal ultrasound (TRUS) to magnetic resonance (MR) registration, a clinically important procedure that benefits prostate cancer biopsy. Our experimental results demonstrate that for MR-TRUS registration, a deep neural network embedded with the cross-modal attention block outperforms other advanced CNN-based networks with ten times its size. We also incorporated visualization techniques to improve the interpretability of our network, which helps bring insights into the deep learning based image registration methods. The source code of our work is available at https://github.com/DIAL-RPI/Attention-Reg.

Artificial Intelligence in Kidney Cancer.

Artificial intelligence is rapidly expanding into nearly all facets of life, particularly within the field of medicine. The diagnosis, characterization, management, and treatment of kidney cancer is ripe with areas for improvement that may be met with the promises of artificial intelligence. Here, we explore the impact of current research work in artificial intelligence for clinicians caring for patients with renal cancer, with a focus on the perspectives of radiologists, pathologists, and urologists. Promising preliminary results indicate that artificial intelligence may assist in the diagnosis and risk stratification of newly discovered renal masses and help guide the clinical treatment of patients with kidney cancer. However, much of the work in this field is still in its early stages, limited in its broader applicability, and hampered by small datasets, the varied appearance and presentation of kidney cancers, and the intrinsic limitations of the rigidly structured tasks artificial intelligence algorithms are trained to complete. Nonetheless, the continued exploration of artificial intelligence holds promise toward improving the clinical care of patients with kidney cancer.

Polar transform network for prostate ultrasound segmentation with uncertainty estimation.

Automatic and accurate prostate ultrasound segmentation is a long-standing and challenging problem due to the severe noise and ambiguous/missing prostate boundaries. In this work, we propose a novel polar transform network (PTN) to handle this problem from a fundamentally new perspective, where the prostate is represented and segmented in the polar coordinate space rather than the original image grid space. This new representation gives a prostate volume, especially the most challenging apex and base sub-areas, much denser samples than the background and thus facilitate the learning of discriminative features for accurate prostate segmentation. Moreover, in the polar representation, the prostate surface can be efficiently parameterized using a 2D surface radius map with respect to a centroid coordinate, which allows the proposed PTN to obtain superior accuracy compared with its counterparts using convolutional neural networks while having significantly fewer (18%∼41%) trainable parameters. We also equip our PTN with a novel strategy of centroid perturbed test-time augmentation (CPTTA), which is designed to further improve the segmentation accuracy and quantitatively assess the model uncertainty at the same time. The uncertainty estimation function provides valuable feedback to clinicians when manual modifications or approvals are required for the segmentation, substantially improving the clinical significance of our work. We conduct a three-fold cross validation on a clinical dataset consisting of 315 transrectal ultrasound (TRUS) images to comprehensively evaluate the performance of the proposed method. The experimental results show that our proposed PTN with CPTTA outperforms the state-of-the-art methods with statistical significance on most of the metrics while exhibiting a much smaller model size. Source code of the proposed PTN is released at https://github.com/DIAL-RPI/PTN.

Shadow-Consistent Semi-Supervised Learning for Prostate Ultrasound Segmentation.

Prostate segmentation in transrectal ultrasound (TRUS) image is an essential prerequisite for many prostate-related clinical procedures, which, however, is also a long-standing problem due to the challenges caused by the low image quality and shadow artifacts. In this paper, we propose a Shadow-consistent Semi-supervised Learning (SCO-SSL) method with two novel mechanisms, namely shadow augmentation (Shadow-AUG) and shadow dropout (Shadow-DROP), to tackle this challenging problem. Specifically, Shadow-AUG enriches training samples by adding simulated shadow artifacts to the images to make the network robust to the shadow patterns. Shadow-DROP enforces the segmentation network to infer the prostate boundary using the neighboring shadow-free pixels. Extensive experiments are conducted on two large clinical datasets (a public dataset containing 1,761 TRUS volumes and an in-house dataset containing 662 TRUS volumes). In the fully-supervised setting, a vanilla U-Net equipped with our Shadow-AUG&Shadow-DROP outperforms the state-of-the-arts with statistical significance. In the semi-supervised setting, even with only 20% labeled training data, our SCO-SSL method still achieves highly competitive performance, suggesting great clinical value in relieving the labor of data annotation. Source code is released at https://github.com/DIAL-RPI/SCO-SSL.

