RAPHIA: A deep learning pipeline for the registration of MRI and whole-mount histopathology images of the prostate

Image registration can map the ground truth extent of prostate cancer from histopathology images onto MRI, facilitating the development of machine learning methods for early prostate cancer detection. Here, we present RAdiology PatHology Image Alignment (RAPHIA), an end-to-end pipeline for efficient and accurate registration of MRI and histopathology images. RAPHIA automates several time-consuming manual steps in existing approaches including prostate segmentation, estimation of the rotation angle and horizontal flipping in histopathology images, and estimation of MRI-histopathology slice correspondences. By utilizing deep learning registration networks, RAPHIA substantially reduces computational time. Furthermore, RAPHIA obviates the need for a multimodal image similarity metric by transferring histopathology image representations to MRI image representations and vice versa. With the assistance of RAPHIA, novice users achieved expert-level performance, and their mean error in estimating histopathology rotation angle was reduced by 51% (12 degrees vs 8 degrees), their mean accuracy of estimating histopathology flipping was increased by 5% (95.3% vs 100%), and their mean error in estimating MRI-histopathology slice correspondences was reduced by 45% (1.12 slices vs 0.62 slices). When compared to a recent conventional registration approach and a deep learning registration approach, RAPHIA achieved better mapping of histopathology cancer labels, with an improved mean Dice coefficient of cancer regions outlined on MRI and the deformed histopathology (0.44 vs 0.48 vs 0.50), and a reduced mean per-case processing time (51 vs 11 vs 4.5 min). The improved performance by RAPHIA allows efficient processing of large datasets for the development of machine learning models for prostate cancer detection on MRI. Our code is publicly available at: https://github.com/pimed/RAPHIA.

AZGP1 deficiency promotes angiogenesis in prostate cancer

BACKGROUND

Loss of AZGP1 expression is a biomarker associated with progression to castration resistance, development of metastasis, and poor disease-specific survival in prostate cancer. However, high expression of AZGP1 cells in prostate cancer has been reported to increase proliferation and invasion. The exact role of AZGP1 in prostate cancer progression remains elusive.

METHOD

AZGP1 knockout and overexpressing prostate cancer cells were generated using a lentiviral system. The effects of AZGP1 under- or over-expression in prostate cancer cells were evaluated by in vitro cell proliferation, migration, and invasion assays. Heterozygous AZGP1± mice were obtained from European Mouse Mutant Archive (EMMA), and prostate tissues from homozygous knockout male mice were collected at 2, 6 and 10 months for histological analysis. In vivo xenografts generated from AZGP1 under- or over-expressing prostate cancer cells were used to determine the role of AZGP1 in prostate cancer tumor growth, and subsequent proteomics analysis was conducted to elucidate the mechanisms of AZGP1 action in prostate cancer progression. AZGP1 expression and microvessel density were measured in human prostate cancer samples on a tissue microarray of 215 independent patient samples.

RESULT

Neither the knockout nor overexpression of AZGP1 exhibited significant effects on prostate cancer cell proliferation, clonal growth, migration, or invasion in vitro. The prostates of AZGP1-/- mice initially appeared to have grossly normal morphology; however, we observed fibrosis in the periglandular stroma and higher blood vessel density in the mouse prostate by 6 months. In PC3 and DU145 mouse xenografts, over-expression of AZGP1 did not affect tumor growth. Instead, these tumors displayed decreased microvessel density compared to xenografts derived from PC3 and DU145 control cells, suggesting that AZGP1 functions to inhibit angiogenesis in prostate cancer. Proteomics profiling further indicated that, compared to control xenografts, AZGP1 overexpressing PC3 xenografts are enriched with angiogenesis pathway proteins, including YWHAZ, EPHA2, SERPINE1, and PDCD6, MMP9, GPX1, HSPB1, COL18A1, RNH1, and ANXA1. In vitro functional studies show that AZGP1 inhibits human umbilical vein endothelial cell proliferation, migration, tubular formation and branching. Additionally, tumor microarray analysis shows that AZGP1 expression is negatively correlated with blood vessel density in human prostate cancer tissues.

CONCLUSION

AZGP1 is a negative regulator of angiogenesis, such that loss of AZGP1 promotes angiogenesis in prostate cancer. AZGP1 likely exerts heterotypical effects on cells in the tumor microenvironment, such as stromal and endothelial cells. This study sheds light on the anti-angiogenic characteristics of AZGP1 in the prostate and provides a rationale to target AZGP1 to inhibit prostate cancer progression.

Mortality Risks Associated with Depression in Men with Prostate Cancer

BACKGROUND

Men diagnosed with prostate cancer (PC) have an increased risk of depression; however, it is unclear to what extent depression affects long-term survival. A better understanding of such effects is needed to improve long-term care and outcomes for men with PC.

OBJECTIVE

To determine the associations between major depression and mortality in a national cohort of men with PC.

DESIGN, SETTING, AND PARTICIPANTS

A national cohort study was conducted of all 180 189 men diagnosed with PC in Sweden during 1998-2017. Subsequent diagnoses of major depression were ascertained from nationwide outpatient and inpatient records through 2018.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Deaths were identified from nationwide records through 2018. Cox regression was used to compute hazard ratios (HRs) for all-cause mortality associated with major depression, adjusting for sociodemographic factors and comorbidities. Subanalyses assessed differences by PC treatment during 2005-2017. PC-specific mortality was examined using competing risks models.

RESULTS AND LIMITATIONS

In 1.3 million person-years of follow-up, 16 134 (9%) men with PC were diagnosed with major depression and 65 643 (36%) men died. After adjusting for sociodemographic factors and comorbidities, major depression was associated with significantly higher all-cause mortality in men with high-risk PC (HR, 1.50; 95% confidence interval [CI], 1.44-1.55) or low- or intermediate-risk PC (1.64; 1.56-1.71). These risks were elevated regardless of PC treatment or age at PC diagnosis, except for youngest men (<55 yr) in whom the risks were nonsignificant. Major depression was also associated with increased PC-specific mortality in men with either high-risk PC (HR, 1.35; 95% CI, 1.28-1.43) or low- or intermediate-risk PC (1.42; 1.27-1.59). This study was limited to Sweden and will need replication in other countries when feasible.

CONCLUSIONS

In this national cohort of men with PC, major depression was associated with ∼50% higher all-cause mortality. Men with PC need timely detection and treatment of depression to support their long-term outcomes and survival.

PATIENT SUMMARY

In this report, we examined the effects of depression on survival in men with prostate cancer. We found that among all men with prostate cancer, those who developed depression had a 50% higher risk of dying than those without depression. Men with prostate cancer need close monitoring for the detection and treatment of depression to improve their long-term health outcomes.

68Ga-RM2 PET-MRI versus MRI alone for evaluation of patients with biochemical recurrence of prostate cancer: a single-centre, single-arm, phase 2/3 imaging trial

BACKGROUND

National Comprehensive Cancer Network guidelines include prostate-specific membrane antigen (PSMA)-targeted PET for detection of biochemical recurrence of prostate cancer. However, targeting a single tumour characteristic might not be sufficient to reflect the full extent of disease. Gastrin releasing peptide receptors (GRPR) have been shown to be overexpressed in prostate cancer. In this study, we aimed to evaluate the diagnostic performance of the GRPR-targeting radiopharmaceutical 68Ga-RM2 in patients with biochemical recurrence of prostate cancer.

METHODS

This single-centre, single-arm, phase 2/3 trial was done at Stanford University (USA). Adult patients (aged ≥18 years) with biochemical recurrence of prostate cancer, a Karnofsky performance status of 50 or higher, increasing prostate-specific antigen concentration 0·2 ng/mL or more after prostatectomy or 2 ng/mL or more above nadir after radiotherapy, and non-contributory conventional imaging (negative CT or MRI, and bone scan) were eligible. All participants underwent 68Ga-RM2 PET-MRI. The primary outcome was the proportion of patients with PET-positive findings on 68Ga-RM2 PET-MRI compared with MRI alone after initial therapy, at a per-patient and per-lesion level. The primary outcome would be considered met if at least 30% of patients had one or more lesions detected by 68Ga-RM2 PET-MRI and the detection by 68Ga-RM2 PET-MRI was significantly greater than for MRI. Each PET scan was interpreted by three independent masked readers using a standardised evaluation criteria. This study is registered with ClinicalTrials.gov, NCT02624518, and is complete.

FINDINGS

Between Dec 12, 2015, and July 27, 2021, 209 men were screened for eligibility, of whom 100 were included in analyses. Median follow-up was 49·3 months (IQR 36·7-59·2). The primary endpoint was met; 68Ga-RM2 PET-MRI was positive in 69 (69%) patients and MRI alone was positive in 40 (40%) patients (p<0·0001). In the per-lesion analysis 68Ga-RM2 PET-MRI showed significantly higher detection rates than MRI alone (143 vs 96 lesions; p<0·0001). No grade 1 or worse events were reported.

INTERPRETATION

68Ga-RM2 PET-MRI showed better diagnostic performance than MRI alone in patients with biochemical recurrence of prostate cancer. Further prospective comparative studies with PSMA-targeted PET are needed to gain a better understanding of GRPR and PSMA expression patterns in these patients.

FUNDING

The US Department of Defense.

Critical evaluation of artificial intelligence as a digital twin of pathologists for prostate cancer pathology

Prostate cancer pathology plays a crucial role in clinical management but is time-consuming. Artificial intelligence (AI) shows promise in detecting prostate cancer and grading patterns. We tested an AI-based digital twin of a pathologist, vPatho, on 2603 histological images of prostate tissue stained with hematoxylin and eosin. We analyzed various factors influencing tumor grade discordance between the vPatho system and six human pathologists. Our results demonstrated that vPatho achieved comparable performance in prostate cancer detection and tumor volume estimation, as reported in the literature. The concordance levels between vPatho and human pathologists were examined. Notably, moderate to substantial agreement was observed in identifying complementary histological features such as ductal, cribriform, nerve, blood vessel, and lymphocyte infiltration. However, concordance in tumor grading decreased when applied to prostatectomy specimens (κ = 0.44) compared to biopsy cores (κ = 0.70). Adjusting the decision threshold for the secondary Gleason pattern from 5 to 10% improved the concordance level between pathologists and vPatho for tumor grading on prostatectomy specimens (κ from 0.44 to 0.64). Potential causes of grade discordance included the vertical extent of tumors toward the prostate boundary and the proportions of slides with prostate cancer. Gleason pattern 4 was particularly associated with this population. Notably, the grade according to vPatho was not specific to any of the six pathologists involved in routine clinical grading. In conclusion, our study highlights the potential utility of AI in developing a digital twin for a pathologist. This approach can help uncover limitations in AI adoption and the practical application of the current grading system for prostate cancer pathology.

