Search

390 Background: PCa has inherited risk factors including high genetic risk variants such as BRCA1/2, HOXB13, and DNA mismatch repair genes. mpMRI has been shown to be effective for detection and staging of localized PCa. This study follows participants (prts), born biologically male, without a diagnosis of PCa with known germline pathogenic or likely pathogenic variants (PV) in BRCA1/2, MLH1, MSH2, MSH6, PMS2, EPCAM, HOXB13, ATM, NBN, TP53, CHEK2, PALB2, RAD51C/D, BRIP1, or FANCA-FANCM (NCT03805919). Methods: Up to 500 eligible prts 30-75 years old (yo) with a documented germline PV will enroll. Prts undergo biennial clinical exam and mpMRI, and annual PSA monitoring and are followed at 12-month intervals to determine PSA, prostate cancer diagnosis, and/or disease/survival status until death. Indication for prostate biopsy includes clinical or imaging findings. Biopsy specimens undergo molecular analyses. Results: To date, 175 prts have been enrolled: 169 (97%) White, 3 Hispanic (2%), 1 African American (1%), 1 Asian (1%), and 1 biethnic (1%). Median age is 47 yo. The most common monoallelic PV are: 48.6% BRCA2, 25.1% BRCA1, 6.3% CHEK2 and 5.7% MSH2. PVs in ATM, PALB2, HOXB13, PMS2, MLH1, MSH6, BRIP1, EPCAM and RAD51D are ≤4%. One subject carries three distinct PVs ( BRCA2, CHEK2, BRIP1). Indication for biopsy was found in 26.3% of prts with 22/46 (47.8%) with a PIRADS 4 lesion, 6/46 (13.0%) PIRADS 3 lesion, 12/46 (26.1%) elevated PSA (median=2.8 ng/mL) or 6/46 (13.0%) due to clinical discretion. Adenocarcinoma was diagnosed on 13/39 (33.3%) biopsies with median age at diagnosis=59 yo. 9/13 (69%) prts had a PSA

e13555 Background: Small cell lung cancer (SCLC) is a highly malignant cancer that commonly metastasizes throughout the body (mSCLC) and is molecularly classified based on expression of four transcription regulators (NAPY): NeuroD1, ASCL1. POU2F3, and YAP1. Medical imaging is used for treatment monitoring and patient management. Due to the relatively rare occurrence of this disease, large cohorts of data are not available for training. The purpose of this work is to develop a model for automated detection and segmentation of mSCLC liver metastases by leveraging publicly available data and characterize its output against NAPY subtyping. Methods: Two datasets were used: (1) the Liver Tumor Segmentation Challenge (LiTS), a public dataset containing 130 CT scans and (2) 23 patients with mSCLC imaged at our institution with biopsy-confirmed SCLC metastasis to the liver and available NAPY subtyping. N = 12 mSCLC scans underwent ground truth lesion segmentation by an experienced radiologist (> 10yrs) using ITK-Snap. The LiTS dataset was split to 80/12/8 for training/validation/testing. The mSCLC data was reserved for testing. CT scans were resampled to uniform z-spacing of 1mm and Hounsfield range. A liver organ segmentation model based on 2D UNet and liver lesion segmentation model based on DeepLabV3+ were trained using Fast.ai library. Random flipping, rotation, erasing, and variable CT windowing were applied in each batch. Hyperparameters included cross-entropy loss, Adam optimization, batch size 32, input size [512,512], 1e-4 learning rate, 100 epochs, pre-trained weights from ImageNet. The final liver volume was selected as largest 3D-connected component of output and was used as input to lesion segmentation. Segmentation performance was evaluated by Dice Similar Coefficient (DSC). AI-based lesion segmentations were used to extract 106 shape and texture features (PyRadiomics) for association to NAPY subtypes using logistic regression analysis accounting for the clustered nature of lesion data on the patient-level. Results: The liver model achieved median DSC 0.952 (range (0.926-0.964) and 0.958 (range 0.933-0.969) in the LiTs validation and testing sets, respectively. Lesion segmentation results in validation and testing sets achieved median DSC 0.639 (0-0.848) and 0.523 (0-0.878), respectively. This model demonstrated acceptable generalizabilty in the mSCLC patients from our institution with median Dice coefficient 0.696 (0.102-0.882). Overall, the model achieved 62.1% sensitivity and 88.2% PPV with median 1 FP/image (range 0-4) from 53 ground-truth lesions. Sensitivity analysis showed lesions > 2mm3 had excellent detection performance. AI-based Maximum 2D Diameter and GLCM Difference Entropy were significantly associated with NA vs. PY subtypes across 29 lesions with available outcomes. Conclusions: AI-based models have potential for CT-based disease characterization of liver metastases in mSCLC.