A Cascaded Deep Learning-Based Artificial Intelligence Algorithm for Automated Lesion Detection and Classification on Biparametric Prostate Magnetic Resonance Imaging.

RATIONALE AND OBJECTIVES: Prostate MRI improves detection of clinically significant prostate cancer; however, its diagnostic performance has wide variation. Artificial intelligence (AI) has the potential to assist radiologists in the detection and classification of prostatic lesions. Herein, we aimed to develop and test a cascaded deep learning detection and classification system trained on biparametric prostate MRI using PI-RADS for assisting radiologists during prostate MRI read out. MATERIALS AND METHODS: T2-weighted, diffusion-weighted (ADC maps, high b value DWI) MRI scans obtained at 3 Tesla from two institutions (n = 1043 in-house and n = 347 Prostate-X, respectively) acquired between 2015 to 2019 were used for model training, validation, testing. All scans were retrospectively reevaluated by one radiologist. Suspicious lesions were contoured and assigned a PI-RADS category. A 3D U-Net-based deep neural network was used to train an algorithm for automated detection and segmentation of prostate MRI lesions. Two 3D residual neural network were used for a 4-class classification task to predict PI-RADS categories 2 to 5 and BPH. Training and validation used 89% (n = 1290 scans) of the data using 5 fold cross-validation, the remaining 11% (n = 150 scans) were used for independent testing. Algorithm performance at lesion level was assessed using sensitivities, positive predictive values (PPV), false discovery rates (FDR), classification accuracy, Dice similarity coefficient (DSC). Additional analysis was conducted to compare AI algorithm's lesion detection performance with targeted biopsy results. RESULTS: Median age was 66 years (IQR = 60-71), PSA 6.7 ng/ml (IQR = 4.7-9.9) from in-house cohort. In the independent test set, algorithm correctly detected 111 of 198 lesions leading to 56.1% (49.3%-62.6%) sensitivity. PPV was 62.7% (95% CI 54.7%-70.7%) with FDR of 37.3% (95% CI 29.3%-45.3%). Of 79 true positive lesions, 82.3% were tumor positive at targeted biopsy, whereas of 57 false negative lesions, 50.9% were benign at targeted biopsy. Median DSC for lesion segmentation was 0.359. Overall PI-RADS classification accuracy was 30.8% (95% CI 24.6%-37.8%). CONCLUSION: Our cascaded U-Net, residual network architecture can detect, classify cancer suspicious lesions at prostate MRI with good detection, reasonable classification performance metrics.

Is There a Role for Prone Computed Tomography Before Percutaneous Nephrolithotomy?

Background: The majority of percutaneous nephrolithotomies (PCNLs) are performed prone, whereas most preoperative CT scans are done supine. The purpose of this pilot study is to determine if there is utility of prone CT scans in preoperative planning for prone PCNL by identifying patient populations at risk for organ injury and tract length-related complications. Materials and Methods: To represent typical preoperative planning using CT, two-dimensional (2D)-axial-prone/supine percutaneous tract measurements were performed by minimizing the distance from the target calix to the posterior-lateral skin in a single axial plane. The minimum distance and organ interception rates for the 2D-axial planning scans were recorded. Results: A total of 60 CT colonography and 13 CT urography patients were included in analysis. There were 42 women and 31 men with unspecified pathology reports ranging in age from 27 to 86 years and in body mass index (BMI) from 17.1 to 49. Multiple logistic regression identified female gender and low BMI as predictors of organ interception on the left. On multiple linear regression comparing the difference in axial prone/supine lengths; BMI, gender, and age were not significant independent predictors of changes in tract length in any pole when prone vs supine. However, shorter supine tracts tended to lengthen when prone, and longer supine tracts tended to shorten. Conclusions: This pilot study has demonstrated that patients with long and short estimates of tract length in the supine position may have shorter and longer tracts, respectively, with repositioning to prone. Thus, prone CT may have benefit when anticipating exceptionally long (>15 cm) tract lengths. Prone scans also revealed more potential organ interceptions, particularly for low BMI and women in the left upper pole. In patients for whom prone CT demonstrates an organ interception, the urologist should consider an alternate target calix or ultrasound-guided percutaneous access to identify the most appropriate needle trajectory.

