Search

Objective To compare Prostate Health Index (PHI) and prostate‐specific antigen (PSA) density as secondary tests after multiparametric magnetic resonance imaging (mpMRI) in improving the detection accuracy of Gleason grade group (GG) 2‐5 prostate cancer (PCa) and in decreasing unnecessary biopsies in a multiethnic biopsy‐naïve population. Methods From February 2017 to February 2020, we recruited consecutive biopsy‐naïve men in participating urology clinics for elevated PSA levels. They all had a PHI score, mpMRI, and prostate biopsy. Experienced genitourinary radiologists read all mpMRI studies based on PIRADS version 2.0. Logistic regression models were used to generate receiver operating characteristic curves. Models were tested for effect modification between Race (Black vs White) and both PHI and PSA density, and Race and PIRADS to determine if race impacted their prediction accuracy. Sensitivity, specificity, and predictive values of PHI and PSA density thresholds were calculated by PIRADS scores. The primary outcome was GG2‐5 PCa, that is, Gleason score ≥3 + 4. Results The study included 143 men, of which 65 (45.5%) were self‐reported Black. Median age was 62.0 years and 55 men (38.4%) had GG2‐5 PCa. Overall, 18.1% had PIRADS 1‐2, 32.9% had PIRADS 3, and 49.0% had PIRADS 4‐5. For the binary logistic regressions, the interactions between PIRADS and Race (P = .08), Log (PHI) and Race (P = .17), and Log (PSA density) and Race (P = .42) were not statistically significant. Within PIRADS 3 lesions, a PHI ≥49 prevented unnecessary biopsies in 55% of men and missed no GG2‐5 PCa, yielding a negative predictive value of 100%. There was no reliable PHI or PSA density threshold to avoid PCa biopsies in PIRADS 1‐2 or 4‐5. Conclusions PHI and PSA density can be used after mpMRI to improve the detection of GG2‐5 PCa in a biopsy‐naïve cohort. PHI may be superior to PSA density in PIRADS 3 lesions by avoiding 55% of unnecessary biopsies. Using both PHI and PSA density in series may further increase specificity and lead to fewer unnecessary biopsies, but further larger studies are warranted to determine the optimal threshold of each biomarker.

Deep neural networks (DNNs) that predict mutational status from H&E slides of cancers can enable inexpensive and timely precision oncology. Although expert knowledge is reliable for annotating regions informative of malignancy and other known histologic patterns (strong supervision), it is unreliable for identifying regions informative of mutational status. This poses a serious impediment to obtaining higher prognostic accuracy and discovering new knowledge of pathobiology. We used a weakly supervised learning technique to train a DNN to predict BRAF V600E mutational status, determined using DNA testing, in H&E-stained images of thyroid cancer tissue without regional annotations. Our discovery cohort was a tissue microarray of only 85 patients from a single hospital. On a large independent external cohort of 444 patients from other hospitals, the trained model gave an area under the receiver operating characteristic curve of 0.98 (95% CI 0.97-1.00), which is much higher than the previously reported results for detecting any mutation using H&E by DNNs trained using strong supervision. We also developed a visualization technique that can automatically highlight regions the DNN found most informative for predicting mutational status. Our visualization is spatially granular and highly specific in highlighting strong negative and positive regions and moves us toward explainable artificial intelligence. Using t-tests, we confirmed that the proportions of follicular or papillary histology and oncocytic cytology, as noted for each patient by a pathologist who was blinded to the mutational status, were significantly different between mutated and wildtype patients. However, based solely on these features noted by the pathologist, a logistic regression classifier gave an average area under the receiver operating characteristic curve of 0.78 in five-fold cross-validation, which is much lower than that obtained using the DNN. These results highlight the potential of weakly supervised learning for training DNN models for problems where the informative visual patterns and their locations are not known a priori. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

PURPOSE The Genomic Prostate Score (GPS), performed on biopsy tissue, predicts adverse outcome in prostate cancer (PCa) and has shown promise for improving patient selection for active surveillance (AS). However, its impact on treatment choice in high-risk populations of African Americans is largely unknown and, in general, the effect of the GPS on this difficult decision has not been evaluated in randomized trials. METHODS Two hundred men with National Comprehensive Cancer Network very low to low-intermediate PCa from three Chicago hospitals (70% Black, 16% college graduates) were randomly assigned at diagnosis to standard counseling with or without a 12-gene GPS assay. The primary end point was treatment choice at a second postdiagnosis visit. The proportion of patients choosing AS was compared, and multivariable modeling was used to estimate the effects of various factors on AS acceptance. RESULTS AS acceptance was high overall, although marginally lower in the intervention group (77% v 88%; P = .067), and lower still when men with inadequate specimens were excluded ( P = .029). Men with lower health literacy who received a GPS were seven-fold less likely to choose AS compared with controls, whereas no difference was seen in men with higher health literacy ( Pinteraction = .022). Among men with low-intermediate risk, 69% had GPS values consistent with unfavorable intermediate or high-risk cancer. AS choice was also independently associated with a family history of PCa and having health insurance. CONCLUSION In contrast to other studies, the net effect of the GPS was to move patients away from AS, primarily among men with low health literacy. These findings have implications for our understanding of how prognostic molecular assays that generate probabilities of poor outcome can affect treatment decisions in diverse clinical populations.

PURPOSE: The Prostate Health Index is validated for prostate cancer detection but has not been well validated for Gleason grade group 2-5 prostate cancer detection in Black men. We hypothesize that the Prostate Health Index has greater accuracy than prostate specific antigen for detection of Gleason grade group 2-5 prostate cancer. We estimated probability of overall and Gleason grade group 2-5 prostate cancer across previously established Prostate Health Index ranges and identified Prostate Health Index cutoffs that maximize specificity for Gleason grade group 2-5 prostate cancer with sensitivity >90%. MATERIALS AND METHODS: We recruited a “cancer-free” Black control cohort (135 patients) and a cohort of biopsy naïve Black men (158) biopsied for elevated prostate specific antigen. Descriptive statistics compared the prostate cancer cases and controls and the frequency of Gleason grade group 2-5 prostate cancer across Prostate Health Index scores. Receiver operating characteristics compared the discrimination of prostate specific antigen, Prostate Health Index and other prostate specific antigen related biomarkers. Sensitivity and specificity for Gleason grade group 2-5 prostate cancer detection were assessed at prostate specific antigen and Prostate Health Index thresholds alone and in series. RESULTS: Of biopsied subjects 32.9% had Gleason grade group 2-5 prostate cancer. In Blacks with prostate specific antigen from 4.0–10.0 ng/ml, Prostate Health Index and prostate specific antigen had similar discrimination for Gleason grade group 2-5 prostate cancer (0.63 vs 0.57, p=0.27). In Blacks with prostate specific antigen ≤10.0, a threshold of prostate specific antigen ≥4.0 had 90.4% sensitivity for Gleason grade group 2-5 prostate cancer; a threshold of prostate specific antigen ≥4.0 with Prostate Health Index ≥35.0 in series avoided unnecessary biopsy in 33.0% of men but missed 17.3% of Gleason grade group 2-5 prostate cancer. Prostate specific antigen ≥4.0 with Prostate Health Index ≥28.0 in series spared biopsy in 17.9%, while maintaining 90.4% sensitivity of Gleason grade group 2-5 prostate cancer. CONCLUSIONS: The Prostate Health Index has moderate accuracy in detecting Gleason grade group 2-5 prostate cancer in Blacks, but Prostate Health Index ≥28.0 can be safely used to avoid some unnecessary biopsies in Blacks.

Objective To validate the 17-gene Oncotype DX Genomic Prostate Score (GPS) as a predictor of adverse pathology (AP) in African American (AA) men and to assess the distribution of GPS in AA and European American (EA) men with localized prostate cancer. Methods The study populations were derived from 2 multi-institutional observational studies. Between February 2009 and September 2014, AA and EA men who elected immediate radical prostatectomy after a ≥10-core transrectal ultrasound biopsy were included in the study. Logistic regressions, area under the receiver operating characteristics curves (AUC), calibration curves, and predictive values were used to compare the accuracy of GPS. AP was defined as primary Gleason grade 4, presence of any Gleason pattern 5, and/or non-organ-confined disease (≥pT3aN0M0) at radical prostatectomy. Results Overall, 96 AA and 76 EA men were selected and 46 (26.7%) had AP. GPS result was a significant predictor of AP (odds ratio per 20 GPS units [OR/20 units] in AA: 4.58; 95% confidence interval (CI) 1.8-11.5, P = .001; and EA: 4.88; 95% CI 1.8-13.5, P = .002). On multivariate analysis, there was no significant interaction between GPS and race (P >.10). GPS remained significant in models adjusted for either National Comprehensive Cancer Network (NCCN) risk group or Cancer of the Prostate Risk Assessment (CAPRA) score. In race-stratified models, area under the receiver operating characteristics curves for GPS/20 units was 0.69 for AAs vs 0.74 for EAs (P = .79). The GPS distributions were not statistically different by race (all P >.05). Conclusion In this clinical validation study, the Oncotype DX GPS is an independent predictor of AP at prostatectomy in AA and EA men with similar predictive accuracy and distributions.