Purpose Desmoid tumors (aggressive fibromatosis) arise from connective tissue cells or fibroblasts. In general, they are slow growing and do not metastasize; however, locally aggressive desmoid tumors can cause severe morbidity and loss of function. Disease recurrence after surgery and/or radiation and diagnosis of multifocal desmoid tumors highlight the need to develop effective systemic treatments for this disease. In this study, we evaluate objective response rate after therapy with the γ-secretase inhibitor PF-03084014 in patients with recurrent, refractory, progressive desmoid tumors. Patients and Methods Seventeen patients with desmoid tumors received PF-03084014 150 mg orally twice a day in 3-week cycles. Response to treatment was evaluated at cycle 1 and every six cycles, that is, 18 weeks, by RECIST (Response Evaluation Criteria in Solid Tumors) version 1.1. Patient-reported outcomes were measured at baseline and at every restaging visit by using the MD Anderson Symptoms Inventory. Archival tumor and blood samples were genotyped for somatic and germline mutations in APC and CTNNB1. Results Of 17 patients accrued to the study, 15 had mutations in APC or CTNNB1 genes. Sixteen patients (94%) were evaluable for response; five (29%) experienced a confirmed partial response and have been on study for more than 2 years. Another five patients with prolonged stable disease as their best response remain on study. Patient-reported outcomes confirmed clinician reporting that the investigational agent was well tolerated and, in subgroup analyses, participants who demonstrated partial response also experienced clinically meaningful and statistically significant improvements in symptom burden. Conclusion PF-03084014 was well tolerated and demonstrated promising clinical benefit in patients with refractory, progressive desmoid tumors who receive long-term treatment.

206 Background: While active surveillance (AS) has become the preferred management strategy for patients with NCCN very-low and low risk prostate cancer, its use in the setting of intermediate risk disease continues to be controversial. Current guidelines state that AS can be considered in favorable intermediate risk disease, but data on outcomes in this cohort of patients is lacking. We aim to report on our experience with AS of Gleason grade group (GG) 2 at the National Cancer Institute (NCI)/National Institutes of Health (NIH). Methods: Our IRB-approved, institutional database was reviewed for men who enrolled on our AS protocol at NIH with GG2 disease from 2007-2020. All patients received MRI-targeted and systematic biopsy at the time of enrollment, diagnosing or confirming GG2 disease. MRI and PSA were performed annually, and majority of patients underwent annual combined MRI targeted and systematic prostate biopsy. If PSA and MRI were stable, annual biopsy was postponed in several patients based on their preference. Differences in PSA and PSAD at enrollment on AS and at progression were compared using Wilcoxon signed rank test. P value < 0.05 was considered significant. Results: 98 patients with GG2 disease enrolled in AS at NIH. Average age at enrollment was 64 years old and the majority of patients were Caucasian. 36/98(37%) of these patients progressed to GG3 or higher while on AS. Median PSA at time of progression was significantly higher than at the time of enrollment on AS (5.2 [IQR 4.0-8.9] vs 8.5 [IQR 5.8-10.9], z = -3.12, p < 0.01). PSAD was also significantly higher at time of progression (0.12[IQR 0.1-0.16] vs 0.13[IQR 0.10-0.22], z = -2.65, p < 0.01). Highest PIRADS score on MRI was largely unchanged. Median time to progression was 71 months. The majority of patients progressed to GG3, and progression was the trigger for definitive treatment. All patients were alive at the end of the follow up period. Conclusions: AS is a reasonable option for compliant patients diagnosed with GG2 prostate cancer. While a significant percentage of these men will progress on AS, they do so at a median of 71 months, avoiding treatment-related harms of definitive therapy for over 5 years. Further research with larger sample sizes are needed to better evaluate the oncologic outcomes of AS for GG2 disease, as well as predictors of more aggressive disease that may be better served with upfront definitive treatment. [Table: see text]