Prostate-specific Antigen Testing in Men with Disabilities: A Cross-sectional Analysis of the Health Information National Trends Survey.

BACKGROUND: Patients with disabilities represent a unique minority population. The incidence of prostate-specific antigen (PSA) testing among this population is unknown. OBJECTIVE: To compare PSA testing rates and associated predictors among men with and without reported disabilities in the USA. DESIGN, SETTING, AND PARTICIPANTS: A cross-sectional study of the Health Information National Trends Survey (HINTS) for the years 2012, 2013, 2017 and 2019 was conducted in men with reported disabilities. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Baseline demographics of the entire cohort were stratified based on their reported disabilities (none, disabled, deaf, and blind). Each disability was compared separately and in combination with the cohort without disabilities. Multivariable logistic regression models determined clinically significant predictors of PSA testing in men with disabilities compared with those without. RESULTS AND LIMITATIONS: Overall, 782 (15%) men with disabilities were compared with 4569 (85%) men without disabilities. The former cohort was older with a median (interquartile range) age of 65 (56-75) versus 57 (43-67) yr (p < 0.001). On multivariable analysis, men with any disability were less likely to undergo PSA testing (odds ratio 0.77, 95% confidence interval 0.62-0.96, p = 0.018). Variables associated with increased PSA testing included age, having a health care provider, health insurance, and living with a partner. CONCLUSIONS: Inequalities in PSA testing exist among men with disabilities in the USA, especially among the deaf and blind, being less likely to undergo PSA testing. Further research is required to identify and deal with any obstacles in the implementation of equal PSA testing in this unique population. PATIENT SUMMARY: In the USA, men with reported disabilities are less likely to undergo PSA testing than patients without reported disabilities.

Comprehensive genomic profiling of histologic subtypes of urethral carcinomas.

BACKGROUND: Carcinoma of the urethra (UrthCa) is an uncommon Genitourinary (GU) malignancy that can progress to advanced metastatic disease. METHODS: One hundred twenty-seven metastatic UrthCa underwent hybrid capture-based comprehensive genomic profiling to evaluate all classes of genomic alterations (GA). Tumor mutational burden was determined on up to 1.1 Mbp of sequenced DNA, and microsatellite instability was determined on 114 loci. PD-L1 expression was determined by IHC (Dako 22C3). RESULTS: Forty-nine (39%) urothelial (UrthUC), 31 (24%) squamous (UrthSCC), 24 (19%) adenocarcinomas NOS (UrthAC), and 12 (9%) clear cell (UrthCC) were evaluated. UrthUC and UrthSCC are more common in men; UrthAC and UrthCC are more common in women. Ages were similar in all 4 groups. GA in PIK3CA were the most frequent potentially targetable GA; mTOR pathway GA in PTEN were also identified. GA in other potentially targetable genes were also identified including ERBB2 (6% in UrthUC, 3% in UrthSCC, and 12% in UrthAC), FGFR1-3 (3% in UrthSCC), BRAF (3% in UrthAC), PTCH1 (8% in UrthCC), and MET (8% in UrthCC). Possibly reflecting their higher GA/tumor status, potential for immunotherapy benefit associated with higher tumor mutational burden and PD-L1 staining levels were seen in UrthUC and UrthSCC compared to UrthAC and UrthCC. Microsatellite instability high status was absent throughout. CONCLUSIONS: Comprehensive genomic profiling reveals GA that may be predictive of both targeted and immunotherapy benefit in patients with advanced UrthCa and that could potentially be used in future adjuvant, neoadjuvant, and metastatic disease trials.

Harnessing clinical annotations to improve deep learning performance in prostate segmentation.