UCHL1 is a potential molecular indicator and therapeutic target for neuroendocrine carcinomas

Neuroendocrine carcinomas, such as neuroendocrine prostate cancer and small-cell lung cancer, commonly have a poor prognosis and limited therapeutic options. We report that ubiquitin carboxy-terminal hydrolase L1 (UCHL1), a deubiquitinating enzyme, is elevated in tissues and plasma from patients with neuroendocrine carcinomas. Loss of UCHL1 decreases tumor growth and inhibits metastasis of these malignancies. UCHL1 maintains neuroendocrine differentiation and promotes cancer progression by regulating nucleoporin, POM121, and p53. UCHL1 binds, deubiquitinates, and stabilizes POM121 to regulate POM121-associated nuclear transport of E2F1 and c-MYC. Treatment with the UCHL1 inhibitor LDN-57444 slows tumor growth and metastasis across neuroendocrine carcinomas. The combination of UCHL1 inhibitors with cisplatin, the standard of care used for neuroendocrine carcinomas, significantly delays tumor growth in pre-clinical settings. Our study reveals mechanisms of UCHL1 function in regulating the progression of neuroendocrine carcinomas and identifies UCHL1 as a therapeutic target and potential molecular indicator for diagnosing and monitoring treatment responses in these malignancies.

Identification and characterization of intact glycopeptides in human urine

Glycoproteins in urine have the potential to provide a rich class of informative molecules for studying human health and disease. Despite this promise, the urine glycoproteome has been largely uncharacterized. Here, we present the analysis of glycoproteins in human urine using LC-MS/MS-based intact glycopeptide analysis, providing both the identification of protein glycosites and characterization of the glycan composition at specific glycosites. Gene enrichment analysis reveals differences in biological processes, cellular components, and molecular functions in the urine glycoproteome versus the urine proteome, as well as differences based on the major glycan class observed on proteins. Meta-heterogeneity of glycosylation is examined on proteins to determine the variation in glycosylation across multiple sites of a given protein with specific examples of individual sites differing from the glycosylation trends in the overall protein. Taken together, this dataset represents a potentially valuable resource as a baseline characterization of glycoproteins in human urine for future urine glycoproteomics studies.

Spatial transcriptomics identifies candidate stromal drivers of benign prostatic hyperplasia

Benign prostatic hyperplasia (BPH) is the nodular proliferation of the prostate transition zone in older men, leading to urinary storage and voiding problems that can be recalcitrant to therapy. Decades ago, John McNeal proposed that BPH originates with the "reawakening" of embryonic inductive activity by adult prostate stroma, which spurs new ductal proliferation and branching morphogenesis. Here, by laser microdissection and transcriptional profiling of the BPH stroma adjacent to hyperplastic branching ducts, we identified secreted factors likely mediating stromal induction of prostate glandular epithelium and coinciding processes. The top stromal factors were insulin-like growth factor 1 (IGF1) and CXC chemokine ligand 13 (CXCL13), which we verified by RNA in situ hybridization to be coexpressed in BPH fibroblasts, along with their cognate receptors (IGF1R and CXCR5) on adjacent epithelium. In contrast, IGF1 but not CXCL13 was expressed in human embryonic prostate stroma. Finally, we demonstrated that IGF1 is necessary for the generation of BPH-1 cell spheroids and patient-derived BPH cell organoids in 3D culture. Our findings partially support historic speculations on the etiology of BPH and provide what we believe to be new molecular targets for rational therapies directed against the underlying processes driving BPH.

Deep learning-based automated pipeline for blood vessel detection and distribution analysis in multiplexed prostate cancer images

Introduction: Prostate cancer is a highly heterogeneous disease, presenting varying levels of aggressiveness and response to treatment. Angiogenesis is one of the hallmarks of cancer, providing oxygen and nutrient supply to tumors. Micro vessel density has previously been correlated with higher Gleason score and poor prognosis. Manual segmentation of blood vessels (BVs) In microscopy images is challenging, time consuming and may be prone to inter-rater variabilities. In this study, an automated pipeline is presented for BV detection and distribution analysis in multiplexed prostate cancer images. Methods: A deep learning model was trained to segment BVs by combining CD31, CD34 and collagen IV images. In addition, the trained model was used to analyze the size and distribution patterns of BVs in relation to disease progression in a cohort of prostate cancer patients (N = 215). Results: The model was capable of accurately detecting and segmenting BVs, as compared to ground truth annotations provided by two reviewers. The precision (P), recall (R) and dice similarity coefficient (DSC) were equal to 0.93 (SD 0.04), 0.97 (SD 0.02) and 0.71 (SD 0.07) with respect to reviewer 1, and 0.95 (SD 0.05), 0.94 (SD 0.07) and 0.70 (SD 0.08) with respect to reviewer 2, respectively. BV count was significantly associated with 5-year recurrence (adjusted p = 0.0042), while both count and area of blood vessel were significantly associated with Gleason grade (adjusted p = 0.032 and 0.003 respectively). Discussion: The proposed methodology is anticipated to streamline and standardize BV analysis, offering additional insights into the biology of prostate cancer, with broad applicability to other cancers.

High expression of Trop2 is associated with aggressive localized prostate cancer and is a candidate urinary biomarker

Distinguishing indolent from clinically significant localized prostate cancer is a major clinical challenge and influences clinical decision-making between treatment and active surveillance. The development of novel predictive biomarkers will help with risk stratification, and clinical decision-making, leading to a decrease in over or under-treatment of patients with prostate cancer. Here, we report that Trop2 is a prognostic tissue biomarker for clinically significant prostate cancer by utilizing the Canary Prostate Cancer Tissue Microarray (CPCTA) cohort composed of over 1100 patients from a multi-institutional study. We demonstrate that elevated Trop2 expression is correlated with worse clinical features including Gleason score, age, and pre-operative PSA levels. More importantly, we demonstrate that elevated Trop2 expression at radical prostatectomy predicts worse overall survival in men undergoing radical prostatectomy. Additionally, we detect shed Trop2 in urine from men with clinically significant prostate cancer. Our study identifies Trop2 as a novel tissue prognostic biomarker and a candidate non-invasive marker for prostate cancer.

Risks of Depression, Anxiety, and Suicide in Partners of Men with Prostate Cancer: A National Cohort Study

BACKGROUND

A diagnosis of prostate cancer (PC) may cause psychosocial distress not only in a man but also his intimate partner. However, long-term risks of depression, anxiety, or suicide in partners of men with PC are largely unknown.

METHODS

A national cohort study was conducted of 121,530 partners of men diagnosed with PC during 1998-2017 and 1,093,304 population-based controls in Sweden. Major depression, anxiety disorder, and suicide death were ascertained through 2018. Cox regression was used to compute hazard ratios (HRs) while adjusting for sociodemographic factors.

RESULTS

Partners of men with high-risk PC had increased risks of major depression (adjusted HR, 1.34; 95% CI, 1.30-1.39) and anxiety disorder (1.25; 1.20-1.30), which remained elevated ≥10 years later. Suicide death was increased in partners of men with distant metastases (adjusted HR, 2.38; 95% CI, 1.08-5.22) but not other high-risk PC (1.14; 0.70-1.88). Among partners of men with high-risk PC, risks of major depression and anxiety disorder were highest among those aged ≥80 years (adjusted HR, 1.73; 95% CI, 1.53-1.96; and 1.70; 1.47-1.96, respectively), whereas suicide death was highest among those aged <60 years (7.55; 2.20-25.89). In contrast, partners of men with low- or intermediate-risk PC had modestly or no increased risks of these outcomes.

CONCLUSIONS

In this large cohort, partners of men with high-risk PC had increased risks of major depression and anxiety disorder, which persisted for ≥10 years. Suicide death was increased 2-fold in partners of men with distant metastases. Partners as well as men with PC need psychosocial support and close follow-up for psychosocial distress.

Patient and physician perspectives on treatments for low-risk prostate cancer: a qualitative study

BACKGROUND

Patients diagnosed with low-risk prostate cancer (PCa) are confronted with a difficult decision regarding whether to undergo definitive treatment or to pursue an active surveillance protocol. This is potentially further complicated by the possibility that patients and physicians may place different value on factors that influence this decision. We conducted a qualitative investigation to better understand patient and physician perceptions of factors influencing treatment decisions for low-risk PCa.

METHODS

Semi-structured interviews were conducted among 43 racially and ethnically diverse patients diagnosed with low-risk PCa, who were identified through a population-based cancer registry, and 15 physicians who were selected to represent a variety of practice settings in the Greater San Francisco Bay Area.

RESULTS

Patients and physicians both described several key individual (e.g., clinical) and interpersonal (e.g., healthcare communications) factors as important for treatment decision-making. Overall, physicians' perceptions largely mirrored patients' perceptions. First, we observed differences in treatment preferences by age and stage of life. At older ages, there was a preference for less invasive options. However, at younger ages, we found varying opinions among both patients and physicians. Second, patients and physicians both described concerns about side effects including physical functioning and non-physical considerations. Third, we observed differences in expectations and the level of difficulty for clinical conversations based on information needs and resources between patients and physicians. Finally, we discovered that patients and physicians perceived patients' prior knowledge and the support of family/friends as facilitators of clinical conversations.

CONCLUSIONS

Our study suggests that the gap between patient and physician perceptions on the influence of clinical and communication factors on treatment decision-making is not large. The consensus we observed points to the importance of developing relevant clinical communication roadmaps as well as high quality and accessible patient education materials.