207 Background: Active surveillance (AS) is considered the standard of care for managing patients with low-risk prostate cancer. Identifying risk factors for disease progression on AS can improve risk stratification. As previously described, the presence of bilateral disease may affect disease progression. Current systematic biopsy techniques inadequately predict presence of bilateral disease. We hypothesize that combined mpMRI-guided fusion biopsy and systematic biopsy will improve detection of bilateral disease. Methods: Our institutional database of patients referred for AS from 2007-2020 was queried. All AS patients received combined 12-core systematic biopsy and mpMRI-Fusion biopsy (Cbx), and those with Gleason Grade group 1 (GG1) on confirmatory Cbx were included. Cox regression analysis identified baseline characteristics associated with AS progression, defined as progression to GG ≥ 2. The distributions of GG and National Comprehensive Cancer Network (NCCN) risk categories were also compared. Results: 103 patients with prostate cancer and at least one follow-up biopsy at our institution were included in the analysis. 30% (31/103) had bilateral disease and 70% (72/103) had unilateral disease. On univariable and multivariable analysis, both PSAD (univariable HR = 1.6; 95% CI: 1.1-2.2; multivariable HR = 1.7; CI: 1.2-2.5) and presence of bilateral disease (univariable HR = 2.2; 95% CI: 1.3-4.0; multivariable HR = 2.2; 95% CI: 1.2-4.0) at time of confirmatory biopsy were significantly associated with AS progression. On Kaplan-Meier analysis, median time from Cbx to AS progression ≥ GG2 for the whole cohort was 52 months (95% CI: 44 - 68). The median time to progression for patients with unilateral and bilateral disease was 68 months and 52 months, respectively (log-rank test, p = 0.0055). Bilateral disease was marginally associated with receiving an NCCN score of high or very high on upgraded biopsy (RR: 3.16, 95% CI: 1.004-9.932). NCCN score at time of confirmatory biopsy was not associated with progression. However, NCCNs scores and GG at time of progression were higher among patients with bilateral disease (p = 0.015 and p = 0.024). Ultimately, use of combined biopsy resulted in 16.1% more bilateral cancers detected when compared to systematic biopsy. Conclusions: Detection of bilateral disease using combined biopsy was significantly associated with disease progression in patients on AS. Notably, combined biopsy resulted in greater detection of bilateral disease than systematic biopsy alone. Identifying presence of bilateral disease using combined biopsy could ultimately contribute to management decisions in patients on AS.

208 Background: Following treatment of prostate cancer with radical prostatectomy (RP), biochemical recurrence (BCR) can be detected with elevated PSA. This may be attributed to either cancer recurrence or retained benign prostatic gland tissue. Options for detecting malignancy after RP currently entail diagnostic imaging and biopsy with transrectal ultrasound (TRUS). TRUS alone has limited accuracy in detecting recurrence in the prostate bed. MRI fusion-guided biopsy (Fbx) may be a more accurate method of detecting post-RP local recurrence. We hypothesize that Fbx for diagnosing benign versus malignant recurrence in the prostate bed is feasible and produces clinically meaningful results. Methods: Our institutional database was queried for patients who received RP and demonstrated BCR between February 2015 and July 2020. All patients with evidence of prostate bed recurrence on mpMRI were included in this analysis. Cancer detection via mpMRI-guided fusion biopsy using the UroNav platform was evaluated and patient variables including final Gleason Grade group (GG), margin involvement, PSA at BCR, and prostate bed lesion size were analyzed with univariate logistic regression. Results: 40 patients (median age = 68) with BCR underwent post-RP mpMRI. 25/40 (62.5%) patients had MRI-visible lesions, and among those, 17/25 (68%) patients underwent Fbx of the prostate bed. 15/17 (88.2%) Fbxs detected prostate tissue (either benign or cancer), 11/17 (64.7%) contained cancer, and 4/17 (23.5%) contained benign prostate glands. Median cores per biopsy was 4 (IQR 4-6). Among the 83 cores obtained, 57 (68.6%) cores contained prostate gland tissue and 26 (31.3%) contained fibromuscular tissue. Of those 57 with gland tissue, 33 (57.9%) cores contained cancer, and 24 (42.1%) contained benign prostate tissue. Among patients with benign biopsies, none had further evidence of metastasis at median follow-up of 13.5 months after Fbx and 182 months after RP. On final RP pathology, 4 patients had GG1 disease, 4 had GG2, 4 had GG3, 2 had GG4, and 3 had GG5. 6/17 (35%) patients had positive RP margins. Median prostate bed lesion size was 1.3 cm (IQR 0.9-1.5). Prostate bed lesion size (cm) was the only variable significantly associated with cancer on Fbx (OR = 2.20, 95% CI:1.29-3.76, p = 0.011). Conclusions: mpMRI-Fbx is a feasible method for reliably targeting prostate bed lesions. With this technique, we found improved accuracy for biopsy-proven recurrence in the prostate bed. This technique will help direct treatment planning of salvage therapies among patients with detectable PSA post-RP. [Table: see text]