PURPOSE: Developing large-scale datasets with research-quality annotations is challenging due to the high cost of refining clinically generated markup into high precision annotations. We evaluated the direct use of a large dataset with only clinically generated annotations in development of high-performance segmentation models for small research-quality challenge datasets. MATERIALS AND METHODS: We used a large retrospective dataset from our institution comprised of 1,620 clinically generated segmentations, and two challenge datasets (PROMISE12: 50 patients, ProstateX-2: 99 patients). We trained a 3D U-Net convolutional neural network (CNN) segmentation model using our entire dataset, and used that model as a template to train models on the challenge datasets. We also trained versions of the template model using ablated proportions of our dataset, and evaluated the relative benefit of those templates for the final models. Finally, we trained a version of the template model using an out-of-domain brain cancer dataset, and evaluated the relevant benefit of that template for the final models. We used five-fold cross-validation (CV) for all training and evaluation across our entire dataset. RESULTS: Our model achieves state-of-the-art performance on our large dataset (mean overall Dice 0.916, average Hausdorff distance 0.135 across CV folds). Using this model as a pre-trained template for refining on two external datasets significantly enhanced performance (30% and 49% enhancement in Dice scores respectively). Mean overall Dice and mean average Hausdorff distance were 0.912 and 0.15 for the ProstateX-2 dataset, and 0.852 and 0.581 for the PROMISE12 dataset. Using even small quantities of data to train the template enhanced performance, with significant improvements using 5% or more of the data. CONCLUSION: We trained a state-of-the-art model using unrefined clinical prostate annotations and found that its use as a template model significantly improved performance in other prostate segmentation tasks, even when trained with only 5% of the original dataset.

Rapidly Progressing Urothelial Carcinoma Due to a Rare TP53 (p.Arg110Pro) Mutation: A Case Report and Review of the Literature.

We present a case of a 69-year-old male patient diagnosed with high grade (T1 HG) urothelial carcinoma of the bladder who progressed rapidly towards muscle invasive disease and eventually death despite neoadjuvant chemotherapy and radical cystectomy. We postulate that this may be due to a deleterious underlying somatic gene mutation. Molecular pathologic data obtained on the initial, non-muscle invasive tumor and the final cystectomy specimen, revealed the same TP53 mutation (p.Arg110Pro) in both specimens with a variant allele frequency of 44%. The tumor was tested for 50 common gene mutations in urothelial carcinoma and no other identifiable DNA repair mutations were found, suggesting that this specific TP53 aberration, one that has never been reported in the bladder cancer literature, could be particularly deleterious. Knowing that bladder cancer cell lines that lack TP53 are more resistant to cisplatin and because the tumor lacked any other DNA mutation, this patient may have been a candidate for upfront surgery without neoadjuvant chemotherapy. In addition to histological analysis of the tumor, early molecular and cytogenetic characterization of resected tissue is essential in predicting progression and eventual prognosis of the disease based on identifiable gene mutations. Further comparative prospective studies are required to clarify the importance of molecular heterogeneity and subtyping in bladder cancer.

Prognostic Features of Biochemical Recurrence of Prostate Cancer Following Radical Prostatectomy Based on Multiparametric MRI and Immunohistochemistry Analysis of MRI-guided Biopsy Specimens.