The Stockholm3 test improves Gleason Grade Group ≥2 (GG ≥ 2) prostate cancer (PC) detection, however it has not been evaluated in an American cohort where clinical practice patterns and ethnicity differ. We aimed to identify subgroups within a Stockholm population with PC risk profiles matching American ethnicity-specific subgroups and compare the detection of PC and describe Stockholm3 performance within these subgroups. All men age 49–70 years presenting for prostate biopsies were evaluated at UIC from 2016 to 2019, as well as men in Stockholm from 2012 to 2014 in the STHLM3 study. Propensity scores (PS) were estimated for each person using logistic regression for age, PSA, prostate volume, family history of PC, 5-alpha reductase inhibitor use, and prior biopsy. 3:1 PS matching was performed for Stockholm to Chicago ethnicity-specific cohorts and odds ratios (OR) were computed to compare detection of GG ≥ 2 PC between groups. 504 Chicago men and 6980 Stockholm men were included. In African American (AA) men, 51% had GG ≥ 2 PC detected, while in risk-matched Stockholm men, 34% had GG ≥ 2 PC detected (OR: 2.1, p

Vitamin D deficiency is associated with an increased risk of prostate cancer (PCa) mortality and is hypothesized to contribute to PCa aggressiveness and disparities in African American populations. The prostate epithelium was recently shown to express megalin, an endocytic receptor that internalizes globulin-bound hormones, which suggests regulation of prostate hormone levels, in contrast to the free hormone hypothesis. Here, we demonstrated that megalin imports testosterone bound to sex hormone-binding globulin into prostate cells. Prostatic loss of Lrp2 (megalin) in a mouse model resulted in reduced prostate testosterone and dihydrotestosterone (DHT) levels. Megalin expression was regulated and suppressed by 25-hydroxyvitamin D (25D) in cell lines, patient-derived prostate epithelial cells, and prostate tissue explants, indicating a negative feedback loop. In patent samples, the relationships between hormones support this feedback mechanism, as prostatic DHT levels are higher in African American men and are inversely correlated with serum 25D status. Megalin levels are reduced in localized PCa by the Gleason grade and in patients with future disease recurrence. Our findings suggest that the free hormone hypothesis should be revisited for testosterone and highlight the impact of vitamin D deficiency on prostate androgen levels, which are known drivers of PCa. Thus, we revealed a mechanistic link between vitamin D and PCa disparities observed in African Americans.

Purpose: Selective progesterone receptor modulators (SPRMs) show preclinical activity against hormone-sensitive breast cancer, but have not been tested in patients with early, treatment-naïve tumors. Patients and Methods: In a double-blind presurgical window trial of oral telapristone acetate (TPA) 12 mg daily versus placebo, 70 patients with early-stage breast cancer were randomized 1:1 (stratified by menopause) and treated for 2 to 10 weeks. The primary endpoint was change in Ki67 between diagnostic biopsy and surgical specimens. Gene expression pre- and posttherapy was assessed using RNA-sequencing and gene set enrichment analysis was performed to determine pathways enriched in response to TPA and placebo treatments. Results: Among 61 evaluable women (29 placebo and 32 telapristone acetate), 91% of tumors were ER/PR positive. The mean Ki67 declined by 5.5% in all women treated with telapristone acetate (P = 0.003), and by 4.2% in all women treated with placebo (P = 0.04). After menopausal stratification, the Ki67 decline remained significant in 22 telapristone acetate–treated premenopausal women (P = 0.03). Differential gene expression analysis showed no significant modulation overall. However, in a subset of tumors that demonstrated ≥30% relative reduction in Ki67 in the telapristone acetate group, genes related to cell-cycle progression, and those in the HER2 amplicon were significantly downregulated. In contrast, no significantly enriched pathways were identified in the placebo group. Conclusions: Patients treated with telapristone acetate whose Ki67 decreased by ≥30% demonstrated a selective antiproliferative signal, with a potentially important effect on HER2 amplicon genes. Evaluation of SPRMs in a neoadjuvant trial is merited, with attention to predictors of response to SPRM therapy, and inclusion of pre- and postmenopausal women.

Purpose Large cell neuroblastomas (LCN) are frequently seen in recurrent, high-risk neuroblastoma but are rare in primary tumors. LCN, characterized by large nuclei with prominent nucleoli, predict a poor prognosis. We hypothesize that LCN can be created with high-dose intra-tumoral chemotherapy and identified by a digital analysis system. Methods Orthotopic mouse xenografts were created using human neuroblastoma and treated with high-dose chemotherapy delivered locally via sustained-release silk platforms, inducing tumor remission. After recurrence, LCN populations were identified on H&E sections manually. Clusters of typical LCN and non-LCN cells were divided equally into training and test sets for digital analysis. Marker-controlled watershed segmentation was used to identify nuclei and characterize their features. Logistic regression was developed to distinguish LCN from non-LCN. Results Image analysis identified 15,000 nuclei and characterized 70 nuclear features. A 19-feature model provided AUC > 0.90 and 100% accuracy when > 30% nuclei/cluster were predicted as LCN. Overall accuracy was 87%. Conclusions We recreated LCN using high-dose chemotherapy and developed an automated method for defining LCN histologically. Features in the model provide insight into LCN nuclear phenotypic changes that may be related to increased activity. This model could be adapted to identify LCN in human tumors and correlated with clinical outcomes.

PURPOSE Although alterations of concentrations in circulating steroids have been linked to single nucleotide polymorphisms (SNPs) of steroidogenic enzymes, we hypothesized that SNPs of such enzymes located within the breast affect local steroid concentrations more than products of such SNPs absorbed from the circulation. METHODS Steroids (estradiol, estrone, testosterone, androstenedione, DHEA, DHEA sulfate, progesterone) in nipple aspirate fluid (NAF) were purified by HPLC and they along with serum steroids were quantified by immunoassays. Polymorphisms of the transporter SLCO2B1 and enzymes HSD3B1, CYP19A1, HSD17B12, AKR1C3, CYP1B1, and SRD5A1 were measured in white blood cell DNA. RESULTS Steroid concentrations in NAF of subjects with homozygous minor genotypes differed from those with heterozygotes, i.e., SLCO2B1 (rs2851069) decreased DHEAS (p = 0.04), HSD17B12 (rs11555762) increased estradiol (p

Aim: We examined methylation patterns with aggressive tumor phenotypes and investigated demographic, socioeconomic and reproductive predictors of gene methylation. Materials & methods: Pyrosequencing quantified methylation of BRCA1, EGFR, GSTM2, RASSF1, TFF1 and Sat 2. We used quantile regression models to calculate adjusted median methylation values by estrogen and progesterone receptor (ER/PR) status. Bivariate associations between participant characteristics and methylation were examined. Results: Higher percent methylation of GSTM2 was observed in ER/PR-negative compared with ER/PR-positive tumors in ductal carcinoma in situ (14 vs 2%) and invasive (35 vs 3%) tissue components. Trends in aberrant GSTM2 methylation across tissue components were stronger among ER/PR-negative tumors (p-interaction

Objective To determine whether a computer vision-based approach applied to haematoxylin and eosin (H&E) prostate biopsy images can distinguish dutasteride-treated tissue from placebo, and identify features associated with degree of responsiveness to 5α-reductase inhibitor (5ARI) therapy. Subjects and methods Our study population comprised 100 treatment-adherent men without prostate cancer assigned to dutasteride or placebo in the REDUCE trial, who had slides available from mandatory year-4 biopsies. Half of the men also provided slides from a year-2 biopsy. We obtained 20× whole-slide images and used specialized software to generate a library of 1 300 epithelial and stromal features from objects comprising superpixels and several types of nuclei, including spatial relations among objects between and within each hierarchical level. We used penalized logistic regression and fivefold cross-validation to find optimal combinations of histological features in the year-4 biopsies. Feature data from the year-2 biopsies were fitted to a final model for independent validation. Two pathologists, blinded to treatment, scored each image for focal atrophy and histological features previously linked to 5AR1 treatment. Results Consensus classification by pathologists obtained a discrimination accuracy equivalent to chance. A 21-feature computer vision model gave a cross-validation area under the curve of 0.97 (95% confidence interval [CI] 0.95-0.99) in the year-4 biopsies and 0.79 (95% CI: 0.65-0.92) in the set-aside year-2 biopsies. Histology scores were not correlated with change in prostate-specific antigen level, serum dihydrotestosterone level or gland volume. Key features associated with dutasteride treatment included greater shape and colour uniformity in stroma, irregular clustering of epithelial nuclei, and greater variation in lumen shape. Conclusion The present findings show that a computer vision approach can detect subtle histological effects attributable to dutasteride, resulting in a continuous measure of responsiveness to the drug that could eventually be used to predict individual patient response in the context of BPH treatment or cancer chemoprevention.

BackgroundPredictive models that take race into account like the Prostate Cancer Prevention Trial Risk Calculator 2.0 (PCPT RC) and the new Prostate Biopsy Collaborative Group (PBCG) RC have been developed to equitably mitigate the overdiagnosis of prostate specific antigen (PSA) screening. Few studies have compared the performance of both calculators across racial groups.MethodsFrom 1485 prospectively recruited participants, 954 men were identified undergoing initial prostate biopsy for abnormal PSA or digital rectal examination in five Chicago hospitals between 2009 and 2014. Discrimination, calibration, and frequency of avoided biopsies were calculated to assess the performance of both risk calculators.ResultsOf 954 participants, 463 (48.5%) were Black, 355 (37.2%) were White, and 136 (14.2%) identified as Other. Biopsy results were as follows: 310 (32.5%) exhibited no cancer, 323 (33.9%) indolent prostate cancer, and 321 (33.6%) clinically significant prostate cancer (csPCa). Differences in area under the curve (AUC)s for the detection of csPCa between PCPT and PBCG were not statistically different across all racial groups. PBCG did not improve calibration plots in Blacks and Others, as it showed higher levels of overprediction at most risk thresholds. PCPT led to an increased number of avoidable biopsies in minorities compared to PBCG at the 30% threshold (68% vs. 28% of all patients) with roughly similar rates of missed csPCa (23% vs. 20%).ConclusionSignificant improvements were noticed in PBCG’s calibrations and net benefits in Whites compared to PCPT. Since PBCG’s improvements in Blacks are disputable and potentially biases a greater number of low risk Black and Other men towards unnecessary biopsies, PCPT may lead to better biopsy decisions in racial minority groups. Further comparisons of commonly used risk calculators across racial groups is warranted to minimize excessive biopsies and overdiagnosis in ethnic minorities.