ACAA2 is a novel molecular indicator for cancers with neuroendocrine phenotype

BACKGROUND

Neuroendocrine phenotype is commonly associated with therapy resistance and poor prognoses in small-cell neuroendocrine cancers (SCNCs), such as neuroendocrine prostate cancer (NEPC) and small-cell lung cancer (SCLC). Expression levels of current neuroendocrine markers exhibit high case-by-case variability, so multiple markers are used in combination to identify SCNCs. Here, we report that ACAA2 is elevated in SCNCs and is a potential molecular indicator for SCNCs.

METHODS

ACAA2 expressions in tumour xenografts, tissue microarrays (TMAs), and patient tissues from prostate and lung cancers were analysed via immunohistochemistry. ACAA2 mRNA levels in lung and prostate cancer (PC) patients were assessed in published datasets.

RESULTS

ACAA2 protein and mRNA levels were elevated in SCNCs relative to non-SCNCs. Medium/high ACAA2 intensity was observed in 78% of NEPC PDXs samples (N = 27) relative to 33% of adeno-CRPC (N = 86), 2% of localised PC (N = 50), and 0% of benign prostate specimens (N = 101). ACAA2 was also elevated in lung cancer patient tissues with neuroendocrine phenotype. 83% of lung carcinoid tissues (N = 12) and 90% of SCLC tissues (N = 10) exhibited medium/high intensity relative to 40% of lung adenocarcinoma (N = 15).

CONCLUSION

ACAA2 expression is elevated in aggressive SCNCs such as NEPC and SCLC, suggesting it is a potential molecular indicator for SCNCs.

Artificial Intelligence-Based PTEN Loss Assessment as an Early Predictor of Prostate Cancer Metastasis After Surgery: A Multicenter Retrospective Study

Phosphatase and tensin homolog (PTEN) loss is associated with adverse outcomes in prostate cancer and can be measured via immunohistochemistry. The purpose of the study was to establish the clinical application of an in-house developed artificial intelligence (AI) image analysis workflow for automated detection of PTEN loss on digital images for identifying patients at risk of early recurrence and metastasis. Postsurgical tissue microarray sections from the Canary Foundation (n = 1264) stained with anti-PTEN antibody were evaluated independently by pathologist conventional visual scoring (cPTEN) and an automated AI-based image analysis pipeline (AI-PTEN). The relationship of PTEN evaluation methods with cancer recurrence and metastasis was analyzed using multivariable Cox proportional hazard and decision curve models. Both cPTEN scoring by the pathologist and quantification of PTEN loss by AI (high-risk AI-qPTEN) were significantly associated with shorter metastasis-free survival (MFS) in univariable analysis (cPTEN hazard ratio [HR], 1.54; CI, 1.07-2.21; P = .019; AI-qPTEN HR, 2.55; CI, 1.83-3.56; P < .001). In multivariable analyses, AI-qPTEN showed a statistically significant association with shorter MFS (HR, 2.17; CI, 1.49-3.17; P < .001) and recurrence-free survival (HR, 1.36; CI, 1.06-1.75; P = .016) when adjusting for relevant postsurgical clinical nomogram (Cancer of the Prostate Risk Assessment [CAPRA] postsurgical score [CAPRA-S]), whereas cPTEN does not show a statistically significant association (HR, 1.33; CI, 0.89-2; P = .2 and HR, 1.26; CI, 0.99-1.62; P = .063, respectively) when adjusting for CAPRA-S risk stratification. More importantly, AI-qPTEN was associated with shorter MFS in patients with favorable pathological stage and negative surgical margins (HR, 2.72; CI, 1.46-5.06; P = .002). Workflow also demonstrated enhanced clinical utility in decision curve analysis, more accurately identifying men who might benefit from adjuvant therapy postsurgery. This study demonstrates the clinical value of an affordable and fully automated AI-powered PTEN assessment for evaluating the risk of developing metastasis or disease recurrence after radical prostatectomy. Adding the AI-qPTEN assessment workflow to clinical variables may affect postoperative surveillance or management options, particularly in low-risk patients.

Long-term Risks of Depression and Suicide Among Men with Prostate Cancer: A National Cohort Study

BACKGROUND

A diagnosis of prostate cancer (PC) may cause psychosocial distress that worsens quality of life; however, long-term mental health outcomes are unclear.

OBJECTIVE

To determine the long-term risks of major depression and death by suicide in a large population-based cohort.

DESIGN, SETTING, AND PARTICIPANTS

This was a national cohort study of 180 189 men diagnosed with PC during 1998-2017 and 1 801 890 age-matched, population-based, control men in Sweden.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Major depression and death by suicide were ascertained from nationwide outpatient, inpatient, and death records up to 2018. Cox regression was used to compute hazard ratios (HRs) adjusted for sociodemographic factors and comorbidities. Subanalyses assessed differences by PC treatment during 2005-2017.

RESULTS AND LIMITATIONS

Men diagnosed with high-risk PC had higher relative rates of major depression (adjusted HR [aHR] 1.82, 95% confidence interval [CI] 1.75-1.89) and death by suicide (aHR 2.43, 95% CI 2.01-2.95). These associations persisted for ≥10 yr after PC diagnosis. The relative increase in major depression was lower among those treated with radiation (aHR 1.44, 95% CI 1.31-1.57) or surgery (aHR 1.60, 95% CI 1.31-1.95) in comparison to androgen deprivation therapy (ADT) alone (aHR 2.02, 95% CI 1.89-2.16), whereas the relative rate of suicide death was higher only among those treated solely with ADT (aHR 2.83, 95% CI 1.80-4.43). By contrast, men with low- or intermediate-risk PC had a modestly higher relative rate of major depression (aHR 1.19, 95% CI 1.16-1.23) and higher relative rate of suicide death at 3-12 mo after PC diagnosis (aHR 1.88, 95% CI 1.11-3.18) but not across the entire follow-up period (aHR 1.02, 95% CI 0.84-1.25). This study was limited to Sweden and will need replication in other populations.

CONCLUSIONS

In this large cohort, high-risk PC was associated with substantially higher relative rates of major depression and death by suicide, which persisted for ≥10 yr after PC diagnosis. PC survivors need close follow-up for timely detection and treatment of psychosocial distress.

PATIENT SUMMARY

In a large Swedish population, men with aggressive prostate cancer had higher long-term relative rates of depression and suicide.

Identification of age- and immune-related gene signatures for clinical outcome prediction in lung adenocarcinoma

BACKGROUND

The understanding of the factors causing decreased overall survival (OS) in older patients compared to younger patients in lung adenocarcinoma (LUAD) remains.

METHODS

Gene expression profiles of LUAD were obtained from publicly available databases by Kaplan-Meier analysis was performed to determine whether age was associated with patient OS. The immune cell composition in the tumor microenvironment (TME) was evaluated using CIBERSORT. The fraction of stromal and immune cells in tumor samples were also using assessed using multiple tools including ESTIMATE, EPIC, and TIMER. Differentially expressed genes (DEGs) from the RNA-Seq data that were associated with age and immune cell composition were identified using the R package DEGseq. A 22-gene signature composed of DEGs associated with age and immune cell composition that predicted OS were constructed using Least Absolute Shrinkage and Selection Operator (LASSO).

RESULTS

In The Cancer Genome Atlas (TCGA)-LUAD dataset, we found that younger patients (≤70) had a significant better OS compared to older patients (>70). In addition, older patients had significantly higher expression of immune checkpoint proteins including inhibitory T cell receptors and their ligands. Moreover, analyses using multiple bioinformatics tools showed increased immune infiltration, including CD4+ T cells, in older patients compared to younger patients. We identified a panel of genes differentially expressed between patients >70 years compared to those ≤70 years, as well as between patients with high or low immune scores and selected 84 common genes to construct a prognostic gene signature. A risk score calculated based on 22 genes selected by LASSO predicted 1, 3, and 5-year OS, with an area under the curve (AUC) of 0.72, 0.72, 0.69, receptively, in TCGA-LUAD dataset and an independent validation dataset available from the European Genome-phenome Archive (EGA).

CONCLUSION

Our results demonstrate that age contributes to OS of LUAD patients atleast in part through its association with immune infiltration in the TME.

Risks of alcohol and drug use disorders in prostate cancer survivors: a national cohort study

BACKGROUND

Prostate cancer (PC) survivors may potentially use substances to cope with psychological distress or poorly controlled physical symptoms. Little is known, however, about the long-term risks of alcohol use disorder (AUD) or drug use disorders in men with PC.

METHODS

A national cohort study was conducted in Sweden of 180 189 men diagnosed with PC between 1998 and 2017 and 1 801 890 age-matched population-based control men. AUD and drug use disorders were ascertained from nationwide records through 2018. Cox regression was used to compute hazard ratios (HRs) while adjusting for sociodemographic factors and prior psychiatric disorders. Subanalyses examined differences by PC treatment from 2005 to 2017.

RESULTS

Men with high-risk PC had increased risks of both AUD (adjusted HR = 1.44, 95% confidence interval [CI] = 1.33 to 1.57) and drug use disorders (adjusted HR = 1.93, 95% CI = 1.67 to 2.24). Their AUD risk was highest in the first year and was no longer significantly elevated 5 years after PC diagnosis, whereas their drug use disorders risk remained elevated 10 years after PC diagnosis (adjusted HR = 2.26, 95% CI = 1.45 to 3.52), particularly opioid use disorder (adjusted HR = 3.07, 95% CI = 1.61 to 5.84). Those treated only with androgen-deprivation therapy had the highest risks of AUD (adjusted HR = 1.91, 95% CI = 1.62 to 2.25) and drug use disorders (adjusted HR = 2.23, 95% CI = 1.70 to 2.92). Low- or intermediate-risk PC was associated with modestly increased risks of AUD (adjusted HR = 1.38, 95% CI = 1.30 to 1.46) and drug use disorders (adjusted HR = 1.19, 95% CI = 1.06 to 1.34).

CONCLUSIONS

In this large cohort, men with PC had significantly increased risks of both AUD and drug use disorders, especially those with high-risk PC and treated only with androgen-deprivation therapy. PC survivors need long-term psychosocial support and timely detection and treatment of AUD and drug use disorders.