Active surveillance (AS) has emerged as a standard management option for men with very low-risk and low-risk prostate cancer, and contemporary data indicate that use of AS is increasing in the United States and abroad. In the favorable-risk population, reports from multiple prospective cohorts indicate a less than 1% likelihood of metastatic disease and prostate cancer–specific mortality over intermediate-term follow-up (median 5–6 years). Higher-risk men participating in AS appear to be at increased risk of adverse outcomes, but these populations have not been adequately studied to this point. Although monitoring on AS largely relies on serial prostate biopsy, a procedure associated with considerable morbidity, there is a need for improved diagnostic tools for patient selection and monitoring. Revisions from the 2014 International Society of Urologic Pathology consensus conference have yielded a more intuitive reporting system and detailed reporting of low-intermediate grade tumors, which should facilitate the practice of AS. Meanwhile, emerging modalities such as multiparametric magnetic resonance imaging and tissue-based molecular testing have shown prognostic value in some populations. At this time, however, these instruments have not been sufficiently studied to consider their routine, standardized use in the AS setting. Future studies should seek to identify those platforms most informative in the AS population and propose a strategy by which promising diagnostic tools can be safely and efficiently incorporated into clinical practice.

e17567 Background: Patients diagnosed with prostate cancer undergo computed tomography (CT) for pretreatment staging to rule out bone metastases. However, detection and classification of bone lesions on CT is challenging and subject to inter-reader variability. We present a cascaded deep learning algorithm for automatic detection and classification of bone lesions on staging CT in patients diagnosed with prostate cancer. Methods: CT scans from 56 patients with histopathologically proven prostate cancer were included. An expert radiologist annotated the extent of individual bone lesions (N = 4217) and labelled all regions as either benign or malignant. All scans were anonymized and normalized at the patient-level prior to training. Our method can be described as a two-stage framework, 1) A detection algorithm: Inspired by Yolo-v3 detection method, we designed a network with a backbone of darknet-53 pretrained on Coco dataset and four final scaling blocks to compensate for wide range of lesion diameters, 2) A classification algorithm: we formed a binary classifier based on ResNet-50 pretrained using the ImageNet dataset. We used a train/validation split equal to 90%/10% for this study. To facilitate the learning process, horizontal flipping, relative zooming and mean weighted averaging were used for data augmentation in stage 1. Instead, the classification algorithm took advantage of synthesized patches generated by Deep Convolutional Generative Adversarial Network (DC-GAN) for augmentation. Results: We could achieve a real-time (~120ms per slice) performance on our validation set with a median penalty of 0.3(0.02-0.78) false positives per true positive within each patient. Overall performance of our detection algorithm was 81% sensitivity and 86% positive predictive value. In stage 2, we obtained an accuracy of 89% for correct classification of benign from malignant bone lesions with no augmentation which was improved to 91% when we incorporated the augmented data for training. Conclusions: Our 2-stage algorithm sequentially detects and classifies bone lesions on CT of prostate cancer patients with a significant performance. To further improve our results and for generalizability we are accruing more data from different centers. Eventually, with greater dataset, both algorithms will be cascaded and trained as a whole unit to become one single tool for fully automatic detection and classification which serves as an aid for radiologists who read the staging CTs.

293 Background: Active surveillance (AS) is now considered a well-accepted alternative for low-favorable intermediate risk prostate cancer over definitive therapy. Few studies have incorporated the use of multi-parametric MRI (mpMRI) into the treatment paradigm. In this study we investigate imaging findings that are predictive of a patient dropping off AS. Methods: Our institutional database was queried for all patients who met criteria for active surveillance from 11/2003 to 5/2017. Criteria for inclusion included ≥ 2 mpMRIs, ≥ 2 prostate biopsies, and a diagnosis of Gleason Grade group (GG) 1 or higher. Patients were excluded if they received any other therapy for the treatment of their prostate cancer such as radiation, chemotherapy, focal therapy, or immunologic therapy. Patient demographics, mpMRI, biopsy and most recent follow-up data were recorded. Factors, including PSA density (PSAD), PSA, lesion size, and PI-RADS category, associated with AS progression were evaluated in Cox Proportional Hazards Model. Results: An analysis of a total of 212 patients were performed during the study time interval. 88 patients were dropped from AS during this time and of those patients the median amount of time before removal was 4.70 years (range, 0.7-10.5). On univariable analysis, PI-RADS category (HR, 1.302 for every increase in 1 unit of the PI-RADS category; 95% CI, 1.046-1.62, p = 0.01) and PSAD (HR, 4.98 for every increase in 0.001 ng/mL/cc; 95% CI, 2.127-11.66; p < 0.001) were found to be associated with being removed from AS. On multivariable analysis, both PI-RADS score (HR, 1.281 for every increase in 1 unit of the PI-RADS category; 95% CI, 1.025-1.6; p = 0.003) and PSAD (HR, 4.188 for every increase in 0.001 ng/mL/cc; 95% CI, 1.640-10.7; p < 0.001) remained associated with being removed from AS. Conclusions: PI-RADS categories and PSAD predict the risk of a patient to drop off active surveillance. AS. Patients with these criteria should be considered high risk in any current AS protocol.[Table: see text]