Background Fibrostenosis, the most serious eosinophilic esophagitis (EoE) complication, is mediated by epithelial-mesenchymal transition (EMT). Transitioned cells contribute to pathogenesis by overproducing extracellular matrix. Objective Our aim was to determine whether RPC4046 (anti‒IL-13 mAb) modulates EMT biomarkers in biopsy samples from adults with active EoE in a substudy of a double-blind, placebo-controlled phase 2 trial. Methods Baseline and week 16 esophageal biopsy samples were taken from 69 patients who were randomized to weekly treatment with subcutaneous RPC4046, 180 mg (n = 19), 360 mg (n = 26), or placebo (n = 24). Duplex immunofluorescence slides stained for E-cadherin and vimentin were digitally analyzed by mapping each epithelial cell and recording fluorescence intensities. End points included change from baseline to week 16 in percentage of vimentin-positive epithelial cells (primary), total E-cadherin expression, and vimentin-to–E-cadherin ratio per cell (an average of 47,000 cells per biopsy sample analyzed). Results The mean percentage of vimentin-positive cells decreased by 0.94%, 2.75%, and 4.24% in the placebo, low-dose, and high-dose groups, respectively (P =.032 for the high-dose vs placebo group). Mean E-cadherin expression per cell increased 5.6-fold in both dose groups versus in the placebo group (high-dose group P = .047). The increases in E-cadherin expression per cell from baseline to week 16 were correlated with improvements in histology, eosinophil counts, endoscopic findings, and symptoms. Conclusion RPC4046 significantly reduced EMT markers in adults with active EoE, with greater effects at 360 mg. Together with results for eosinophil density and clinical end points from the main trial, these data support the hypothesis that pharmacologic IL-13 inhibition ameliorates both inflammatory and remodeling pathways and could potentially reduce the risk of fibrostenotic complications.

Background Emerging evidence suggests airborne metals may be associated with breast cancer risk. However, breast cancer is heterogenous and associations with heavy metals vary by subtype. Heavy metals possess both carcinogenic and xenoestrogenic properties which may be related to different tumor etiologies. Therefore, we tested for etiologic heterogeneity, using a case-series approach, to determine whether associations between residential airborne metal concentrations and breast cancer differed by tumor subtype. Methods Between 2005 and 2008, we enrolled incident breast cancer cases into the Breast Cancer Care in Chicago study. Tumor estrogen and progesterone receptors status was determined by medical record abstraction and confirmed immunohistochemically (N = 696; 147 ER/PR-negative). The 2002 USEPA's National Air Toxics Assessment census-tract estimates of metal concentrations (antimony, arsenic, beryllium, cadmium, chromium, cobalt, lead, manganese, mercury, nickel and selenium) were matched to participants' residences of the same year. Adjusted logistic regression models were used to examine whether the airborne heavy metal associations differed by tumor ER/PR status. Principal component analysis was performed to assess associations by metal co-exposures. Results Comparing the highest and lowest quintiles, higher concentrations of antimony (odds ratio[OR]: 1.8, 95% confidence interval[CI]: 0.9, 3.7, P-trend: 0.05), cadmium (OR: 2.3, 95% CI: 1.2, 4.4, P-trend: 0.04) and cobalt (OR: 2.0, 95% CI: 0.9, 4.4, P-trend: 0.04) were associated with ER/PR-negative breast cancer. Mixture analysis using principal components suggested co-exposures to multiple airborne heavy metals may drive associations with tumor receptor status. Conclusions Among women diagnosed with breast cancer, metallic air pollutants were associated with increased odds of developing ER/PR-negative breast cancer.

Prostate cancer (PCa) disproportionately affects African American men. Early detection reduces risk of mortality. The United States Preventive Services Task Force (USPSTF) issued an updated recommendation statement on serum Prostate Specific Antigen (PSA)-based screening for PCa. Specifically, in 2012, the USPSTF recommended against PSA-based screening due to risk for overdiagnosis and overtreatment. However, the updated 2018 guidelines recommend consideration of screening for certain at risk men and revised the recommendation rating from “D” to “C.” This new guideline recommends providers to educate high-risk men on the benefits and harms of PSA-based PCa screening so that they can make an informed decision. The Affordable Care Act (ACA) includes provisions of service coverage for patient navigators who can help patients decide whether screening is appropriate, given potential risks and benefits, and training of health care providers in shared-decision regarding screening/treatment. These services can be utilized to support health care providers to better adhere to the new guideline. However, recommendations that are given a C rating or lower are not consistently reimbursed through many plans, including those offered through the ACA marketplace. The Society of Behavioral Medicine (SBM) supports the USPSTF guideline for the consideration of prostate cancer screening for high-risk men between the ages of 55 and 69. SBM encourages policymakers to include provisions for coverage of patient navigation services in the ACA to facilitate shared decision-making between providers and patients regarding screening.

Introduction and Objective: The low specificity of PSA in prostate cancer (PCa) screening has caused excessive biopsies and over-detection of PCa. The multi-parametric MRI (mpMRI) with its Prostate Imaging Reporting and Data Systems (PIRADS) score have emerged as an important tool to enhance detection of clinically significant PCa, i.e. Gleason grade group (GG) 2-5 PCa. Use of serum-based biomarkers, such as the Prostate Health Index (PHI), may be useful adjuncts to patients with lesions identified on mpMRI in detecting GG2-5 PCa and determining who would most benefit from a prostate biopsy. No study to date has investigated whether PHI improves accuracy of mpMRI in detecting GG2-5 PCa in biopsy-naïve men nor with adequate representation of Black men. The objective of this study was to determine the utility of PHI as a secondary reflex test to mpMRI in improving the accuracy for detecting GG2-5 PCa and improving specificity in a multi-ethnic biopsy-naïve population. Methods: From February 2017-2020, 143 men who were referred to participating urology clinics for elevated PSA levels were recruited into this prospective study and underwent both mpMRI and prostate biopsy. Enrolled men had to be between the ages of 40-79 years and prostate biopsy naïve to participate. Men with a history of a previous prostate cancer were excluded from the study. The mpMRI studies were blinded and read by two highly experienced genitourinary radiologists based on the standardized PIRADS criteria version 2.0. Logistic regression models were used to generate receiver operating characteristic curves. Models were tested for effect modification between Race and PHI and Race and PIRADS to determine if race impacted the prediction accuracy of either tool. The sensitivity, specificity, and predictive values of PHI thresholds were calculated by PIRADS scores. The primary outcome was GG2-5 PCa on biopsy, defined as any Gleason score ≥3+4. Results: The study included 143 men, of which 65 (45.5%) identified as Black. Median age was 62.0 years and 55 men (38.4%) had GG2-5 PCa. PIRADS risk groups were as follows: 18.1% PIRADS 1-2, 32.9% PIRADS 3, and 49.0% PIRADS 4-5. For the logistic regression models developed, the interactions between PIRADS and Race (p=0.08) and Log(PHI) and Race (p =0.17) variables were not statistically significant. A PHI score of ≥49 prevented unnecessary biopsies in 55% of men and missed zero cases of GG2-5 PCa, yielding a positive predictive value of 32.6% and negative predictive value of 100%. There was no reliable PHI threshold to exclude PCa biopsies in PIRADS 1-2 or 4-5. Conclusion: PHI and mpMRI of the prostate are independent predictors that aid in the detection of GG2-5 PCa in a Black enriched prostate biopsy cohort. PHI is particularly useful in men with equivocal PIRADS 3 lesions where a threshold of >49 safely avoids over 55% of unnecessary biopsies without missing GG2-5 PCa in our cohort. Citation Format: Samuel Carbunaru, Rilwan Babajide, Edward M. Schaeffer, Peter H. Gann, Adam B. Murphy. Predictive value of prostate health index (PHI) in multi-parametric MRI in an ethnically diverse cohort [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 2546.

Simple Summary Breast cancer, one of the leading causes of death among women, is a complex disease in which several factors, such as psychosocial stress, have been implicated in its initiation and progression. The glucocorticoid receptor (GCR) is one of the molecules that transfers the stress signal into the body. We measured the genetic variation and protein expression of GCR and the genes that regulate GCR function or response and examined whether these variations were associated with breast cancer. We found several genetic variants of functionally important SNPs associated with later disease stage, higher grade, and hormone receptor-negative status. The GCR protein expression was reduced in breast cancer tissue and correlated with the basal cell marker CK5/6. Abstract Background: Glucocorticoid, one of the primary mediators of stress, acts via its receptor, the glucocorticoid receptor (GCR/NR3C1), to regulate a myriad of physiological processes. We measured the genetic variation and protein expression of GCR, and the genes that regulate GCR function or response and examined whether these alterations were associated with breast cancer clinicopathological characteristics. Method: We used samples from a multiracial cohort of breast cancer patients to assess the association between breast cancer characteristics and the genetic variants of single nucleotide polymorphisms (SNPs) in GCR/NR3C1, FKBP5, Sgk1, IL-6, ADIPOQ, LEPR, SOD2, CAT, and BCL2. Results: Several SNPs were associated with breast cancer characteristics, but statistical significance was lost after adjustment for multiple comparisons. GCR was detected in all normal breast tissues and was predominantly located in the nuclei of the myoepithelial cell layer, whereas the luminal layer was negative for GCR. GCR expression was significantly decreased in all breast cancer tissue types, compared to nontumor tissue, but was not associated with breast cancer characteristics. We found that high nuclear GCR expression was associated with basal cell marker cytokeratin 5/6 positivity. Conclusion: GCR expression is reduced in breast cancer tissue and correlates with the basal cell marker CK5/6.