Proteomics analysis of urine and catheter-associated biofilms in spinal cord injury patients

After spinal cord injury (SCI), use chronic urinary catheters for bladder management is common, making these patients especially vulnerable to catheter-associated complications. Chronic catheterization is associated with bacterial colonization and frequent catheter-associated urinary tract infections (CAUTI). One determinant of infection success and treatment resistance is production of catheter-associated biofilms, composed of microorganisms and host- and microbial-derived components. To better understand the biofilm microenvironment, we performed proteomics analysis of catheter-associated biofilms and paired urine samples from four people with SCI with chronic indwelling urinary catheters. We developed a novel method for the removal of adhered cellular components on catheters that contained both human and microbial homologous proteins. Proteins from seven microbial species were identified including: Escherichia coli, Klebsiella species (spp), Enterococcus spp, Proteus mirabilis, Pseudomonas spp, Staphylococcus spp, and Candida spp. Peptides identified from catheter biofilms were assigned to 4,820 unique proteins, with 61% of proteins assigned to the biofilm-associated microorganisms, while the remainder were human-derived. Contrastingly, in urine, only 51% were assigned to biofilm-associated microorganisms and 4,554 proteins were identified as a human-derived. Of the proteins assigned to microorganisms in the biofilm and paired urine, Enterococcus, Candida spp, and P. mirabilis had greater associations with the biofilm phase, whereas E. coli and Klebsiella had greater associations with the urine phase, thus demonstrating a significant difference between the urine and adhered microbial communities. The microbial proteins that differed significantly between the biofilm and paired urine samples mapped to pathways associated with amino acid synthesis, likely related to adaptation to high urea concentrations in the urine, and growth and protein synthesis in bacteria in the biofilm. Human proteins demonstrated enrichment for immune response in the catheter-associated biofilm. Proteomic analysis of catheter-associated biofilms and paired urine samples has the potential to provide detailed information on host and bacterial responses to chronic indwelling urinary catheters and could be useful for understanding complications of chronic indwelling catheters including CAUTIs, urinary stones, and catheter blockages.

Distinguishing Renal Cell Carcinoma From Normal Kidney Tissue Using Mass Spectrometry Imaging Combined With Machine Learning

PURPOSE

Accurately distinguishing renal cell carcinoma (RCC) from normal kidney tissue is critical for identifying positive surgical margins (PSMs) during partial and radical nephrectomy, which remains the primary intervention for localized RCC. Techniques that detect PSM with higher accuracy and faster turnaround time than intraoperative frozen section (IFS) analysis can help decrease reoperation rates, relieve patient anxiety and costs, and potentially improve patient outcomes.

MATERIALS AND METHODS

Here, we extended our combined desorption electrospray ionization mass spectrometry imaging (DESI-MSI) and machine learning methodology to identify metabolite and lipid species from tissue surfaces that can distinguish normal tissues from clear cell RCC (ccRCC), papillary RCC (pRCC), and chromophobe RCC (chRCC) tissues.

RESULTS

From 24 normal and 40 renal cancer (23 ccRCC, 13 pRCC, and 4 chRCC) tissues, we developed a multinomial lasso classifier that selects 281 total analytes from over 27,000 detected molecular species that distinguishes all histological subtypes of RCC from normal kidney tissues with 84.5% accuracy. On the basis of independent test data reflecting distinct patient populations, the classifier achieves 85.4% and 91.2% accuracy on a Stanford test set (20 normal and 28 RCC) and a Baylor-UT Austin test set (16 normal and 41 RCC), respectively. The majority of the model's selected features show consistent trends across data sets affirming its stable performance, where the suppression of arachidonic acid metabolism is identified as a shared molecular feature of ccRCC and pRCC.

CONCLUSION

Together, these results indicate that signatures derived from DESI-MSI combined with machine learning may be used to rapidly determine surgical margin status with accuracies that meet or exceed those reported for IFS.

HostSeq: a Canadian whole genome sequencing and clinical data resource

HostSeq was launched in April 2020 as a national initiative to integrate whole genome sequencing data from 10,000 Canadians infected with SARS-CoV-2 with clinical information related to their disease experience. The mandate of HostSeq is to support the Canadian and international research communities in their efforts to understand the risk factors for disease and associated health outcomes and support the development of interventions such as vaccines and therapeutics. HostSeq is a collaboration among 13 independent epidemiological studies of SARS-CoV-2 across five provinces in Canada. Aggregated data collected by HostSeq are made available to the public through two data portals: a phenotype portal showing summaries of major variables and their distributions, and a variant search portal enabling queries in a genomic region. Individual-level data is available to the global research community for health research through a Data Access Agreement and Data Access Compliance Office approval. Here we provide an overview of the collective project design along with summary level information for HostSeq. We highlight several statistical considerations for researchers using the HostSeq platform regarding data aggregation, sampling mechanism, covariate adjustment, and X chromosome analysis. In addition to serving as a rich data source, the diversity of study designs, sample sizes, and research objectives among the participating studies provides unique opportunities for the research community.

Fusion Gene Detection in Prostate Cancer Samples Enhances the Prediction of Prostate Cancer Clinical Outcomes from Radical Prostatectomy through Machine Learning in a Multi-Institutional Analysis

Prostate cancer remains one of the most fatal malignancies in men in the United States. Predicting the course of prostate cancer is challenging given that only a fraction of prostate cancer patients experience cancer recurrence after radical prostatectomy or radiation therapy. This study examined the expressions of 14 fusion genes in 607 prostate cancer samples from the University of Pittsburgh, Stanford University, and the University of Wisconsin-Madison. The profiling of 14 fusion genes was integrated with Gleason score of the primary prostate cancer and serum prostate-specific antigen level to develop machine-learning models to predict the recurrence of prostate cancer after radical prostatectomy. Machine-learning algorithms were developed by analysis of the data from the University of Pittsburgh cohort as a training set using the leave-one-out cross-validation method. These algorithms were then applied to the data set from the combined Stanford/Wisconsin cohort (testing set). The results showed that the addition of fusion gene profiling consistently improved the prediction accuracy rate of prostate cancer recurrence by Gleason score, serum prostate-specific antigen level, or a combination of both. These improvements occurred in both the training and testing cohorts and were corroborated by multiple models.

NUSAP1 Binds ILF2 to Modulate R-Loop Accumulation and DNA Damage in Prostate Cancer

Increased expression of NUSAP1 has been identified as a robust prognostic biomarker in prostate cancer and other malignancies. We have previously shown that NUSAP1 is positively regulated by E2F1 and promotes cancer invasion and metastasis. To further understand the biological function of NUSAP1, we used affinity purification and mass spectrometry proteomic analysis to identify NUSAP1 interactors. We identified 85 unique proteins in the NUSAP1 interactome, including ILF2, DHX9, and other RNA-binding proteins. Using proteomic approaches, we uncovered a function for NUSAP1 in maintaining R-loops and in DNA damage response through its interaction with ILF2. Co-immunoprecipitation and colocalization using confocal microscopy verified the interactions of NUSAP1 with ILF2 and DHX9, and RNA/DNA hybrids. We showed that the microtubule and charged helical domains of NUSAP1 were necessary for the protein-protein interactions. Depletion of ILF2 alone further increased camptothecin-induced R-loop accumulation and DNA damage, and NUSAP1 depletion abolished this effect. In human prostate adenocarcinoma, NUSAP1 and ILF2 mRNA expression levels are positively correlated, elevated, and associated with poor clinical outcomes. Our study identifies a novel role for NUSAP1 in regulating R-loop formation and accumulation in response to DNA damage through its interactions with ILF2 and hence provides a potential therapeutic target.

Factors that influence treatment decisions: A qualitative study of racially and ethnically diverse patients with low- and very-low risk prostate cancer

BACKGROUND

Factors that influence prostate cancer treatment decisions are complex, multifaceted, and personal, and may vary by race/ethnicity. Although research has been published to quantify factors involved in decision-making, these studies have been limited to primarily white, and to a lesser extent, Black patients, and quantitative studies are limited for discerning the cultural and contextual processes that shape decision-making.

METHODS

We conducted 43 semi-structured interviews with a racially and ethnically diverse sample of patients diagnosed with low- and very-low risk prostate cancer who had undergone treatment for their prostate cancer. Interviews were transcribed, independently coded, and analyzed to identify themes salient for decision-making, with attention to sociocultural differences.

RESULTS

We found racial and ethnic differences in three areas. First, we found differences in how socialized masculinity influenced patient's feelings about different treatment options. Second, we found that for some men, religion and spirituality alleviated anxiety associated with the active surveillance protocol. Finally, for racially and ethnically minoritized patients, we found descriptions of how historic and social experiences within the healthcare system influenced decision-making.

CONCLUSIONS

Our study adds to the current literature by expounding on racial and ethnic differences in the multidimensional, nuanced factors related to decision-making. Our findings suggest that factors associated with prostate cancer decision-making can manifest differently across racial and ethnic groups, and provide some guidance for future research.

PlexusNet: A neural network architectural concept for medical image classification

State-of-the-art (SOTA) convolutional neural network models have been widely adapted in medical imaging and applied to address different clinical problems. However, the complexity and scale of such models may not be justified in medical imaging and subject to the available resource budget. Further increasing the number of representative feature maps for the classification task decreases the model explainability. The current data normalization practice is fixed prior to model development and discounting the specification of the data domain. Acknowledging these issues, the current work proposed a new scalable model family called PlexusNet; the block architecture and model scaling by the network's depth, width, and branch regulate PlexusNet's architecture. The efficient computation costs outlined the dimensions of PlexusNet scaling and design. PlexusNet includes a new learnable data normalization algorithm for better data generalization. We applied a simple yet effective neural architecture search to design PlexusNet tailored to five clinical classification problems that achieve a performance noninferior to the SOTA models ResNet-18 and EfficientNet B0/1. It also does so with lower parameter capacity and representative feature maps in ten-fold ranges than the smallest SOTA models with comparable performance. The visualization of representative features revealed distinguishable clusters associated with categories based on latent features generated by PlexusNet. The package and source code are at https://github.com/oeminaga/PlexusNet.git.