368 Background: PCa is driven by androgen receptor (AR) signaling and neoadjuvant therapy with AR inhibitors offer an opportunity to improve cure rates in high-risk PCa particularly with utilization of multiparametric MRI (mpMRI). A loss of AR-regulated lineage characteristics and genomic loss of tumor suppressors RB1 and TP53 or mutations in DNA damage repair (DDR) genes can represent aggressive prostate variants. We conducted a feasibility study using mpMRI to evaluate tumor responses and resistance in newly diagnosed, high-risk PCa (NCT02430480). Methods: Pts were treated with androgen deprivation therapy (ADT) + enzalutamide (enza) 160 mg daily for 6 months (mos). Pts underwent 2 mpMRIs: baseline and post 6 mos treatment (trt). Post-trt mpMRI was followed by radical prostatectomy (RP). Primary endpoint: feasibility of mpMRI for localization and detection of PCa before and after ADT + enza. Results: 39 pts were enrolled on-study with 36 pts completing 6 mos trt and undergoing RP. Of 39 pts, 3 had disease progression. Conclusions: Neoadjuvant intense testosterone lowering therapy shows activity in PCa but a subset of pts not respond to AR-targeted therapies through lineage plasticity enabled by characteristic loss of RB1 and TP53 or due to genetic alterations. Identification of this high-risk patient population, along with development of treatment options, needs further investigation. Clinical trial information: NCT02430480. [Table: see text]

342 Background: The differential aggressiveness of potentially independent prostate cancer clones remains largely unknown. Appropriate prostate cancer staging using mpMRI and biopsy tissue can be confounded by sampling error. To date, there has been no understanding of whether clonal variability influences management decisions for localized prostate tumors. We sought to identify the sensitivity and genomic profile of distinct localized tumors from a patient following systemic intense neoadjuvant androgen deprivation therapy (ADT). Methods: A 66-year-old man with high risk prostate cancer enrolled in a Phase 2 study of intense neoadjuvant ADT (goserelin + enzalutamide; inADT). Baseline mpMRI showed a single semi-contiguous lesion encompassing the right apical-mid PZ extending into the left distal apical PZ. MR/US-fusion targeted biopsy was performed before 6 months of inADT. A second mpMRI was performed before radical prostatectomy. Whole exome sequencing on microdissected tumor foci identified somatic mutations and copy number alterations, which were further used with immunohistochemistry to assess tumor clonal architecture and genomic/phenotypic evolution of treatment resistant tumor. Results: We found two clonally independent tumors exhibited intrinsic heterogeneity at baseline which correlated with response or resistance. Biopsies of distinct left- and right-sided tumors showed differing histologies. mpMRI and pathology showed near complete response of the left-sided tumor and substantial resistance of the right-sided tumor, which exhibited a large intraductal component. Histology and whole exome data highlighted a divergence in the status of PTEN and TP53, tumor suppressor genes implicated in prostate cancer progression. Conclusions: These data highlight that even nascent prostate cancer is heterogenous and neoadjuvant therapeutic strategies will need to consider this for clinical optimization. Evolutionary trajectories that resulted in tumor heterogeneity in this case likely contributed to our observation that two independent prostate tumor nodules with distinct genetic alterations responded differently to neoadjuvant intense ADT. Clinical trial information: NCT02430480.

283 Background: MRI/US fusion guided prostate biopsy (FBx) has been shown to detect clinically significant prostate cancer (csCaP) at higher rates and with fewer cores than standard prostate biopsy. Size plays an important role in assigning a suspicion level (PI-RADS) to lesions identified on MRI. However, tumor characteristics may pose challenges to accurately characterizing the lesion despite the size. This study sought to determine if there are size cutoffs at which a lesion may be accurately characterized as clinically significant cancer with a single biopsy core. Methods: A retrospective analysis of a prospectively maintained database of all patients undergoing FBx at an academic referral center between May 2014 and January 2018 was conducted. At least two FBx cores were taken from each lesion identified on mpMRI. GEE-based univariate logistic regression model with exchangeable correlation was used determine if size was a significant predictor of positive and negative agreement. Predictability of size as a significant continuous predictor was quantified by AUC. Size thresholds at which multiple cores per lesion are needed to avoid missing > 2% of csCaP were calculated, allowing for a 25% discordance rate. Results: An analysis of a total of 1141 FBx of 2200 lesions was performed during the study time interval. Size was a significant predictor of both positive (OR = 2.43, 1.83-3.23, p < 0.01) and negative (OR = 0.58, 0.44-0.76, p < 0.01) agreement of csCaP. AUC% for positive and negative agreement was 65.8 and 57.6, respectively. Size thresholds of 0.65 and 1.70 cm limited CS cancers missed by skipping a second targeted biopsy core to 2% while allowing for a 25% discordance. Conclusions: These data indicate that clinically significant prostate cancer in lesions less than 0.65 cm and greater than 1.70 cm may be characterized with a single targeted biopsy core, sparing 33.5% of lesions (21% patients) a double core targeted biopsy.