The Oncotype DX Prostate Cancer Assay® (ODX), a 17-gene expression profile performed on biopsy tissue, has been validated for prediction of adverse pathology at surgery and 10-year risk of metastasis or death. The assay output provided to the patient and clinician is a predicted probability. While these studies confirm the ability of the assay to discriminate indolent from aggressive prostate cancer (PCa), the assay’s actual clinical utility, in terms of its impact on treatment choice and psychological factors, has not been determined. Moreover, the assay’s utility for African American (AA) men remains an important knowledge gap for programs aimed at reducing PCa disparities through judicious use of active surveillance (AS). ENACT (Engaging Newly Diagnosed Men About Cancer Treatment Options) is an ongoing randomized trial comparing standard National Comprehensive Cancer Network (NCCN) based treatment counseling plus ODX to NCCN-based counseling alone for a largely AA population of men eligible for AS. The primary aim is to quantify the effect of ODX on treatment, testing the hypothesis that the assay causes a net migration towards AS. Secondary aims measure effects on psychological metrics dealing with cancer-related anxiety, decision conflict, and decision regret. Completion of the trial is expected in Fall 2019. Patients are enrolled at the UI Hospital, Stroger-Cook County Hospital and the Jesse Brown VA Medical Center in Chicago. Eligibility criteria include: biopsy-proven PCa with NCCN risk level Very Low, Low or Low Intermediate, age ⇐ 76, and estimated life expectancy >10 years. Thus far, 201 patients have enrolled and 180 have completed all study visits. Randomization (stratified by race and NCCN risk) occurs at Visit 1, when patients receive their diagnosis and initial counseling, and treatment decision is finalized at Visit 2, 2-3 weeks later. At Visit 3, ≈4 weeks after the decision but before treatment, surveys capture information about decision conflict and changes from Visit 1 in perceived cancer-related risk/mood. At Visit 4, ≈4 months after treatment starts, final surveys capture data on regret and evaluation of the decision process. Additional surveys record treatment preference of the attending urologist before and after the final decision. The enrolled population thus far is 71% AA, 35% age Citation Format: Adam B Murphy, Samuel Carbunaru, Michael Abern, Courtney MP Hollowell, Roohollah Sharifi, Peter H. Gann. ENACT: A randomized trial assessing the impact of the Oncotype DX Prostate Cancer Assay on treatment decision in a racially diverse population of men eligible for active surveillance [abstract]. In: Proceedings of the Twelfth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2019 Sep 20-23; San Francisco, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl_2):Abstract nr A082.

Prostate cancer (PCa) is a hormonally driven cancer and is currently the third most common cancer in the US. African American (AA) men are disproportionately at risk for both PCa and vitamin D (vitD) deficiency compared to white men. The numerous chemopreventative properties of vitD and epidemiological relationship of vitD status with PCa aggressiveness and mortality has led to the hypothesis that vitD deficiency is a biological contributor to PCa disparity in AA men. Our lab recently reported an unexpected relationship between serum and intraprostatic vitD metabolites 25-hydroxyvitamin D (25(OH)D) and 1,25-dihydroxyvitamin D in AA men. We also observed that Megalin, a multi-liganded endocytic membrane receptor encoded by the gene LRP2, was expressed in the prostate epithelium and is regulated by vitD. Extra-renal activity of Megalin has not been well studied as the widely accepted Free Hormone Hypothesis assumes passive diffusion of circulating free hormones into tissues. The presence of megalin suggests that globulin bound hormones from the circulation, including 25(OH)D bound to vitamin D binding protein (DBP) and testosterone (T) bound to sex hormone binding globulin (SHBG), are imported into the prostate in a regulated manner. Here we examine megalin as a potential mechanism to regulate globulin bound hormone import into the prostate. 25(OH)D decreased expression of LRP2 in primary prostate epithelial cells and fresh human prostate tissue slice explants. DBP-bound 25(OH)D and SHBG-bound T were imported into these prostate models and transcriptionally active. Lastly, we quantified T and its active metabolite dihydrotestosterone (DHT) in the patient cohort from our prior study. Prostatic DHT levels inversely correlated with serum 25(OH)D status. AA men had higher levels of DHT in prostate tissue compared to white men. These clinical findings support our hypothesis that vitD status regulates intraprostatic hormone levels. In summary, we report the presence of a negative feedback loop in which vitD deficiency increases hormone import into prostate epithelium via megalin. Therefore, the upregulation of megalin in the setting of vitamin D deficiency may facilitate increased import of circulating sex steroids into the prostate contributing to carcinogenesis in AA men. Citation Format: Jason Garcia, Zachary Richards, Morgan Zenner, Yancai Wang, Peter Gann, Gail Prins, Larisa Nonn. Regulation of megalin by vitamin D as the mechanism for differential levels of intra-prostatic androgens between African American and Caucasian men [abstract]. In: Proceedings of the Twelfth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2019 Sep 20-23; San Francisco, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl_2):Abstract nr B094.

Objective: To validate the 17-gene Oncotype DX Genomic Prostate Score (GPS) biopsy-based gene expression assay as a predictor of adverse pathology (AP: Pathologic Gleason score ≥4+3, presence of any Gleason 5, and/or ≥pT3) in African American (AA) men. Methods: Between February 2009 and September 2014, AA and European American (EA) men with very low, low, and intermediate risk prostate cancer (PCa) enrolled in a multi-institutional prospective study of vitamin D impacts of biopsy outcomes. The subset who proceeded to immediate radical prostatectomy (RP) after biopsy with available biopsy tumor blocks was included in a comparative effectiveness analysis of GPS on biopsy and its association with surgical AP on RP using logistic regression and receiver operating characteristic curves. Multiplicative interactions tested for differential prediction of GPS accuracy by race (AA vs. EA). Results: Overall, 102 AA and 76 EA men elected RP, out of which 51 (47.2%) had AP. GPS result was a significant predictor of AP (odds ratio per 20 GPS units [OR/20 units] in AA: 4.78; 95% CI 1.8-12.5, P =0.001; and EA: 4.41; 95% CI 1.6-11.9, P =0.003) in univariable analysis. On multivariate analysis, there was a significant interaction between GPS and race (P=0.01). On race stratified binary logistic regression, AP remained significant after adjustment for NCCN risk group in both AA and EA men (OR/20 units in AA: 3.29; 95% CI 1.2-9.1, P =0.02; and EA: 4.24; 95% CI 1.4-12.6, P =0.01). Area under the curves for AP using GPS/20 units was 0.719 for AAs vs. 0.745 for EAs (P=0.39). Conclusion: In this AA validation study, the Oncotype Dx PCa assay was confirmed as an independent predictor of AP at prostatectomy in AA and EA men with similar predictive accuracy, though there was evidence of effect modification by race. Citation Format: Samuel Carbunaru, Virgilia Marcias, Peter Gann, Roohollah Sharifi, Ximing Yang, Michael Dixon, Chase Gorbein, Borko Jovanovic, Andre Kajdacsy-Balla, Adam B. Murphy. Oncotype DX assay has similar predictive accuracy for adverse pathology at radical prostatectomy in African American and European American men [abstract]. In: Proceedings of the Twelfth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2019 Sep 20-23; San Francisco, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl_2):Abstract nr A067.

Introduction: The low specificity of PSA in prostate cancer (PCa) screening has caused excessive biopsies and PCa over-detection. PHI is more specific for clinically significant (cs) PCa (i.e. Gleason >6 PCa) and includes serum PSA, [-2]proPSA, and free PSA in its formula. Validation studies in diverse populations have suggested that PHI should be used as a reflex test and the optimal cutoffs for PHI vary across ethnic groups, but have not been determined in Black men. We sought to 1) assess the distributions of PHI and PSA in healthy controls and men with high-grade PCa and 2) assess the accuracy of PHI vs. PSA for detection of csPCa in Black men with elevated PSA. Methods: The present study population consists of two biopsy-naïve cohorts. The first cohort serves as healthy controls and includes Black men with PHI and PSA drawn at community screening events through the social networks of eight lay Black male Citizen Scientists. The second PHI Biopsy Study cohort consists of Black men referred with elevated PSA and/or abnormal digital rectal exam (DRE) who had PHI drawn immediately before prostate biopsy. We excluded men with prior biopsies or known PCa. Distributions of PHI and PSA were compared across controls and men with negative biopsies, Gleason 6 PCa and csPCa. Among men with biopsy, we compared the area under the receiver operating characteristics curves (AUC) for PSA and PHI for detecting csPCa and assessed specificity at selected sensitivities as a proxy of avoided biopsies. We descriptively assessed theoretically avoided biopsies and missed csPCa at previously established PHI cut points [27.0, 28.6, 29.9, and 35.0]. Results: For the first objective, 139 men (42.5%) were included as controls from the Citizens Scientist Study and 188 (57.5%) were from the PHI Biopsy Study cohort, out of which 82 (43.6%) had a negative biopsy for PCa, 40 (21.1%) had low-grade PCa, and 66 (35.1%) had a Gleason >6 PCa. Compared to the healthy control group, the PHI Biopsy Study cohort was older (median age 61 vs. 55 years), with higher PSA levels (7.1ng/mL vs. 0.9ng/mL) and PHI scores (61.1 vs. 20.1), and more commonly had a family history of PCa (19.4% vs. 5.0%) and benign prostatic hyperplasia diagnosis (21.3% vs. 2.2%) (all p Citation Format: Samuel Carbunaru, Oluwarotimi Nettey, Edward M Schaeffer, Peter H Gann, Michael Abern, Courtney M.P Hollowell, Karriem Watson, Tiffany McDowell, Rick A Kittles, Adam B Murphy. Performance of the prostate health index (PHI) assay in black men [abstract]. In: Proceedings of the Twelfth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2019 Sep 20-23; San Francisco, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl_2):Abstract nr A066.