Associations of Renal Cell Carcinoma Subtype with Patient Demographics, Comorbidities, and Neighborhood Socioeconomic Status in the California Population

BACKGROUND

Renal cell carcinoma (RCC) subtypes differ in molecular characteristics and prognosis. We investigated the associations of RCC subtype with patient demographics, comorbidity, and neighborhood socioeconomic status (nSES).

METHODS

Using linked California Cancer Registry and Office of Statewide Health Planning and Development data, we identified history of hypertension, diabetes, and kidney disease prior to RCC diagnosis in Asian/Pacific Islander, non-Latino Black, Latino, and non-Latino White adults diagnosed with their first pathologically confirmed RCC from 2005 through 2015. We used multinomial multivariable logistic regression to model the association of demographics, comorbidity, and nSES with clear-cell, papillary, and chromophobe RCC subtype.

RESULTS

Of the 40,016 RCC cases included, 62.6% were clear cell, 10.9% papillary, and 5.9% chromophobe. The distribution of subtypes differed strikingly by race and ethnicity, ranging from 40.4% clear cell and 30.4% papillary in non-Latino Black adults to 70.7% clear cell and 4.5% papillary in Latino adults. In multivariable analysis, non-Latino Black individuals had a higher likelihood of presenting with papillary (OR, 3.99; 95% confidence interval, 3.61-4.42) and chromophobe (OR, 1.81; 1.54-2.13) versus clear-cell subtype compared with non-Latino White individuals. Both hypertension (OR, 1.19; 1.10-1.29) and kidney disease (OR, 2.38; 2.04-2.77 end-stage disease; OR, 1.52; 1.33-1.72 non-end-stage disease) were associated with papillary subtype. Diabetes was inversely associated with both papillary (OR, 0.63; 0.58-0.69) and chromophobe (OR, 0.61; 0.54-0.70) subtypes.

CONCLUSIONS

RCC subtype is independently associated with patient demographics, and comorbidity.

IMPACT

Targeted RCC treatments or RCC prevention efforts may have differential impact across population subgroups.

Prosteria - National Trends and Outcomes of More Frequent Than Guideline Recommended Prostate Specific Antigen Screening

OBJECTIVE

To characterize national trends in and associated outcomes of more often than annual prostate-specific antigen (PSA) screening, which we term "prosteria."

METHODS

Men in the Optum Clinformatics Data Mart with ≥2 years from first PSA test to censoring at the end of insurance or available data (January 2003 to June 2019) or following exclusionary diagnoses or procedures, such as PCa treatment, were included. PSAs within 90 days were treated as one PSA. Prosteria was defined as having ≥3 PSA testing intervals of ≤270 days.

RESULTS

A total of 9,734,077 PSAs on 2,958,923 men were included. The average inter-PSA testing interval was 1.5 years, and 4.5% of men had prosteria, which increased by 0.53% per year. Educated, wealthy, non-White patients were more likely to have prosteria. Men within the recommended screening age (ie 55-69) had lower rates of prosteria. Prosteria patients had higher average PSA values (2.5 vs 1.4 ng/mL), but lower values at PCa diagnosis. Prosteria was associated with biopsy and PCa diagnosis; however, there were comparable rates of treatment within 2 years of diagnosis.

CONCLUSION

In this large cohort study, prosteria was common, increased over time, and was associated with demographic characteristics. Importantly, there were no clinically meaningful differences in PSA values at diagnosis or rates of early treatment, suggesting prosteria leads to both overdiagnosis and overtreatment. These results support current AUA and USPTF guidelines and can be used to counsel men seeking more frequent PSA screening.

Abstract A13: Siglec-7/9-sialic acid interactions inhibit T cell immune response in prostate cancer

Immunotherapy has rapidly expanded in the care of patients with cancer with the discovery of immune checkpoints like PD-1 and CTLA-4. However, current immune checkpoint inhibitors are largely ineffective for prostate cancer. Recent studies suggest an alternative immune evasion pathway through the interactions between Sialic acid-binding immunoglobulin-type lectin proteins (Siglec) and their ligands, sialylated glycoprotein. Siglec-7/9-sialic acid interactions are reported to inhibit the immune cell response in several cancer types including leukemia, melanoma, and non-small cell lung cancer. Here, we demonstrated that Siglec-7/9 ligands were expressed in both prostate cancer tumor tissues and cell lines. We promoted T cell-mediated cytotoxicity of cancer cells by disrupting the interactions between Siglec-7/9 and their ligands. We discovered that FXYD5 and CD59 are potential Siglec-7 and Siglec-9 ligands, respectively, with CRISPRi screen. We then found that FXYD5 and CD59 knockout cells had reduced Siglec-7/9 binding capacity and enhanced T cells mediated killing effects on prostate cancer cells. These results provide a rationale for novel immune checkpoints and potential approaches for targeting Siglec-7/FXYD5 and Siglec-9/CD59 immune checkpoints for prostate cancer.

Citation Format: Ru M Wen, Jessica C. Stark C Stark, Fernando García-Marqués, Hongjuan Zhao, Rosie Nolley, Carolyn R Bertozzi, Sharon J Pitteri, James D. Brooks. Siglec-7/9-sialic acid interactions inhibit T cell immune response in prostate cancer [abstract]. In: Proceedings of the AACR Special Conference: Tumor Immunology and Immunotherapy; 2022 Oct 21-24; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2022;10(12 Suppl):Abstract nr A13.

Multiparametric Magnetic Resonance Imaging and Metabolic Characterization of Patient-Derived Xenograft Models of Clear Cell Renal Cell Carcinoma

Patient-derived xenografts (PDX) are high-fidelity cancer models typically credentialled by genomics, transcriptomics and proteomics. Characterization of metabolic reprogramming, a hallmark of cancer, is less frequent. Dysregulated metabolism is a key feature of clear cell renal cell carcinoma (ccRCC) and authentic preclinical models are needed to evaluate novel imaging and therapeutic approaches targeting metabolism. We characterized 5 PDX from high-grade or metastatic ccRCC by multiparametric magnetic resonance imaging (MRI) and steady state metabolic profiling and flux analysis. Similar to MRI of clinical ccRCC, T2-weighted images of orthotopic tumors of most PDX were homogeneous. The increased hyperintense (cystic) areas observed in one PDX mimicked the cystic phenotype typical of some RCC. The negligible hypointense (necrotic) areas of PDX grown under the highly vascularized renal capsule are beneficial for preclinical studies. Mean apparent diffusion coefficient (ADC) values were equivalent to those of ccRCC in human patients. Hyperpolarized (HP) [1-13C]pyruvate MRI of PDX showed high glycolytic activity typical of high-grade primary and metastatic ccRCC with considerable intra- and inter-tumoral variability, as has been observed in clinical HP MRI of ccRCC. Comparison of steady state metabolite concentrations and metabolic flux in [U-13C]glucose-labeled tumors highlighted the distinctive phenotypes of two PDX with elevated levels of numerous metabolites and increased fractional enrichment of lactate and/or glutamate, capturing the metabolic heterogeneity of glycolysis and the TCA cycle in clinical ccRCC. Culturing PDX cells and reimplanting to generate xenografts (XEN), or passaging PDX in vivo, altered some imaging and metabolic characteristics while transcription remained like that of the original PDX. These findings show that PDX are realistic models of ccRCC for imaging and metabolic studies but that the plasticity of metabolism must be considered when manipulating PDX for preclinical studies.

Dietary Patterns and Risk of Gleason Grade Progression among Men on Active Surveillance for Prostate Cancer: Results from the Canary Prostate Active Surveillance Study

Modifiable lifestyle factors, such as following a healthy dietary pattern may delay or prevent prostate cancer (PCa) progression. However, few studies have evaluated whether following specific dietary patterns after PCa diagnosis impacts risk of disease progression among men with localized PCa managed by active surveillance (AS). 564 men enrolled in the Canary Prostate Active Surveillance Study, a protocol-driven AS study utilizing a pre-specified prostate-specific antigen monitoring and surveillance biopsy regimen, completed a food frequency questionnaire (FFQ) at enrollment and had ≥ 1 surveillance biopsy during follow-up. FFQs were used to evaluate adherence to the Dietary Guidelines for Americans (Healthy Eating index (HEI))-2015, alternative Mediterranean Diet (aMED), and Dietary Approaches to Stop Hypertension (DASH) dietary patterns. Multivariable-adjusted hazards ratios (HRs) and 95% confidence intervals were estimated using Cox proportional hazards models. During a median follow-up of 7.8 years, 237 men experienced an increase in Gleason score on subsequent biopsy (grade reclassification). Higher HEI-2015, aMED or DASH diet scores after diagnosis were not associated with significant reductions in the risk of grade reclassification during AS. However, these dietary patterns have well-established protective effects on chronic diseases and mortality and remain a prudent choice for men with prostate cancer managed by AS.

Impact of Prostate Health Index Results for Prediction of Biopsy Grade Reclassification During Active Surveillance

PURPOSE

We assessed whether Prostate Health Index results improve prediction of grade reclassification for men on active surveillance.

METHODS AND MATERIALS

We identified men in Canary Prostate Active Surveillance Study with Grade Group 1 cancer. Outcome was grade reclassification to Grade Group 2+ cancer. We considered decision rules to maximize specificity with sensitivity set at 95%. We derived rules based on clinical data (R1) vs clinical data+Prostate Health Index (R3). We considered an "or"-logic rule combining clinical score and Prostate Health Index (R4), and a "2-step" rule using clinical data followed by risk stratification based on Prostate Health Index (R2). Rules were applied to a validation set, where values of R2-R4 vs R1 for specificity and sensitivity were evaluated.

RESULTS

We included 1,532 biopsies (n = 610 discovery; n = 922 validation) among 1,142 men. Grade reclassification was seen in 27% of biopsies (23% discovery, 29% validation). Among the discovery set, at 95% sensitivity, R2 yielded highest specificity at 27% vs 17% for R1. In the validation set, R3 had best performance vs R1 with Δsensitivity = -4% and Δspecificity = +6%. There was slight improvement for R3 vs R1 for confirmatory biopsy (AUC 0.745 vs R1 0.724, ΔAUC 0.021, 95% CI 0.002-0.041) but not for subsequent biopsies (ΔAUC -0.012, 95% CI -0.031-0.006). R3 did not have better discrimination vs R1 among the biopsy cohort overall (ΔAUC 0.007, 95% CI -0.007-0.020).