294 Background: PTEN loss is associated with adverse outcomes in prostate cancer and has the potential to be clinically implemented as a prognostic biomarker. Deep learning algorithms applied to digital pathology can provide automated and objective assessment of biomarkers. The objective of this work was to develop an artificial intelligence (AI) system for automated detection and localization of PTEN loss in prostate cancer samples. Methods: Immunohistochemistry (IHC) was used to measure PTEN protein levels on prostate tissue microarrays (TMA) from two institutions (in-house n=272 and external n=125 patients). TMA cores were visually scored for PTEN loss by pathologists and, if present, spatially annotated. In-house cohort (N=1239 cores) were divided into 70/20/10 training/validation/testing sets. Two algorithms were developed: a) Class I=core-based, to label each core for biomarker status and b) Class II=pixel-based, to spatially distinguish areas of PTEN loss within each core. ResNet101 architecture was used to train a multi-resolution ensemble of classifiers at 5x, 10x, and 20x for Class I task and a single classifier at simulated 40x for Class II segmentation. Results: For Class I algorithm, accuracy of PTEN status was 88.3% and 93.4% in validation and testing cohorts, respectively (Table). AI-based probability of PTEN loss was higher in cores with complete loss vs partial loss. Accuracy was improved to 90.7% in validation and 93.5% in test cohorts using the Class II region-based algorithm, with median dice scores 0.833 and 0.831, respectively. Direct application to external set demonstrated a high false positive rate. Loading trained model and conservatively re-training (“fine-tuning”) on 48/320 external cohort cores improved accuracy to 93.4%. Conclusions: Results demonstrate feasibility and robustness for fully automated detection and localization of PTEN loss in prostate cancer tissue samples and possibility for time/cost-effectiveness of sample processing/scoring in research and clinical laboratories.[Table: see text]

TPS4159 Background: Pheochromocytoma/paraganglioma (PHEO/PGL) is a rare malignancy that arises from chromaffin cells of typically the adrenal medulla but can also be of extra-adrenal origin. These tumors produce excessive catecholamines such as epinephrine and norepinephrine which causes labile hypertension, tachycardia, and flushing. Lu-177-DOTATATE (Lu-177-dodecanetetraacetic acid‐tyrosine-3-octreotate) is a radiolabeled somatostatin analog that is FDA approved for somatostatin receptor-positive neuroendocrine tumors. It is being being investigated in PHEO/PGL, which overexpress somatostatin receptors. Amino acid solutions containing lysine/arginine (L/A) are routinely co-administered with Lu-177-DOTATATE for renal radioprotection, although solutions containing other amino acids are also used. Methods: This is a prospective, single center, open label Phase 2 study evaluating the efficacy of Lu-177-DOTATATE in PHEO/PGL. Ninety patients will be enrolled, divided into two cohorts of 45 patients each ( SDHx mutation vs. apparent sporadic). Lu-177-DOTATATE is given at a fixed dose of 200 mCi with a co-administration of L/A amino acid solution q8 weeks for 4 cycles. The primary endpoint is the progression-free-survival (PFS) rate at 6 months. Secondary endpoints include response rate, overall survival, time to progression, quality of life measures, and examination of potential biomarkers such as biochemical profiles, Ga-68-DOTATATE PET, and F-18-FDG PET scans. Eligibility criteria include inoperable disease (including non-metastatic), histological confirmation of PHEO/PGL, evidence of disease progression by RECIST 1.1, ECOG performance status of 1 or better, and a positive Ga-68-DOTATATE PET scan. Exclusion criteria include prior treatment with systemic radionuclide therapy such as I-131-MIBG, brain parenchymal metastases, and standard organ dysfunction limitations. Interim analysis using a Simon two-stage optimal design will be performed separately for each cohort after enrollment of 18 patients. First patient accrual to this ongoing study was in the Fall of 2017, and as of February 2019, fourteen patients have been accrued. Preliminary results will be reported at the completion of stage 1 for each cohort. Clinical trial information: NCT03206060.