Background Although prostate MRI is routinely used for the detection and staging of localized prostate cancer, imaging-based assessment and targeted molecular sampling for risk stratification are an active area of research. Purpose To evaluate features of preoperative MRI and MRI-guided biopsy immunohistochemistry (IHC) findings associated with biochemical recurrence (BCR) of prostate cancer after surgery. Materials and Methods In this retrospective case-control study, patients underwent multiparametric MRI before MRI-guided biopsy followed by radical prostatectomy between 2008 and 2016. Lesions were retrospectively scored with the Prostate Imaging Reporting and Data System (PI-RADS) (version 2) by radiologists who were blinded to the clinical-pathologic results. The IHC staining, including stains for the ETS-related gene, phosphatase and tensin homolog, androgen receptor, prostate specific antigen, and p53, was performed with targeted biopsy specimens of the index lesion (highest suspicion at MRI and pathologic grade) and scored by pathologists who were blinded to clinical-pathologic outcomes. Cox proportional hazards regression analysis was used to evaluate associations with recurrence-free survival (RFS). Results The median RFS was 31.7 months (range, 1-101 months) for 39 patients (median age, 62 years; age range, 47-76 years) without BCR and 14.6 months (range, 1-61 months) for 40 patients (median age, 59 years; age range, 47-73 years) with BCR. MRI features that showed a significant relationship with the RFS interval included an index lesion with a PI-RADS score of 5 (hazard ratio [HR], 2.10; 95% CI: 1.05, 4.21; P = .04); index lesion burden, defined as ratio of index lesion volume to prostate volume (HR, 1.55; 95% CI: 1.2, 2.1; P = .003); and suspicion of extraprostatic extension (EPE) (HR, 2.18; 95% CI: 1.1, 4.2; P = .02). Presurgical multivariable analysis indicated that suspicion of EPE at MRI (adjusted HR, 2.19; 95% CI: 1.1, 4.3; P = .02) and p53 stain intensity (adjusted HR, 2.22; 95% CI: 1.0, 4.7; P = .04) were significantly associated with RFS. Conclusion MRI features, including Prostate Imaging Reporting and Data System score, index lesion burden, extraprostatic extension, and preoperative guided biopsy p53 immunohistochemistry stain intensity are associated with biochemical relapse of prostate cancer after surgery. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Costa in this issue.

Comprehensive genomic profiling of metastatic collecting duct carcinoma, renal medullary carcinoma, and clear cell renal cell carcinoma.

INTRODUCTION AND OBJECTIVE: Unlike clear cell renal cell carcinoma (CCRCC), collecting duct carcinoma (CDC) and renal medullary carcinoma (RMC) are rare tumors that progress rapidly and appear resistant to current systemic therapies. We queried comprehensive genomic profiling to uncover opportunities for targeted therapy and immunotherapy. MATERIAL AND METHODS: DNA was extracted from 40 microns of formalin-fixed, paraffin-embedded specimen from relapsed, mCDC (n = 46), mRMC (n = 24), and refractory and metastatic (m) mCCRCC (n = 626). Comprehensive genomic profiling was performed, and Tumor mutational burden (TMB) and microsatellite instability (MSI) were calculated. We analyzed all classes of genomic alterations. RESULTS: mCDC had 1.7 versus 2.7 genomic alterations/tumor in mCCRCC ( = 0.04). Mutations in VHL (P < 0.0001) and TSC1 (P = 0.04) were more frequent in mCCRCC. SMARCB1 (P < 0.0001), NF2 (P = 0.0007), RB1 (P = 0.02) and RET (P = 0.0003) alterations were more frequent in mCDC versus mCCRCC. No VHL alterations in mRMC and mCDC were identified. SMARCB1 genomic alterations were significantly more frequent in mRMC than mCDC (P = 0.0002), but were the most common alterations in both subtypes. Mutations to EGFR, RET, NF2, and TSC2 were more frequently identified in mCDC versus mRMC. The median TMB and MSI-High status was low with <1% of mCCRC, mCDC, and mRMC having ≥ 20 mut/Mb. CONCLUSION: Genomic alteration patterns in mCDC and mRMC differ significantly from mCCRCC. Targeted therapies for mCDC and mRMC appear limited with rare opportunities to target alterations in receptor tyrosine kinase and MTOR pathways. Similarly, TMB and absence of MSI-High status in mCDC and mRMC suggest resistance to immunotherapies.

Risk of adverse pathology at prostatectomy in the era of MRI and targeted biopsies; rethinking active surveillance for intermediate risk prostate cancer patients.