CONCLUSIONS

Among active surveillance patients, using Prostate Health Index with clinical data modestly improved prediction of grade reclassification on confirmatory biopsy and did not improve prediction on subsequent biopsies.

Measuring the multifaceted roles of mucin-domain glycoproteins in cancer

Mucin-domain glycoproteins are highly O-glycosylated cell surface and secreted proteins that serve as both biochemical and biophysical modulators. Aberrant expression and glycosylation of mucins are known hallmarks in numerous malignancies, yet mucin-domain glycoproteins remain enigmatic in the broad landscape of cancer glycobiology. Here we review the multifaceted roles of mucins in cancer through the lens of the analytical and biochemical methods used to study them. We also describe a collection of emerging tools that are specifically equipped to characterize mucin-domain glycoproteins in complex biological backgrounds. These approaches are poised to further elucidate how mucin biology can be understood and subsequently targeted for the next generation of cancer therapeutics.

Analysis of separate training and validation radical prostatectomy cohorts identifies 0.25 mm diameter as an optimal definition for "large" cribriform prostatic adenocarcinoma

Cribriform growth pattern is well-established as an adverse pathologic feature in prostate cancer. The literature suggests "large" cribriform glands associate with aggressive behavior; however, published studies use varying definitions for "large". We aimed to identify an outcome-based quantitative cut-off for "large" vs "small" cribriform glands. We conducted an initial training phase using the tissue microarray based Canary retrospective radical prostatectomy cohort. Of 1287 patients analyzed, cribriform growth was observed in 307 (24%). Using Kaplan-Meier estimates of recurrence-free survival curves (RFS) that were stratified by cribriform gland size, we identified 0.25 mm as the optimal cutoff to identify more aggressive disease. In univariable and multivariable Cox proportional hazard analyses, size >0.25 mm was a significant predictor of worse RFS compared to patients with cribriform glands ≤0.25 mm, independent of pre-operative PSA, grade, stage and margin status (p < 0.001). In addition, two different subset analyses of low-intermediate risk cases (cases with Gleason score ≤ 3 + 4 = 7; and cases with Gleason score = 3 + 4 = 7/4 + 3 = 7) likewise demonstrated patients with largest cribriform diameter >0.25 mm had a significantly lower RFS relative to patients with cribriform glands ≤0.25 mm (each subset p = 0.004). Furthermore, there was no significant difference in outcomes between patients with cribriform glands ≤ 0.25 mm and patients without cribriform glands. The >0.25 mm cut-off was validated as statistically significant in a separate 419 patient, completely embedded whole-section radical prostatectomy cohort by biochemical recurrence, metastasis-free survival, and disease specific death, even when cases with admixed Gleason pattern 5 carcinoma were excluded. In summary, our findings support reporting cribriform gland size and identify 0.25 mm as an optimal outcome-based quantitative measure for defining "large" cribriform glands. Moreover, cribriform glands >0.25 mm are associated with potential for metastatic disease independent of Gleason pattern 5 adenocarcinoma.

Expanding the Secondary Use of Prostate Cancer Real World Data: Automated Classifiers for Clinical and Pathological Stage

Background

Explicit documentation of stage is an endorsed quality metric by the National Quality Forum. Clinical and pathological cancer staging is inconsistently recorded within clinical narratives but can be derived from text in the Electronic Health Record (EHR). To address this need, we developed a Natural Language Processing (NLP) solution for extraction of clinical and pathological TNM stages from the clinical notes in prostate cancer patients.

Methods

Data for patients diagnosed with prostate cancer between 2010 and 2018 were collected from a tertiary care academic healthcare system's EHR records in the United States. This system is linked to the California Cancer Registry, and contains data on diagnosis, histology, cancer stage, treatment and outcomes. A randomly selected sample of patients were manually annotated for stage to establish the ground truth for training and validating the NLP methods. For each patient, a vector representation of clinical text (written in English) was used to train a machine learning model alongside a rule-based model and compared with the ground truth.

Results

A total of 5,461 prostate cancer patients were identified in the clinical data warehouse and over 30% were missing stage information. Thirty-three to thirty-six percent of patients were missing a clinical stage and the models accurately imputed the stage in 21–32% of cases. Twenty-one percent had a missing pathological stage and using NLP 71% of missing T stages and 56% of missing N stages were imputed. For both clinical and pathological T and N stages, the rule-based NLP approach out-performed the ML approach with a minimum F1 score of 0.71 and 0.40, respectively. For clinical M stage the ML approach out-performed the rule-based model with a minimum F1 score of 0.79 and 0.88, respectively.

Conclusions

We developed an NLP pipeline to successfully extract clinical and pathological staging information from clinical narratives. Our results can serve as a proof of concept for using NLP to augment clinical and pathological stage reporting in cancer registries and EHRs to enhance the secondary use of these data.

Bridging the gap between prostate radiology and pathology through machine learning

Background

Prostate cancer remains the second deadliest cancer for American men despite clinical advancements. Currently, magnetic resonance imaging (MRI) is considered the most sensitive non‐invasive imaging modality that enables visualization, detection, and localization of prostate cancer, and is increasingly used to guide targeted biopsies for prostate cancer diagnosis. However, its utility remains limited due to high rates of false positives and false negatives as well as low inter‐reader agreements.

Purpose

Machine learning methods to detect and localize cancer on prostate MRI can help standardize radiologist interpretations. However, existing machine learning methods vary not only in model architecture, but also in the ground truth labeling strategies used for model training. We compare different labeling strategies and the effects they have on the performance of different machine learning models for prostate cancer detection on MRI.

Methods

Four different deep learning models (SPCNet, U‐Net, branched U‐Net, and DeepLabv3+) were trained to detect prostate cancer on MRI using 75 patients with radical prostatectomy, and evaluated using 40 patients with radical prostatectomy and 275 patients with targeted biopsy. Each deep learning model was trained with four different label types: pathology‐confirmed radiologist labels, pathologist labels on whole‐mount histopathology images, and lesion‐level and pixel‐level digital pathologist labels (previously validated deep learning algorithm on histopathology images to predict pixel‐level Gleason patterns) on whole‐mount histopathology images. The pathologist and digital pathologist labels (collectively referred to as pathology labels) were mapped onto pre‐operative MRI using an automated MRI‐histopathology registration platform.

Results

Radiologist labels missed cancers (ROC‐AUC: 0.75‐0.84), had lower lesion volumes (~68% of pathology lesions), and lower Dice overlaps (0.24‐0.28) when compared with pathology labels. Consequently, machine learning models trained with radiologist labels also showed inferior performance compared to models trained with pathology labels. Digital pathologist labels showed high concordance with pathologist labels of cancer (lesion ROC‐AUC: 0.97‐1, lesion Dice: 0.75‐0.93). Machine learning models trained with digital pathologist labels had the highest lesion detection rates in the radical prostatectomy cohort (aggressive lesion ROC‐AUC: 0.91‐0.94), and had generalizable and comparable performance to pathologist label‐trained‐models in the targeted biopsy cohort (aggressive lesion ROC‐AUC: 0.87‐0.88), irrespective of the deep learning architecture. Moreover, machine learning models trained with pixel‐level digital pathologist labels were able to selectively identify aggressive and indolent cancer components in mixed lesions on MRI, which is not possible with any human‐annotated label type.

Conclusions

Machine learning models for prostate MRI interpretation that are trained with digital pathologist labels showed higher or comparable performance with pathologist label‐trained models in both radical prostatectomy and targeted biopsy cohort. Digital pathologist labels can reduce challenges associated with human annotations, including labor, time, inter‐ and intra‐reader variability, and can help bridge the gap between prostate radiology and pathology by enabling the training of reliable machine learning models to detect and localize prostate cancer on MRI.

Germline mutations in penetrant cancer predisposition genes are rare in men with prostate cancer selecting active surveillance

BACKGROUND

Pathogenic germline mutations in several rare penetrant cancer predisposition genes are associated with an increased risk of aggressive prostate cancer (PC). Our objectives were to determine the prevalence of pathogenic germline mutations in men with low-risk PC on active surveillance, and assess whether pathogenic germline mutations associate with grade reclassification or adverse pathology, recurrence, or metastases, in men treated after initial surveillance.

METHODS

Men prospectively enrolled in the Canary Prostate Active Surveillance Study (PASS) were retrospectively sampled for the study. Germline DNA was sequenced utilizing a hereditary cancer gene panel. Mutations were classified according to the American College of Clinical Genetics and Genomics' guidelines. The association of pathogenic germline mutations with grade reclassification and adverse characteristics was evaluated by weighted Cox proportional hazards modeling and conditional logistic regression, respectively.

RESULTS

Overall, 29 of 437 (6.6%) study participants harbored a pathogenic germline mutation of which 19 occurred in a gene involved in DNA repair (4.3%). Eight participants (1.8%) had pathogenic germline mutations in three genes associated with aggressive PC: ATM, BRCA1, and BRCA2. The presence of pathogenic germline mutations in DNA repair genes did not associate with adverse characteristics (univariate analysis HR = 0.87, 95% CI: 0.36-2.06, p = 0.7). The carrier rates of pathogenic germline mutations in ATM, BRCA1, and BRCA2did not differ in men with or without grade reclassification (1.9% vs. 1.8%).

CONCLUSION

The frequency of pathogenic germline mutations in penetrant cancer predisposition genes is extremely low in men with PC undergoing active surveillance and pathogenic germline mutations had no apparent association with grade reclassification or adverse characteristics.