101 Background: Prostate multi-parametric magnetic resonance imaging (mpMRI) can precisely depict prostate cancer (PCa) location and adverse pathologic features. Surgeons can utilize this information to maximize sparing of the neurovascular bundles (NVBs) during radical prostatectomy (RP) while avoiding a positive surgical margin (PSM). We detail the technique of using preoperative mpMRI to quantify its effect regarding nerve-sparing and rates of PSMs. Methods: A prospectively maintained database was queried for robotic-assisted RPs (RARPs) with preoperative mpMRI between 2007-2017. Imaging margin risk factors (iMRF) were defined on mpMRI as frank extraprostatic extension (EPE), possible EPE, and capsular irregularity (capsular bulge, lesion-capsule contact, or lesion adjacency to the neurovascular bundles). Surgical adjustments to nerve-sparing technique (full sparing, partial sparing, and wide excision) were made based on these findings. Results: Five hundred thirty-two patients comprising 1041 prostate sides were included for analysis. Overall, PSM rate was found in 80/1041 (7.7%) sides of the prostate. iMRF were seen in 313/1041 (30.1%) prostate sides, for which adjustments were made in 244/313 (78.0%) of these. In the 69/244 (22.0%) cases where full nerve-sparing was performed despite iMRF, PSM rate was 20/69 (29%) compared to 33/244 (13.5%), p = 0.002. MRI-guided surgical adjustments decreased PSM risk by 68% and 15% in pT3 and pT2 cases, respectively. On multivariable analysis, logPSA (odds ratio [OR] 4.06, [95% CI 2.40-12.3], p < 0.001) and iMRF (OR 1.78, [95% CI 1.01-3.16], p = 0.047) were significantly associated with PSM while nerve-sparing adjustment was significantly associated with decreased risk of PSM (OR 0.38 [95% CI 0.22-0.66], p = 0.001). Conclusions: MRI effectively detects risks for PSM and guides surgical adjustments to decrease PSM rates. As prostate MRI is more frequently acquired for PCa screening and biopsy, we show its additional value for RP planning and potentially improved outcomes.

108 Background: MRI fusion prostate biopsy has been shown to improve detection of clinically significant prostate cancer, however the degree of this benefit is poorly characterized in large clinical trials. Methods: 1750 MRI targeted plus sextant biopsies were performed in 1742 male patients from 2007 to 2017. Patient demographics, PSA, prostate volume, primary and secondary Gleason scores, Johns Hopkins Grade Groups, number of MRI targeted lesions, number of cores obtained, and biopsy yield were recorded. Results: The patient population consisted of men averaging 62.9-year-old (36-86) with a mean PSA 9.6ng/mL, and prostate volume of 59.2 ml. A total of 804 cancers were detected on sextant biopsy and 839 were detected on MRI targeted biopsy. Relative to targeted biopsy, sextant biopsy detected only significantly more Gleason 6 disease (14% vs 21.5%, p < 0.0001) than targeted biopsy. Targeted biopsy detected more Gleason 7 (21% vs 16.6%, p = 0.0009) and Gleason 8-10 (13.4% vs 9.4%). Additionally, Gleason 7 sub-stratification demonstrated substantially more Gleason 4+3 detection in targeted group vs sextant biopsy (4% vs 0.5%, p < 0.0001). When stratified by Grade Group targeted biopsy detected 76% more Grade Group 3-5 cancers (p < 0.0001) and 17.7% less Gleason Group 1-2 cancers (p < 0.0001). Only 1.7% of Grade Group 3-5 cancers were detected on sextant biopsy alone, where as 15.7% of Grade Group 3-5 cancers were detected on targeted biopsy alone. Conclusions: MRI targeted biopsy significantly increases the likelihood of detecting clinically significant cancer and decreases the risk of indolent cancer detection. These finding strongly support the use of MRI targeted biopsy when possible.