INTRODUCTION AND OBJECTIVE:The Stockholm3 test has not been validated in an American cohort where clinical pathways to biopsy and race differ. We aimed to match clinical data stratified by race and...

Glutathione peroxidase 1 (GPX1) is an extensively studied selenium-dependent protein that reduces hydrogen and lipid peroxides to water. Because of its antioxidant function and its responsiveness to dietary intakes of selenium, an essential trace element whose levels are inversely associated with prostate cancer risk, GPX1 levels were assessed in a prostate cancer tissue microarray, comparing cases of recurrent prostate cancer following prostatectomy to non-recurrent controls. While GPX1 is generally considered as a protein that resides in both the cytoplasm and mitochondria, we detected strong nuclear staining by immunofluorescence using GPX1-specific antibodies. Nuclear localization of GPX1 was also observed in both primary prostate epithelial cells and the immortalized prostate-derived cell line RWPE-1, but not in LNCaP or PC3 prostate tumor-derived cell lines. Quantification of GPX1 levels in the entire cell, the cytoplasm, and the nucleus did not indicate any association of either its levels or subcellular distribution with prostate cancer recurrence. While GPX1 levels may not have an impact on survival among men with prostate cancer, the data indicates that this extensively characterized protein may have a novel function in the nucleus of prostate epithelial cells.

A subset of men with prostate cancer develops aggressive disease. We sought to determine whether miR-182, an miRNA with reported oncogenic functions in the prostate, is associated with biochemical recurrence and aggressive disease. Prostate epithelial miR-182 expression was quantified via in situ hybridization of two prostate tissue microarrays and by laser-capture microdissection of prostate epithelium. miR-182 was significantly higher in cancer epithelium than adjacent benign epithelium (P

A diverse body of evidence suggests that lycopene might inhibit prostate cancer development. We conducted a 6-mo repeat biopsy randomized trial among men with high-grade prostatic intraepithelial neoplasia (HGPIN). Here we report results for serum lycopene, prostate specific antigen (PSA) and insulin-like growth factor (IGF) proteins, histopathological review, and tissue markers for proliferation [minichromosome maintenance protein 2 (MCM-2)] and cell cycle inhibition (p27). Participants consumed placebo or tomato extract capsules containing 30 mg/day lycopene. Pre- and posttreatment biopsies were immunostained and digitally scored. Serum lycopene was determined by LC-MS-MS. In secondary analyses, pathologists blindly reviewed each biopsy to score histological features. Fifty-eight men completed the trial. Serum lycopene increased 0.55 μmol/L with treatment and declined 0.29 μmol/L with placebo. We observed no meaningful differences in PSA, IGF-1, or IGF binding protein 3 concentrations between groups, nor any differences in expression of MCM-2 or p27 in epithelial nuclei. Prevalences of cancer, HGPIN, atrophy, or inflammation posttreatment were similar; however, more extensive atrophy and less extensive HGPIN was more common in the lycopene group. Despite large differences in serum lycopene following intervention, no treatment effects were apparent on either the serum or benign tissue endpoints. Larger studies are warranted to determine whether changes observed in extent of HGPIN and focal atrophy can be replicated.

Background Selenium status is inversely associated with the incidence of prostate cancer. However, supplementation trials have not indicated a benefit of selenium supplementation in reducing cancer risk. Polymorphisms in the gene encoding selenoprotein 15 (SELENOF) are associated with cancer incidence/mortality and present disproportionately in African Americans. Relationships among the genotype of selenoproteins implicated in increased cancer risk, selenium status, and race with prostate cancer were investigated. Methods Tissue microarrays were used to assess SELENOF levels and cellular location in prostatic tissue. Sera and DNA from participants of the Chicago-based Adiposity Study Cohort were used to quantify selenium levels and genotype frequencies of the genes for SELENOF and the selenium-carrier protein selenoprotein P (SELENOP). Logistic regression models for dichotomous patient outcomes and regression models for continuous outcome were employed to identify both clinical, genetic, and biochemical characteristics that are associated with these outcomes. Results SELENOF is dramatically reduced in prostate cancer and lower in tumors derived from African American men as compared to tumors obtained from Caucasians. Differing frequency of SELENOF polymorphisms and lower selenium levels were observed in African Americans as compared to Caucasians. SELENOF genotypes were associated with higher histological tumor grade. A polymorphism in SELENOP was associated with recurrence and higher serum PSA. Conclusions These results indicate an interaction between selenium status and selenoprotein genotypes that may contribute to the disparity in prostate cancer incidence and outcome experienced by African Americans.

BACKGROUND Arylsulfatase B (ARSB) removes the 4-sulfate group from chondroitin 4-sulfate (C4S) and dermatan sulfate and is required for their degradation. Prior work showed that ARSB immunohistochemical scores were lower in malignant prostate tissue, and were associated with higher Gleason scores and recurrence. OBJECTIVE This study aims to confirm that ARSB immunostaining of prostate tissue obtained at the time of radical prostatectomy is prognostic for prostate cancer recurrence. METHODS Intensity and distribution of ARSB immunostaining were digitally analyzed in a large, well-annotated, prostate cancer tissue microarray (TMA). Scores were calculated for stroma and epithelium and compared for 191 cases, including 36 recurrences, defined as PSA > 0.2 ng/ml. RESULTS Epithelial scores were significantly lower in the recurrences (p= 0.010), and among subgroups with age > 60, initial PSA > 6 ng/ml, or Gleason grade = 7. ARSB score did not improve the prediction of recurrence in multifactorial analysis. CONCLUSIONS Study findings validate previous findings and provide further evidence that lower ARSB is associated with prostate cancer recurrence. Additional studies are required to assess if there are specific cutoff values that may help predict recurrence.

The majority of patients with colon cancer will develop advanced disease, with the liver being the most common site of metastatic disease. Patients with increased numbers of tumor-infiltrating lymphocytes in primary colon tumors and liver metastases have improved outcomes. However, the molecular factors that could empower antitumor immune responses in this setting remain to be elucidated. We reported that the immunostimulatory cytokine LIGHT (TNFSF14) in the microenvironment of colon cancer metastases associates with improved patient survival, and here we demonstrate in an immunocompetent murine model that colon tumors expressing LIGHT stimulate lymphocyte proliferation and tumor cell–specific antitumor immune responses. In this model, increasing LIGHT expression in the microenvironment of either primary tumors or liver metastases triggered regression of established tumors and slowed the growth of liver metastases, driven by cytotoxic T-lymphocyte–mediated antitumor immunity. These responses corresponded with significant increases in tumor-infiltrating lymphocytes and increased expression of lymphocyte-homing signals in the metastatic tumors. Furthermore, we demonstrated evidence of durable tumor-specific antitumor immunity. In conclusion, increasing LIGHT expression increased T-cell proliferation, activation, and infiltration, resulting in enhanced tumor-specific immune-mediated tumor regressions in primary tumors and colorectal liver metastases. Mechanisms to increase LIGHT in the colon cancer microenvironment warrant further investigation and hold promise as an immunotherapeutic strategy. Cancer Res; 77(8); 1880–91. ©2017 AACR.

Accurate prediction of the treatment outcome is important for cancer treatment planning. We present an approach to predict prostate cancer (PCa) recurrence after radical prostatectomy using tissue images. We used a cohort whose case vs. control (recurrent vs. non-recurrent) status had been determined using post-treatment follow up. Further, to aid the development of novel biomarkers of PCa recurrence, cases and controls were paired based on matching of other predictive clinical variables such as Gleason grade, stage, age, and race. For this cohort, tissue resection microarray with up to four cores per patient was available. The proposed approach is based on deep learning, and its novelty lies in the use of two separate convolutional neural networks (CNNs) – one to detect individual nuclei even in the crowded areas, and the other to classify them. To detect nuclear centers in an image, the first CNN predicts distance transform of the underlying (but unknown) multi-nuclear map from the input HE image. The second CNN classifies the patches centered at nuclear centers into those belonging to cases or controls. Voting across patches extracted from image(s) of a patient yields the probability of recurrence for the patient. The proposed approach gave 0.81 AUC for a sample of 30 recurrent cases and 30 non-recurrent controls, after being trained on an independent set of 80 case-controls pairs. If validated further, such an approach might help in choosing between a combination of treatment options such as active surveillance, radical prostatectomy, radiation, and hormone therapy. It can also generalize to the prediction of treatment outcomes in other cancers.