Development and validation of a quantitative reactive stroma biomarker (qRS) for prostate cancer prognosis

To develop and validate a new tissue-based biomarker that improves prediction of outcomes in localized prostate cancer by quantifying the host response to tumor. We use digital image analysis and machine learning to develop a biomarker of the prostate stroma called quantitative reactive stroma (qRS). qRS is a measure of percentage tumor area with a distinct, reactive stromal architecture. Kaplan Meier analysis was used to determine survival in a large retrospective cohort of radical prostatectomy samples. qRS was validated in two additional, distinct cohorts that include international cases and tissue from both radical prostatectomy and biopsy specimens. In the developmental cohort (Baylor College of Medicine, n = 482), patients whose tumor had qRS > 34% had increased risk of prostate cancer-specific death (HR 2.94; p = 0.039). This result was replicated in two validation cohorts, where patients with qRS > 34% had increased risk of prostate cancer-specific death (MEDVAMC; n = 332; HR 2.64; p = 0.02) and also biochemical recurrence (Canary; n = 988; HR 1.51; p = 0.001). By multivariate analysis, these associations were shown to hold independent predictive value when compared to currently used clinicopathologic factors including Gleason score and PSA. qRS is a new, validated biomarker that predicts prostate cancer death and biochemical recurrence across three distinct cohorts. It measures host-response rather than tumor-based characteristics, and provides information not represented by standard prognostic measurements.

Protein signatures to distinguish aggressive from indolent prostate cancer

BACKGROUND

Distinguishing men with aggressive from indolent prostate cancer is critical to decisions in the management of clinically localized prostate cancer. Molecular signatures of aggressive disease could help men overcome this major clinical challenge by reducing unnecessary treatment and allowing more appropriate treatment of aggressive disease.

METHODS

We performed a mass spectrometry-based proteomic analysis of normal and malignant prostate tissues from 22 men who underwent surgery for prostate cancer. Prostate cancer samples included Grade Groups (3-5), with 8 patients experiencing recurrence and 14 without evidence of recurrence with a mean of 6.8 years of follow-up. To better understand the biological pathways underlying prostate cancer aggressiveness, we performed a systems biology analysis and gene enrichment analysis. Proteins that distinguished recurrent from nonrecurrent cancer were chosen for validation by immunohistochemical analysis on tissue microarrays containing samples from a larger cohort of patients with recurrent and nonrecurrent prostate cancer.

RESULTS

In all, 24,037 unique peptides (false discovery rate < 1%) corresponding to 3,313 distinct proteins were identified with absolute abundance ranges spanning seven orders of magnitude. Of these proteins, 115 showed significantly (p < 0.01) different levels in tissues from recurrent versus nonrecurrent cancers. Analysis of all differentially expressed proteins in recurrent and nonrecurrent cases identified several protein networks, most prominently one in which approximately 24% of the proteins in the network were regulated by the YY1 transcription factor (adjusted p < 0.001). Strong immunohistochemical staining levels of three differentially expressed proteins, POSTN, CALR, and CTSD, on a tissue microarray validated their association with shorter patient survival.

CONCLUSIONS

The protein signatures identified could improve understanding of the molecular drivers of aggressive prostate cancer and be used as candidate prognostic biomarkers.

SU086, an inhibitor of HSP90, impairs glycolysis and represents a treatment strategy for advanced prostate cancer

Among men, prostate cancer is the second leading cause of cancer-associated mortality, with advanced disease remaining a major clinical challenge. We describe a small molecule, SU086, as a therapeutic strategy for advanced prostate cancer. We demonstrate that SU086 inhibits the growth of prostate cancer cells in vitro, cell-line and patient-derived xenografts in vivo, and ex vivo prostate cancer patient specimens. Furthermore, SU086 in combination with standard of care second-generation anti-androgen therapies displays increased impairment of prostate cancer cell and tumor growth in vitro and in vivo. Cellular thermal shift assay reveals that SU086 binds to heat shock protein 90 (HSP90) and leads to a decrease in HSP90 levels. Proteomic profiling demonstrates that SU086 binds to and decreases HSP90. Metabolomic profiling reveals that SU086 leads to perturbation of glycolysis. Our study identifies SU086 as a treatment for advanced prostate cancer as a single agent or when combined with second-generation anti-androgens.

•

SU086 inhibits prostate cancer growth in preclinical models of prostate cancer

•

SU086 targets heat shock protein 90

•

SU086 alters prostate cancer glycolysis and decreases intratumoral metabolism

•

SU086 in combination with anti-androgens halts prostate cancer growth

Treatment in the absence of disease reclassification among men on active surveillance for prostate cancer

BACKGROUND

Maintaining men on active surveillance for prostate cancer can be challenging. Although most men who eventually undergo treatment have experienced clinical progression, a smaller subset elects treatment in the absence of disease reclassification. This study sought to understand factors associated with treatment in a large, contemporary, prospective cohort.

METHODS

This study identified 1789 men in the Canary Prostate Cancer Active Surveillance Study cohort enrolled as of 2020 with a median follow-up of 5.6 years. Clinical and demographic data as well as information on patient-reported quality of life and urinary symptoms were used in multivariable Cox proportional hazards regression models to identify factors associated with the time to treatment RESULTS: Within 4 years of their diagnosis, 33% of men (95% confidence interval [CI], 30%-35%) underwent treatment, and 10% (95% CI, 9%-12%) were treated in the absence of reclassification. The most significant factor associated with any treatment was an increasing Gleason grade group (adjusted hazard ratio [aHR], 14.5; 95% CI, 11.7-17.9). Urinary quality-of-life scores were associated with treatment without reclassification (aHR comparing "mostly dissatisfied/terrible" with "pleased/mixed," 2.65; 95% CI, 1.54-4.59). In a subset analysis (n = 692), married men, compared with single men, were more likely to undergo treatment in the absence of reclassification (aHR, 2.63; 95% CI, 1.04-6.66).

CONCLUSIONS

A substantial number of men with prostate cancer undergo treatment in the absence of clinical changes in their cancers, and quality-of-life changes and marital status may be important factors in these decisions.

LAY SUMMARY

This analysis of men on active surveillance for prostate cancer shows that approximately 1 in 10 men will decide to be treated within 4 years of their diagnosis even if their cancer is stable. These choices may be related in part to quality-or-life or spousal concerns.

A review of artificial intelligence in prostate cancer detection on imaging

A multitude of studies have explored the role of artificial intelligence (AI) in providing diagnostic support to radiologists, pathologists, and urologists in prostate cancer detection, risk-stratification, and management. This review provides a comprehensive overview of relevant literature regarding the use of AI models in (1) detecting prostate cancer on radiology images (magnetic resonance and ultrasound imaging), (2) detecting prostate cancer on histopathology images of prostate biopsy tissue, and (3) assisting in supporting tasks for prostate cancer detection (prostate gland segmentation, MRI-histopathology registration, MRI-ultrasound registration). We discuss both the potential of these AI models to assist in the clinical workflow of prostate cancer diagnosis, as well as the current limitations including variability in training data sets, algorithms, and evaluation criteria. We also discuss ongoing challenges and what is needed to bridge the gap between academic research on AI for prostate cancer and commercial solutions that improve routine clinical care.

Machine Learning Applied to Electronic Health Records: Identification of Chemotherapy Patients at High Risk for Preventable Emergency Department Visits and Hospital Admissions

PURPOSE

Acute care use (ACU) is a major driver of oncologic costs and is penalized by a Centers for Medicare & Medicaid Services quality measure, OP-35. Targeted interventions reduce preventable ACU; however, identifying which patients might benefit remains challenging. Prior predictive models have made use of a limited subset of the data in the electronic health record (EHR). We aimed to predict risk of preventable ACU after starting chemotherapy using machine learning (ML) algorithms trained on comprehensive EHR data.

METHODS

Chemotherapy patients treated at an academic institution and affiliated community care sites between January 2013 and July 2019 who met inclusion criteria for OP-35 were identified. Preventable ACU was defined using OP-35 criteria. Structured EHR data generated before chemotherapy treatment were obtained. ML models were trained to predict risk for ACU after starting chemotherapy using 80% of the cohort. The remaining 20% were used to test model performance by the area under the receiver operator curve.

RESULTS

Eight thousand four hundred thirty-nine patients were included, of whom 35% had preventable ACU within 180 days of starting chemotherapy. Our primary model classified patients at risk for preventable ACU with an area under the receiver operator curve of 0.783 (95% CI, 0.761 to 0.806). Performance was better for identifying admissions than emergency department visits. Key variables included prior hospitalizations, cancer stage, race, laboratory values, and a diagnosis of depression. Analyses showed limited benefit from including patient-reported outcome data and indicated inequities in outcomes and risk modeling for Black and Medicaid patients.

CONCLUSION

Dense EHR data can identify patients at risk for ACU using ML with promising accuracy. These models have potential to improve cancer care outcomes, patient experience, and costs by allowing for targeted, preventative interventions.

Clinical laboratory tests associated with survival in patients with metastatic renal cell carcinoma: A Laboratory Wide Association Study (LWAS)

BACKGROUND

Prognostic models for patients with metastatic renal cell carcinoma (mRCC) include select laboratory values. These models have important limitations, including reliance on a limited array of laboratory tests, and use of dichotomous ("high-low") cutoffs. We applied a Laboratory-Wide Association Study (LWAS) framework to systematically evaluate common clinical laboratory results associated with survival for patients diagnosed with mRCC.

METHODS

We used laboratory data for 3,385 patients diagnosed with mRCC from 2002 to 2017. We developed a LWAS framework, to examine the association with 53 common clinical laboratory tests results (641,712 measurements) and overall survival. We employed false-discovery rate to test the association of multiple laboratory tests with survival, and validated these results using 3 separate cohorts to generate a standardized hazard ratio (sHR), reported for a 1 standard deviation unit change in each laboratory test.

RESULTS

The LWAS approach confirmed the association of laboratory values currently used in prognostic models with survival, including calcium (HR 1.35, 95%CI 1.24-1.48), leukocyte count (HR 1.40, 95%CI 1.30-1.51), platelet count (HR 1.36, 95%CI 1.27-1.51), and hemoglobin (HR 0.79, 95%CI 0.72-0.86). Use of these tests as continuous variables improved model performance. LWAS also identified acute phase reactants associated with survival not typically included in prognostic models, including serum albumin (HR 0.66, 95%CI 0.61-0.72), ferritin (HR 1.25, 95%CI 1.08-1.45), alkaline phosphatase (HR 1.31, 95%CI 1.23-1.40), and C-reactive protein (HR 1.70, 95%CI 1.14-2.53).