PURPOSE: Men with intermediate risk (IR) prostate cancer (CaP) are often excluded from active surveillance (AS) due to higher rates of adverse pathology (AP). We determined our rate of AP in men who underwent multiparametric MRI (MpMRI) with combined biopsy (CB) consisting of targeted biopsy (TB) and systematic biopsy (SB) prior to radical prostatectomy (RP). METHODS: A retrospective review was conducted of men with Gleason Grade Group (GG) 2 disease who underwent RP after SB alone or after preoperative MRI with CB. AP was defined as either pathologic stage T3a (AP ≥ T3a) or pathologic stage T3b (AP ≥ T3b) and/or GG upgrading. Rates of AP were determined for both groups and those who fit the National Comprehensive Cancer Network (NCCN) definition of favorable IR (FIR) or the low volume IR (LVIR) criteria. Multivariable logistic regression was used to determine predictive factors. RESULTS: The overall rate of AP ≥ T3b was 21.2% in the SB group vs. 8.6% in the MRI with CB group, P = 0.006. This rate was lowered to 6.8% and 5.6% when men met the definition of NCCN FIR or LVIR, respectively. Suspicion for extraprostatic extension (EPE) (OR 7.65, 95% CI 1.77-33.09, P = 0.006) and positive cores of GG 2 on SB (OR 1.43, 95% CI 1.05-1.96, P = 0.023) were significant for predicting AP ≥ T3b. CONCLUSIONS: Rates of AP at RP after MRI with CB are lower than studies prior to the adoption of this technology, suggesting that more men with IR disease may be considered for AS. However, increasing cores positive on SB and MRI findings suggestive of EPE remain unsafe.

Anti-CD19 CAR T cells potently redirected to kill solid tumor cells.

Successful CAR T cell therapy for the treatment of solid tumors requires exemplary CAR T cell expansion, persistence and fitness, and the ability to target tumor antigens safely. Here we address this constellation of critical attributes for successful cellular therapy by using integrated technologies that simplify development and derisk clinical translation. We have developed a CAR-CD19 T cell that secretes a CD19-anti-Her2 bridging protein. This cell therapy strategy exploits the ability of CD19-targeting CAR T cells to interact with CD19 on normal B cells to drive expansion, persistence and fitness. The secreted bridging protein potently binds to Her2-positive tumor cells, mediating CAR-CD19 T cell cytotoxicity in vitro and in vivo. Because of its short half-life, the secreted bridging protein will selectively accumulate at the site of highest antigen expression, ie. at the tumor. Bridging proteins that bind to multiple different tumor antigens have been created. Therefore, antigen-bridging CAR-CD19 T cells incorporate critical attributes for successful solid tumor cell therapy. This platform can be exploited to attack tumor antigens on any cancer.

Federated learning improves site performance in multicenter deep learning without data sharing.

OBJECTIVE: To demonstrate enabling multi-institutional training without centralizing or sharing the underlying physical data via federated learning (FL). MATERIALS AND METHODS: Deep learning models were trained at each participating institution using local clinical data, and an additional model was trained using FL across all of the institutions. RESULTS: We found that the FL model exhibited superior performance and generalizability to the models trained at single institutions, with an overall performance level that was significantly better than that of any of the institutional models alone when evaluated on held-out test sets from each institution and an outside challenge dataset. DISCUSSION: The power of FL was successfully demonstrated across 3 academic institutions while avoiding the privacy risk associated with the transfer and pooling of patient data. CONCLUSION: Federated learning is an effective methodology that merits further study to enable accelerated development of models across institutions, enabling greater generalizability in clinical use.

Multi-Domain Image Completion for Random Missing Input Data.

Multi-domain data are widely leveraged in vision applications taking advantage of complementary information from different modalities, e.g., brain tumor segmentation from multi-parametric magnetic resonance imaging (MRI). However, due to possible data corruption and different imaging protocols, the availability of images for each domain could vary amongst multiple data sources in practice, which makes it challenging to build a universal model with a varied set of input data. To tackle this problem, we propose a general approach to complete the random missing domain(s) data in real applications. Specifically, we develop a novel multi-domain image completion method that utilizes a generative adversarial network (GAN) with a representational disentanglement scheme to extract shared content encoding and separate style encoding across multiple domains. We further illustrate that the learned representation in multi-domain image completion could be leveraged for high-level tasks, e.g., segmentation, by introducing a unified framework consisting of image completion and segmentation with a shared content encoder. The experiments demonstrate consistent performance improvement on three datasets for brain tumor segmentation, prostate segmentation, and facial expression image completion respectively.