Context: Several therapeutically important mutations in cancers are economically detected using immunohistochemistry (IHC), which highlights the overexpression of specific antigens associated with the mutation. However, IHC panels can be imprecise and relatively expensive in low-income settings. On the other hand, although hematoxylin and eosin (H&E) staining used to visualize the general tissue morphology is a routine and low cost, it does not highlight any specific antigen or mutation. Aims: Using the human epidermal growth factor receptor 2 (HER2) mutation in breast cancer as an example, we strengthen the case for cost-effective detection and screening of overexpression of HER2 protein in H&E-stained tissue. Settings and Design: We use computational methods that reliably detect subtle morphological changes associated with the over-expression of mutation-specific proteins directly from H&E images. Subjects and Methods: We trained a classification pipeline to determine HER2 overexpression status of H&E stained whole slide images. Our training dataset was derived from a single hospital containing 26 (11 HER2+ and 15 HER2–) cases. We tested the classification pipeline on 26 (8 HER2+ and 18 HER2–) held-out cases from the same hospital and 45 independent cases (23 HER2+ and 22 HER2–) from the TCGA-BRCA cohort. The pipeline was composed of a stain separation module and three deep neural network modules in tandem for robustness and interpretability. Statistical Analysis Used: We evaluate our trained model through area under the curve (AUC)-receiver operating characteristic. Results: Our pipeline achieved an AUC of 0.82 (confidence interval [CI]: 0. 65–0. 98) on held-out cases and an AUC of 0.76 (CI: 0. 61–0. 89) on the independent dataset from TCGA. We also demonstrate the region-level correspondence of HER2 overexpression between a patient's IHC and H&E serial sections. Conclusions: Our work strengthens the case for automatically quantifying the overexpression of mutation-specific proteins in H&E-stained digital pathology, and it highlights the importance of multi-stage machine learning pipelines for added robustness and interpretability.

BACKGROUND Alpha-methylacyl-CoA racemase (AMACR) is an enzyme involved in fatty acid metabolism that is markedly over-expressed in virtually all prostate cancers (PCa), relative to benign tissue. One of AMACR's primary substrates, phytanic acid, is derived predominately from red meat and dairy product consumption. Epidemiological evidence suggests links between dairy/red meat intake, as well as phytanic acid levels, and elevated PCa risk. This study investigates the relationships among dietary intake, serum and tissue concentrations of phytanic acid, and AMACR expression (mRNA and protein) in the histologically benign human prostate. METHODS Men undergoing radical prostatectomy for the treatment of localized disease provided a food frequency questionnaire (n = 68), fasting blood (n = 35), benign fresh frozen prostate tissue (n = 26), and formalin-fixed paraffin-embedded (FFPE) sections (n = 67). Serum and tissue phytanic acid concentrations were obtained by gas chromatography–mass spectrometry. We extracted RNA from epithelial cells using laser capture microdissection and quantified mRNA expression of AMACR and other genes involved in the peroxisomal phytanic acid metabolism pathway via qRT-PCR. Immunohistochemistry for AMACR was performed on FFPE sections and subsequently quantified via digital image analysis. Associations between diet, serum, and tissue phytanic acid levels, as well as AMACR and other gene expression levels were assessed by partial Spearman correlation coefficients. RESULTS High-fat dairy intake was the strongest predictor of circulating phytanic acid concentrations (r = 0.35, P = 0.04). Tissue phytanic acid concentrations were not associated with any dietary sources and were only weakly correlated with serum levels (r = 0.29, P = 0.15). AMACR gene expression was not associated with serum phytanic acid (r = 0.13, P = 0.47), prostatic phytanic acid concentrations (r = 0.03, P = 0.88), or AMACR protein expression (r = −0.16, P = 0.20). CONCLUSIONS Our data underscore the complexity of the relationship between AMACR and its substrates and do not support the unifying hypothesis that excess levels of dietary phytanic acid are responsible for both the overexpression of AMACR in prostate cancer and the potential association between PCa risk and intake of dairy foods and red meat. Prostate 75:200–210, 2015. © 2014 Wiley Periodicals, Inc.

Phytanic acid is a saturated branched-chain fatty acid found predominantly in red meat and dairy products, and may contribute to the elevated risks of prostate cancer associated with higher consumption of these foods. Pristanic acid is formed during peroxisomal oxidation of phytanic acid, and is the direct substrate of α-Methyl-CoA-Racemase (AMACR)--an enzyme that is consistently overexpressed in prostate tumors relative to benign tissue. We measured phytanic and pristanic acids as percentages of total fatty acids by gas chromatography-mass spectrometry in prediagnostic blood samples from 300 prostate cancer cases and 300 matched controls, all of whom were participants in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study supplementation trial and follow-up cohort. In addition to providing a fasting blood sample at baseline, all men completed extensive diet, lifestyle, and medical history questionnaires. Among controls, the strongest dietary correlates of serum phytanic and pristanic acids were saturated fat, dairy fat, and butter (r = 0.50 and 0.40, 0.46 and 0.38, and 0.40 and 0.37, respectively; all P-values

We previously identified the intracellular nicotinamide phosphoribosyltransferase (iNAMPT, aka pre–B-cell colony enhancing factor) as a candidate gene promoting acute respiratory distress syndrome (ARDS) and ventilator-induced lung injury (VILI) with circulating nicotinamide phosphoribosyltransferase potently inducing NF-κB signaling in lung endothelium. iNAMPT also synthesizes intracellular nicotinamide adenine dinucleotide (iNAD) in response to extracellular oxidative stress, contributing to the inhibition of apoptosis via ill-defined mechanisms. We now further define the role of iNAMPT activity in the pathogenesis of ARDS/VILI using the selective iNAMPT inhibitor FK-866. C57/B6 mice were exposed to VILI (40 ml/kg, 4 h) or LPS (1.5 mg/kg, 18 h) after osmotic pump delivery of FK-866 (100 mg/kg/d, intraperitoneally). Assessment of total bronchoalveolar lavage (BAL) protein, polymorphonuclear neutrophil (PMN) levels, cytokine levels (TNF-α, IL-6, IL-1α), lung iNAD levels, and injury scores revealed that FK-866–mediated iNAMPT inhibition successfully reduced lung tissue iNAD levels, BAL injury indices, inflammatory cell infiltration, and lung injury scores in LPS- and VILI-exposed mice. FK-866 further increased lung PMN apoptosis, as reflected by caspase-3 activation in BAL PMNs. These findings support iNAMPT inhibition via FK-866 as a novel therapeutic agent for ARDS via enhanced apoptosis in inflammatory PMNs.

Protein Tyrosine kinase 6 (PTK6/BRK) is overexpressed in the majority of human breast tumors and breast tumor cell lines. It is also expressed in normal epithelial linings of the gastrointestinal tract, skin, and prostate. To date, expression of PTK6 has not been extensively examined in the normal human mammary gland. We detected PTK6 mRNA and protein expression in the immortalized normal MCF-10A human mammary gland epithelial cell line, and examined PTK6 expression and activation in a normal human breast tissue microarray, as well as in human breast tumors. Phosphorylation of tyrosine residue 342 in the PTK6 activation loop corresponds with its activation. Similar to findings in the prostate, we detect nuclear and cytoplasmic PTK6 in normal mammary gland epithelial cells, but no phosphorylation of tyrosine residue 342. However, in human breast tumors, striking PTK6 expression and phosphorylation of tyrosine 342 is observed at the plasma membrane. PTK6 is expressed in the normal human mammary gland, but does not appear to be active and may have kinase-independent functions that are distinct from its cancer promoting activities at the membrane. Understanding consequences of PTK6 activation at the plasma membrane may have implications for developing novel targeted therapies against this kinase.

Introduction PAM50 gene profiling assigns individual breast cancers (BCas) to one of five intrinsic subtypes. However, individual tumors vary in their adherence to the assigned subtype, and a subset of these may show admixture with another subtype. We developed a novel metric to quantify this intratumoral heterogeneity using whole transcriptome expression data of individual BCas and studied specific pairwise admixtures for Luminal A (LumA) cases. Methods We obtained whole transcriptome expression data and PAM50 labels of 1081 and 1980 BCa cases from TCGA and METABRIC cohorts, respectively. We combined the two datasets, by identifying a common gene set. METABRIC data was pre-processed to normalize and log-transform expression levels and remove batch effects. For each case, the combined cohort had expression data of 11,379 genes, and probabilities of its adherence to each of the four major subtypes were computed using semi-supervised non-negative matrix factorization. The four probabilities obtained for each case were then combined using the proposed Aggregated Differential Sum (ADS) metric. For each subtype, the cases were subsequently stratified into pure, neither and admixed categories by computing ADS tertiles. For every case, we did comparative analysis of the ADS terms to identify the specific alternate molecular subtype for pairwise admixture analysis. Results Using the proposed ADS criteria, pure and admixed LumA cases spanned tertile 1 (T1) and 3 (T3), respectively. Compared to admixed cases, pure LumA cases were younger and had favorable clinical characteristics including a surrogate LumA profile (ER/PR+ and HER2-), smaller tumor size, lower stage, less prevalence of TP53 mutations, and high likelihood of PIK3CA and CBFB mutations. Kaplan-Meier curves and Cox proportional hazard analysis showed that the admixed cases had worse overall survival and higher mortality risk compared to their pure counterparts. Pairwise admixture analysis revealed that 59% (n=226), 27% (103) and 15% (55) LumA cases in T3 were admixed with Luminal B, Basal and HER2 subtype, respectively. LumA cases admixed with HER2 showed worst tumor characteristics, survival outcomes and highest mortality risk followed by, in order, the admixture with Luminal B and Basal subtypes. Conclusion Our results demonstrate that stratification of LumA BCas by subtype purity using ADS identifies the admixed cases which usually have worse tumor characteristics and survival outcomes. LumA cases were predominantly admixed with Luminal B subtype followed by Basal and HER2 subtypes. Eventually, we will use the pure cases to train computer vision classifiers for morphological identification of subtype admixture in routine H&E images. This would be a cost-effective complement to molecular subtyping that could take advantage of the spatial information available in whole slide images. Citation Format: Neeraj Kumar, Yash Dharmamer, Amit Sethi, Peter Gann. Quantification of intratumoral heterogeneity in individual luminal A breast cancers from whole transcriptome data through semi-supervised learning [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4700.