CONCLUSIONS

Routinely measured laboratory tests can refine current prognostic models, facilitate comparisons across clinical trial cohorts, and match patients with specific systemic therapies.

Diverse patient trajectories during cytotoxic chemotherapy: Capturing longitudinal patient-reported outcomes

Background

High‐value cancer care balances effective treatment with preservation of quality of life. Chemotherapy is known to affect patients’ physical and psychological well‐being negatively. Patient‐reported outcomes (PROs) provide a means to monitor declines in a patients’ well‐being during treatment.

Methods

We identified 741 oncology patients undergoing chemotherapy in our electronic health record (EHR) system who completed Patient‐Reported Outcomes Measurement Information System (PROMIS) surveys during treatment at a comprehensive cancer center, 2013–2018. PROMIS surveys were collected before, during, and after chemotherapy treatment. Linear mixed‐effects models were performed to identify predictors of physical and mental health scores over time. A k‐mean cluster analysis was used to group patient PROMIS score trajectories.

Results

Mean global physical health (GPH) scores were 48.7 (SD 9.3), 47.7 (8.8), and 48.6 (8.9) and global mental health (GMH) scores were 50.4 (8.6), 49.5 (8.8), and 50.6 (9.1) before, during, and after chemotherapy, respectively. Asian race, Hispanic ethnicity, public insurance, anxiety/depression, stage III cancer, and palliative care were predictors of GPH and GMH decline. The treatment time period was also a predictor of both GPH and GMH decline relative to pre‐treatment. Trajectory clustering identified four distinct PRO clusters associated with chemotherapy treatment.

Conclusions

Patient‐reported outcomes are increasingly used to help monitor cancer treatment and are now a part of care reimbursement. This study leveraged routinely collected PROMIS surveys linked to EHRs to identify novel patient trajectories of physical and mental well‐being in oncology patients undergoing chemotherapy and potential predictors. Supportive care interventions in high‐risk populations identified by our study may optimize resource deployment.

Novelty and impact

This study leveraged routinely collected patient‐reported outcome (PROMIS) surveys linked to electronic health records to characterize oncology patients’ quality of life during chemotherapy. Important clinical and demographic predictors of declines in quality of life were identified and four novel trajectories to guide personalized interventions and support. This work highlights the utility of monitoring patient‐reported outcomes not only before and after, but during chemotherapy to help advert adverse patient outcomes and improve treatment adherence.

Abstract 81: Trop2 regulates prostate cancer growth and metastasis through distinct molecular mechanisms

Background: The first line of treatment for men with advanced prostate cancer is androgen deprivation therapy. Although initial responses are observed, prostate cancer commonly relapses in its lethal metastatic form referred to as castration resistant prostate cancer (CRPC) with 1-2 years mean survival time. Neuroendocrine prostate cancer (NEPC) is highly aggressive, AR independent subtype and usually emerges post castration resistance. We recently identified that the cell surface receptor, Trop2, is a new driver of NEPC. Moreover, we demonstrated that Trop2 regulates prostate cancer growth and metastasis. In this study, we set out to delineate the molecular mechanisms through which Trop2 regulates prostate cancer growth and metastasis.

Methods: Proximity-dependent Biotin Identification (BioID) followed by mass spectrometry was performed to identify Trop2 interactome. Lentiviral infection was used to generate prostate cancer cell lines with over-expression of Trop2 and knock-down of NOTCH1, SLC4A7, PLEC, and OCLN shRNA to modulate gene expression levels. In vitro functional assays were performed including colony formation, and Matrigel drop 3D cell invasion assays.

Results: The Trop2 membrane interactome was identified utilizing Proximity-dependent Biotin Identification (BioID) in living cells and uncovered that Trop2-mediated prostate cancer growth and metastasis are orchestrated by distinct downstream pathways including Notch signaling (NOTCH1), control of intracellular pH (SLC4A7), exosome secretion (PLEC), and tight junctions (OCLN). Interaction of Trop2 binding partners including NOTCH1, SLC4A7, PLEC and OCLN with Trop2 were further validated by fluorescence resonance energy transfer (FRET) using confocal microscopy. Moreover, knocking down the Trop2 interacting partners in Trop2 over-expressing prostate cancer cells suppressed Trop2-driven growth and invasion ability.

Conclusions: In a previous study, we identified cell surface receptor, Trop2, as a novel driver of metastatic NEPC. Herein, our new findings reveal that Trop2 interacts with NOTCH1, SLC4A7, PLEC, and OCLN, which may highlight novel biological functions of Trop2 in prostate tumorigenesis and provide new understanding of the potential mechanism of neuroendocrine differentiation and metastasis to provide new therapeutic strategy for metastatic CRPC with neuroendocrine features.

Citation Format: Tanya Ivanova Stoyanova, En-Chi Hsu, Shiqin Liu, Fernando Jose Garcia Marques, Abel Bermudez, Merve Aslan, Michelle Shen, Sharon Pitteri, James D. Brooks. Trop2 regulates prostate cancer growth and metastasis through distinct molecular mechanisms [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 81.

Consumption of cruciferous vegetables and the risk of bladder cancer in a prospective US cohort: data from the NIH-AARP diet and health study

BACKGROUND

Abundant pre-clinical data suggest that consumption of cruciferous vegetables might protect against bladder cancer. While small-scale clinical evidence supports this hypothesis, population-level data is lacking. We tested the hypothesis that consumption of cruciferous vegetables is associated with a lower risk of bladder cancer in a large population-based study.

METHODS

We investigated the association between dietary consumption of cruciferous vegetables and the risk of bladder cancer in the NIH-American Association of Retired Persons (AARP) Diet and Health Study. Diet at baseline was collected with self-administered food-frequency questionnaires. Bladder cancer diagnoses were identified through linkage with state cancer registries. Hazard ratio (HR) and 95% confidence intervals (CI) were estimated with Cox proportional hazards models.

RESULTS

Our analysis included 515,628 individuals. Higher intake of cruciferous vegetables, both overall and when stratified by variety (broccoli vs. brussels sprouts vs. cauliflower), were not associated with bladder cancer risk for men or women. A history of smoking did not affect the results.

CONCLUSIONS

Our study shows no association between dietary consumption of cruciferous vegetables and incident bladder cancer.

Using an Automated Electronic Health Record Score To Estimate Life Expectancy In Men Diagnosed With Prostate Cancer In The Veterans Health Administration

OBJECTIVES

To determine if an automatically calculated electronic health record score can estimate intermediate-term life expectancy in men with prostate cancer to provide guideline concordant care.

METHODS

We identified all men (n = 36,591) diagnosed with prostate cancer in 2013-2015 in the VHA. Of the 36,591, 35,364 (96.6%) had an available Care Assessment Needs (CAN) score (range: 0-99) automatically calculated in the 30 days prior to the date of diagnosis. It was designed to estimate short-term risks of hospitalization and mortality. We fit unadjusted and multivariable Cox proportional hazards regression models to determine the association between the CAN score and overall survival among men with prostate cancer. We compared CAN score performance to two established comorbidity measures: The Charlson Comorbidity Index and Prostate Cancer Comorbidity Index (PCCI).

RESULTS

Among 35,364 men, the CAN score correlated with overall stage, with mean scores of 46.5 ( ± 22.4), 58.0 ( ± 24.4), and 68.1 ( ± 24.3) in localized, locally advanced, and metastatic disease, respectively. In both unadjusted and adjusted models for prostate cancer risk, the CAN score was independently associated with survival (HR = 1.23 95%CI 1.22-1.24 & adjusted HR = 1.17 95%CI 1.16-1.18 per 5-unit change, respectively). The CAN score (overall C-Index 0.74) yielded better discrimination (AUC = 0.76) than PCCI (AUC = 0.65) or Charlson Comorbidity Index (AUC = 0.66) for 5-year survival.

CONCLUSION

The CAN score is strongly associated with intermediate-term survival following a prostate cancer diagnosis. The CAN score is an example of how learning health care systems can implement multi-dimensional tools to provide fully automated life expectancy estimates to facilitate patient-centered cancer care.

Identification of patients at high risk for preventable emergency department visits and inpatient admissions after starting chemotherapy: Machine learning applied to comprehensive electronic health record data

1511

Background: Acute care use is one of the largest drivers of cancer care costs. OP-35: Admissions and Emergency Department Visits for Patients Receiving Outpatient Chemotherapy is a CMS quality measure that will affect reimbursement based on unplanned inpatient admissions (IP) and emergency department (ED) visits. Targeted measures can reduce preventable acute care use but identifying which patients might benefit remains challenging. Prior predictive models have made use of a limited subset of the data available in the Electronic Health Record (EHR). We hypothesized dense, structured EHR data could be used to train machine learning algorithms to predict risk of preventable ED and IP visits. Methods: Patients treated at Stanford Health Care and affiliated community care sites between 2013 and 2015 who met inclusion criteria for OP-35 were selected from our EHR. Preventable ED or IP visits were identified using OP-35 criteria. Demographic, diagnosis, procedure, medication, laboratory, vital sign, and healthcare utilization data generated prior to chemotherapy treatment were obtained. A random split of 80% of the cohort was used to train a logistic regression with least absolute shrinkage and selection operator regularization (LASSO) model to predict risk for acute care events within the first 180 days of chemotherapy. The remaining 20% were used to measure model performance by the Area Under the Receiver Operator Curve (AUROC). Results: 8,439 patients were included, of whom 35% had one or more preventable event within 180 days of starting chemotherapy. Our LASSO model classified patients at risk for preventable ED or IP visits with an AUROC of 0.783 (95% CI: 0.761-0.806). Model performance was better for identifying risk for IP visits than ED visits. LASSO selected 125 of 760 possible features to use when classifying patients. These included prior acute care visits, cancer stage, race, laboratory values, and a diagnosis of depression. Key features for the model are shown in the table. Conclusions: Machine learning models trained on a large number of routinely collected clinical variables can identify patients at risk for acute care events with promising accuracy. These models have the potential to improve cancer care outcomes, patient experience, and costs by allowing for targeted preventative interventions. Future work will include prospective and external validation in other healthcare systems.[Table: see text]