Background: Nipple aspiration fluid (NAF) use as a biosample is limited by the variable yield across studies. We investigated the endocrine determinants of yield in an ongoing breast cancer case–control study. Methods: One-hundred and eighteen women yielding ≥2 μL NAF and 120 non-yielders were included; serum hormones were measured; differences in median hormones were assessed using the Wilcoxon rank-sum test. ORs and 95% confidence intervals (95% CI) for yielder status relative to hormone levels were estimated using logistic regression, adjusting for parity and lactation, and, in premenopausal women, menstrual cycle phase (MCP). Results: Prolactin concentrations were higher in yielders than non-yielders (premenopausal: 7.6 and 2.5 ng/mL, P < 0.01; postmenopausal 5.3 and 2.2 ng/mL; P < 0.01). Among premenopausal-yielders, estradiol was lower (64.3 vs. 90.5 pg/mL, MCP-adjusted P = 0.02). In separate menopausal status and parity-adjusted models, significant case–control differences persisted in prolactin: case OR 1.93 (95% CI, 1.35–2.77), control OR 1.64 (95% CI, 1.17–2.29). Premenopausal control yielders had higher progesterone (OR, 1.70; 95% CI, 1.18–2.46) and sex-hormone binding-globulin (OR, 2.09; 95% CI, 1.08–4.05) than non-yielders. Among parous women, further adjustment for lactation suggested a stronger positive association of serum prolactin with yield in cases than controls. Conclusion: NAF-yielders show higher prolactin than non-yielders, regardless of menopause and parity; implications of this and other endocrine differences on NAF biomarkers of breast cancer risk deserve further study. Impact: NAF yield is associated with a distinct endocrine environment that must be considered in studies of NAF-based breast cancer risk markers. Cancer Epidemiol Biomarkers Prev; 22(12); 2277–84. ©2013 AACR.

BACKGROUND The enzyme Arylsulfatase B (ARSB; N-acetylgalactosamine-4-sulfatase) degrades chondroitin-4-sulfate (C4S) and is reduced in malignant colonic and mammary tissues, but has not previously been evaluated in prostate cancer. METHODS ARSB immunostaining was performed on two tissue microarrays (TMA) and analyzed by digital image analysis, generating ARSB H-scores for prevalence and intensity of epithelial, stromal, and combined epithelial and stromal immunostaining. Also, paired malignant and normal prostate tissues were analyzed for ARSB activity, C4S, total sulfated glycosaminoglycans, and versican content. The quantities of C4S and of the epidermal growth factor receptor that co-immunoprecipitated with versican were determined in the normal and malignant paired prostate tissues. RESULTS 44 cases of prostate cancer were paired by age (± 5y), race, Gleason score (in order), and pathologic TNM score. The pairs differed by recurrence vs. non-recurrence of elevated PSA at 4 or more years. When TMA cores were analyzed for ARSB H-score, 18 of the 22 pairs had lower ARSB H-scores in the recurrent member of the pair, whereas higher initial PSA values were associated with recurrence in only 65% of the paired cases. In a second TMA, Gleason scores 6 and 7 were associated with higher ARSB H-scores than Gleason scores 8 and 9 for stroma, epithelium, and stroma and epithelium combined (p=0.052, p=0.015, p

Purpose Lower serum vitamin D levels have been associated with an increased risk of aggressive prostate cancer. Among men with localized prostate cancer, especially with low- or intermediate-risk disease, vitamin D may serve as an important biomarker of disease aggression. The aim of this study was to assess the relationship between adverse pathology at the time of radical prostatectomy and serum 25-hydroxyvitamin D (25-OH D) levels. Methods This cross-sectional study was carried out from 2009 to 2014, nested within a large epidemiologic study of 1,760 healthy controls and men undergoing prostate cancer screening. In total, 190 men underwent radical prostatectomy in the cohort. Adverse pathology was defined as the presence of primary Gleason 4 or any Gleason 5 disease, or extraprostatic extension. Descriptive and multivariate analyses were performed to assess the relationship between 25-OH D and adverse pathology at the time of prostatectomy. Results Eighty-seven men (45.8%) in this cohort demonstrated adverse pathology at radical prostatectomy. The median age in the cohort was 64.0 years (interquartile range, 59.0 to 67.0). On univariate analysis, men with adverse pathology at radical prostatectomy demonstrated lower median serum 25-OH D (22.7 v 27.0 ng/mL, P = .007) compared with their counterparts. On multivariate analysis, controlling for age, serum prostate specific antigen, and abnormal digital rectal examination, serum 25-OH D less than 30 ng/mL was associated with increased odds of adverse pathology (odds ratio, 2.64; 95% CI, 1.25 to 5.59; P = .01). Conclusion Insufficiency/deficiency of serum 25-OH D is associated with increased odds of adverse pathology in men with localized disease undergoing radical prostatectomy. Serum 25-OH D may serve as a useful biomarker in prostate cancer aggressiveness, which deserves continued study.

BACKGROUND. African American (AA) men are disproportionately affected by both prostate cancer (PCa) and vitamin D deficiency compared with European American (EA) men. Vitamin D deficiency is linked to increased PCa aggressiveness and mortality. Therefore, it has been hypothesized that vitamin D deficiency may contribute to the PCa disparity between AA and EA men. METHODS. We studied a cross sectional group of 60 PCa patients (AA, n = 31; EA, n = 29) who underwent radical prostatectomy. Vitamin D metabolites 25-hydroxyvitamin D (25(OH)D) and 1,25-dihydroxyvitamin D (1,25(OH)2D) were measured in the serum and tissue by uHPLC-MS-MS. Tissue was laser capture microdissected, and gene expression was quantified by microarray. DNA isolated from whole blood was genotyped for West African ancestry markers and vitamin D-related SNPs. RESULTS. Serum concentrations of 25(OH)D were lower in AAs, but concentrations of 1,25(OH)2D in the prostate tissue were higher compared with EAs. Expression of the vitamin D receptor was higher in prostate tissue from AAs. Expression of the extracellular receptor of vitamin D binding protein, LRP2, was positively associated with West African ancestry and inversely associated with tissue 25(OH)D concentrations in AAs. CONCLUSIONS. The relationships between vitamin D binding protein LRP2 and vitamin D metabolites suggest that the prohormone is actively transported into the prostate, followed by intraprostatic conversion to the active hormone, rather than passive diffusion. These findings support the presence of a compensatory response in prostate tissue to vitamin D deficiency in AAs and reveal a previously unknown complexity involving tissue distribution of vitamin D metabolites. FUNDING. Department of Defense Prostate Cancer Research Program Idea Award for Disparities Research PC121923 (LN and RK) and the NIH 1R01MD007105 (RK).

Vitamin D deficiency increases the risk of lethal prostate adenocarcinomas (PCa) and the majority of older men are deficient. Although PCa arises from the epithelium, the surrounding stroma has hormonal regulatory control over the epithelium and contributes to carcinogenesis. Herein, we describe regulation of microRNAs (miRs) by the active hormone dihydroxyvitamin D (1,25(OH) 2 D) in human prostate stroma. 1,25(OH) 2 D binds the vitamin D receptor (VDR) transcription factor to regulate gene expression, including miRs, which have emerged as potent regulators of protein expression. 1,25(OH) 2 D–regulated miRs were identified by profiling in primary human prostatic stromal cells (PrS) and three miRs, miR-126-3p, miR 154-5p and miR-21-5p were subsequently validated in laser-capture micro-dissected prostate stromal tissue from a vitamin D3 clinical trial (N = 45). Regulation of these miRs by 1,25(OH) 2 D was VDR-dependent. Network analysis of known and putative mRNA targets of these miRs was enriched with cancer and inflammation pathways, consistent with known roles of stroma and of vitamin D in carcinogenesis. Expression of the miR processing ribonuclease, DICER1, positively correlated with vitamin D metabolite levels in the clinical trial specimens. High epithelial/stromal ratios of DICER1 were significantly associated biochemical recurrence (OR 3.1, p = 0.03) in a tissue microarray of 170 matched PCa patients. In summary, these results underscore the role of the prostate stroma in regulating responses to the hormone 1,25(OH) 2 D and identified miRs and DICER1 as being regulated in human prostate stroma. Regulation of stromal DICER1 by 1,25(OH) 2 D may also have clinical relevance in protection against aggressive PCa.

Determining the molecular or genomic sub-type of a cancer of a particular organ is important for prognosis and treatment planning. While clinical tests determine the dominant subtype of cancer in a patient, it is believed that some cancer treatments targeting the dominant sub-types ultimately prove ineffective because they ignore the existence of additional sub-types in the same patient. We present a method to detect co-existence of two common sub-types of breast cancer — HER2 and basal-like — in whole slide images of H&E-stained tumors in the same patient. Our main contribution is in formulating this problem in precision medicine and in preparation of its training data. The detectors tested to be highly accurate in classifying patches taken from test cases with well-separated molecular profiles, confirming that image classification was effective in detecting the two sub-types. Interestingly, test patients whose molecular profiles were not very clearly separated appeared to include heterogeneous patches — some HER2 and other basal-like — indicating existence of a secondary sub-type within a single patient. The detection of a secondary sub-type in a single patient may impact the prognosis and treatment of the cancer.

PurposeProstate-specific antigen (PSA) level in midlife predicted future prostate cancer (PCa) mortality in an unscreened Swedish population. Our purpose was to determine if a baseline PSA level during midlife predicts lethal PCa in a US population with opportunistic screening.Materials and MethodsWe conducted a nested case-control study among men age 40 to 59 years who gave blood before random assignment in the Physicians’ Health Study, a randomized, placebo-controlled trial of aspirin and β-carotene among 22,071 US male physicians initiated in 1982 and then transitioned into a prospective cohort with 30 years of follow-up. Baseline PSA levels were available for 234 patients with PCa and 711 age-matched controls. Seventy-one participants who developed lethal PCa were rematched to 213 controls. Conditional logistic regression was used to estimate odds ratios and the area under the receiver operating characteristic curve, with 95% CIs, of the association between baseline PSA and risk of lethal PCa.ResultsMedian PSA among controls was 0.68, 0.88, and 0.96 ng/mL for men age 40 to 49, 50 to 54, and 55 to 59 years, respectively. Risk of lethal PCa was strongly associated with baseline PSA in midlife: odds ratios (95% CIs) comparing PSA in the > 90th percentile versus less than or equal to median were 8.7 (1.0 to 78.2) at 40 to 49 years, 12.6 (1.4 to 110.4) at 50 to 54 years, and 6.9 (2.5 to 19.1) at 55 to 59 years. A total of 82%, 71%, and 86% of lethal cases occurred in men with PSA above the median at ages 40 to 49, 50 to 54, and 55 to 59 years, respectively.ConclusionPSA levels in midlife strongly predict future lethal PCa in a US cohort subject to opportunistic screening. Risk-stratified screening on the basis of midlife PSA should be considered in men age 45 to 59 years.

Background: In vitro and preclinical data indicate that TPA, a selective PR modulator, has activity against hormone-sensitive early breast cancer. We conducted a pre-surgical window trial of oral TPA in Stage 0-II breast cancer to assess the effect of TPA on suppression of cell proliferation (Ki67), and on differential gene expression in responsive and non-responsive tumors. Methods: We enrolled 70 pre and postmenopausal women into a 1:1 randomized, double-blind, placebo-controlled trial of oral TPA 12mg (Repros Therapeutics Inc.) for 2-10 weeks. The primary endpoint was Ki67 labelling, comparing diagnostic core needle biopsy to post-therapy surgical specimens. Ki67 changes were quantitated by dual immunohistochemistry (Ki67/pan-cytokeratin) and image analysis (Aperio ImageScope and Definiens Tissue Studio®). RNA-sequencing (using RNA extracted from the paraffin blocks) was performed with Illumina TruSeq RNA Coding Access method. Differential gene expression pre-post therapy was assessed, followed by Gene Set Enrichment Analysis for pathway analysis. Ki67 changes from baseline were tested with Paired signed-rank test. For gene expression analysis, p values were calculated by Wald test and adjusted for multiple comparisons by Benjamini-Hochberg method (adjusted p Results: Among 61 evaluable women, (29 placebo and 32 TPA) 97% of tumors were ER or PR positive and 91% were ER and PR positive (balanced across arms). A significant 6% decrease in mean %Ki67 was seen in the TPA arm (p= 0.003). When stratified by menopause, the significance held in premenopausal women (n= 22, p= 0.03) but not in postmenopausal women (n=10, p= 0.08). However, a Ki67 decrease (4%) was also observed in placebo group (p = 0.04); this was non-significant after pre- postmenopausal stratification. Overall, differential gene expression analysis showed no significant modulation of genes in either group. Using a pre-specified response parameter (50% relative reduction in Ki67), we identified 12/32 (38%) “responders” in the TPA, and 9/29 (31%) in the placebo arm. In sub-group analysis of these responders, we found 103 genes to be significantly modulated by treatment in the TPA “responders”, but saw no significant change in any gene expression in placebo “responders”. Gene set enrichment analysis for the 103 genes showed that TPA blocked the progression of cell cycle genes (PTTG1, PLK1, UBE2C, HIST1H3F, PSMD3, and etc.) and suppressed PGR and ERBB2 expression. In a pre-planned pooled analysis, these results will be combined with NCT02314156, reported in SABCS abstract 851790. Conclusions: An anti-proliferative (Ki67) signal of TPA was observed in early stage breast cancer patients, but interpretation was limited by placebo group changes. The TPA group demonstrated differential suppression of proliferation-related genes among Ki67 responders, but the placebo group did not. Ongoing analysis will examine signatures related to stemness, metastasis, and immune suppression (potentially better endpoints in trials targeting P signaling). These analyses may help us select the right population and the right biomarkers for future trials. Citation Format: Lee O, Sullivan ME, Xu Y, Shidfar A, Ivancic D, Zeng Z, Singhal H, Helenowski I, Jovanovic B, Hansen N, Bethke K, Gann P, Gradishar WJ, Clare SE, Khan SA. Progesterone receptor (PR) antagonism by telapristone acetate (TPA): A randomized, placebo-controlled phase IIB pre-surgical window trial in women with stage 0-II breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-04-02.

Background/Objectives: The determination of HER2-positivity by IHC or FISH is critical for identifying patients most likely to benefit from anti-HER2 therapy. However, these methods do not always provide an accurate indication of HER2 overactivity, which can occur without gene amplification or overexpression of the HER2 protein. Our study objective was to determine if a deep learning convolutional neural network (CNN) could be trained, using IHC HER2 staining, to learn a morphological signature for HER2 positivity in H&E stained slides. Methods: For training, we used H&E images (whole slides scanned at 40x) from 10 HER+ patients (IHC 3+) and 15 HER2− patients (IHC 0 or 1+) along with their adjacent HER2 IHC images. We first annotated non-cancer regions in H&E images. We then identified tumor regions in HER2 IHC that were positive (intense/complete circumferential stain) or negative (no or weak stain). We ignored any regions with equivocal IHC response as well as whole slides of IHC 2+ patients. For rigorous testing, slides from a separate set of 7 HER2+ and 19 HER2− patients were used. Digitized slides and expert consensus IHC HER2 status for each patient were provided as part of an international HER2 IHC scoring competition organized by the University of Warwick. The computer vision pipeline comprised four-stages. First, we color-normalized the H&E images to reduce unwanted color variation between slides. Second, a pre-trained neural network (NN1) marked all nuclear centroids to make it easier for subsequent stages to focus on nuclear morphology and inter-nuclear spatial arrangements. Third, a neural network (NN2) trained on sub-images of size 100x100 centered at nuclear centroids classified nuclei as non-cancer vs. cancer. Fourth, a final neural network (NN3) sub-classified cancer nuclei into HER2+ or HER−. In H&E images of held-out test patients, the percent of cancer nuclei scored as HER2+ was analyzed. Results: NN2 had a classification accuracy exceeding 97% on a validation set of 25,000 nuclei set aside from training patients, while NN3 had a validation accuracy of 88% on 7,500 test nuclei. On the set aside test patients, among the seven HER2+ patients an average of 49.8% cancer nuclei were scored HER2+, while among the 19 HER2− test patients the corresponding proportion was only 24.7% (p < 0.01). The AUC of binary classification of test patients into HER2+/− based on percent of HER2+ nuclei was 0.815. Upon closer inspection, the H&E morphology of some of the misclassified HER− patients showed visual similarity to the correctly classified HER2− patients and vice versa. Conclusions: Even when morphological patterns associated with cancer subtypes are too subtle for humans to reliably detect, H&E stained slides analyzed by CNNs may be able to geographically map a HER2 signature. It is unclear if misclassification with respect to IHC status is a reflection of morphological confusion or discordance between genomic subtype and IHC response. By training multiple neural networks to detect morphological signatures corresponding to different molecular subtypes of breast cancer, we may be able to detect and study intra-tumor heterogeneity in a cost effective and/or complementary way compared to multi-region sequencing. Citation Format: Dhage S, Anand D, Kumar N, Gann PH, Sethi A. Computer vision detects morphological correlates of HER2 positive breast cancer in H&E stained histological images [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-02-11.

Classification of breast cancers by PAM50 gene profiling assigns each cancer to one of five intrinsic subtypes. However, individual cancers vary in their adherence to the assigned prototype, and some may exhibit expression patterns that indicate admixture with other subtypes. The prevalence of this intratumoral heterogeneity phenomenon and the relationship of subtype admixture to clinical endpoints has not been extensively studied. We obtained microarray data on 1,460 cases from the public Gene Expression Omnibus and processed the data to normalize and log-transform expression levels and control for batch effects. For each case, we extracted levels for PAM50 genes and computed the Mahalanobis distance in multidimensional space for each case from all five centroids, including the one assigned. This metric obtains the standardized distance from a centroid for a single case while accounting for correlation between genes. We then discriminated pure from admixed cases by setting criteria for distance from the assigned centroid as well as distance from nonassigned centroids. Using the criteria that pure cases are those within the 75th percentile of distance from the assigned centroid, while being greater than the 75th percentile from any other centroid, we found that 59% of basal cases were categorized as pure, versus luminal A 31%, luminal B 35%, HER2 34% and normal 27%. We used t-Distributed Stochastic Neighbor Embedding (t-SNE) plots to visualize these admixtures in two dimensions. This confirmed that basal cancers were more heterogeneous than the other types, and that LumA, LumB and normal were more frequently admixed than other combinations. Pure LumA cases were less likely to be high-grade than LumA cases admixed with LumB, basal, HER2 or null cases not near any other centroid (18.5% vs. 7.2%, P < 0.01). Pure basal cases were more likely to be high-grade than basal cases admixed with any other subtype. Kaplan-Meier curves for metastasis-free survival for pure LumA were very similar to admixed cases for the first 5 years, but then diverged, with worse survival when LumA was admixed with worse subtypes. These results support the hypothesis that subtype admixture, or at least the tendency for some breast cancers to exhibit this form of intratumoral heterogeneity, can be identified and that it should be explored for clinical consequences, including failure to respond to subtype-specific therapy. Citation Format: Peter H. Gann, Neeraj Kumar, Dan Zhao, Amit Sethi. PAM50 subtype admixture in individual breast cancers and the relationship of this intratumoral heterogeneity to clinical variables [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1179.