63 Background: Neoadjuvant therapy with novel androgen receptor inhibitors, like enza, offer an opportunity to improve cure rates in men with high risk PC but also emphasizes the need for improved imaging techniques and genomic studies to evaluate treatment (trt) responses. We conducted a feasibility study using mpMRI to evaluate tumor responses and resistance in newly diagnosed, high-risk PC. Methods: Patients (pts) were treated with androgen deprivation therapy (ADT) + enza 160 mg po qdaily for 6 months (mos). Pts underwent 2 mpMRI: baseline and post 6 mos of therapy. Post-trt mpMRI was followed by radical prostatectomy (RP). Primary endpoint was feasibility of mpMRI for localization and detection of PC before and after therapy with ADT + enza. Results: 31 out of 33 pts completed 6 mos of therapy and underwent RP. Median on-study PSA was 9.58 (1.18-984.72 ng/dL). Median PSA post 6 mos of ADT + enza was

110 Background: Epidemiologic data suggests a link between obesity and increased risk of aggressive prostate cancer (PCa). Therefore, the accurate detection of prostate cancer is paramount in obese men. We sought to characterize the PCa lesions seen on multiparametric magnetic resonance imaging (mpMRI) in both obese (OB) and non-obese (nOB) patients. In addition, we examined whether obesity affects the accuracy of fusion biopsy (FBx) versus systematic biopsy (SBx). Methods: Data from patients who underwent FBx, SBx, or both, and subsequent prostatectomy at a single institution between 2007-2017 were reviewed. Men were stratified into three groups based on body mass index (BMI): normal weight (18.5-24.9), overweight (25.0-29.9), and obese (≥30.0). mpMRIs were analyzed to determine prostate size, lesion size, lesion location, and evidence of ≥T3 disease. FBx and SBx were compared to prostatectomy specimens. These variables were then analyzed between the OB and nOB groups. Results: 487 patients were included: 109 (19%) normal weight, 266 (48%) overweight, and 185 (33%) obese. No statistical differences were noted in age, PSA, or clinical stage. mpMRI showed similar prostate size, index and total lesion diameter, lesion location, and Prostate Imaging Reporting and Data System (PIRADS) score between BMI categories. In OB patients, FBx was better able to predict final Gleason score (GS) than SBx, 60% vs 47%, p = 0.0474, OR 1.67 (1.00-2.80). The rate of upgrading after SBx was higher for OB men versus nOB men, 44% vs 34%, p = 0.034, OR 1.52 (1.01-2.29). However, risk of upgrading returned to that of nOB men with FBx, 20% vs 17%, p = 0.538, OR 1.21 (0.658-2.23). Conclusions: FBx is a more accurate measure of the true GS in OB patients than SBx, and there is a significantly greater risk of underrepresenting PCa on SBx in OB patients. These findings were not associated with differences in prostate sizes and lesion characteristics on mpMRI. This suggests that obesity may cause technical challenges with prostate biopsies that may be overcome by improved visualization and targeting of FBx.

104 Background: MRI/US fusion guided prostate biopsy (FBx) has been shown to detect clinically significant prostate cancer (csCaP) at higher rates and with fewer cores than standard prostate biopsy. However, the number of targeted cores needed to accurately characterize lesions identified on multiparametric MRI (mpMRI) is unknown. This study sought to determine factors that predict the number of cores needed to accurately characterize lesions during FBx of patients on active surveillance. Methods: A retrospective analysis of a prospectively maintained database of all patients undergoing FBx at an academic referral center between May 2014 and January 2018 was conducted. At least two FBx cores were taken from each lesion identified on mpMRI. Patient and lesion specific factors were analyzed to determine factors that predict the necessity to obtain additional cores to detect csCaP. GEE-based univariate logistic regression model with exchangeable correlation was used to estimate the effects of clinical characteristics including race, BMI, PSA, PSA density (PSAD), lesion location, and PI-RADS score on the proportion of positive and negative agreement. Predictability of a significant continuous predictor was quantified by AUC. The most significant patient-level predictor (PSAD) was further analyzed to determine thresholds at which multiple cores per lesion are needed to avoid missing csCaP. Results: An analysis of a total of 1141 FBx were performed during the study time interval. PSA (OR=1.57, 1.20-2.05, p2 or greater than 0.26 ng/ml2, lesions may be acceptably characterized with a single targeted biopsy core.

61 Background: AR directed therapies are standard of care for metastatic prostate cancer, and their use for high-risk localized disease may improve survival. Previous trials showed that suppression of AR activity resulting in greater pathologic responses improved survival rates. However, the underlying biology and molecular features of exceptionally responding tumors remain unknown. Methods: Men with locally advanced prostate cancer (median Gleason score 8, median PSA 9.58 ng/dl) received mpMRI and MRI/US-fusion targeted biopsies at baseline. Patients received 6 months of ADT plus enzalutamide, then received a second mpMRI, and underwent radical prostatectomy. RP specimens were whole mounted in the same plane as MRI. The location of each tumor on biopsy was mapped to prostatectomy specimens using pathologic and imaging hallmarks. FFPE sections of baseline and post-treatment tumor were stained, laser capture microdissected, and analyzed using whole exome sequencing. Amongst nonresponding patients (residual tumor burden > 0.05 cc) comparison of pre- and post-treatment tumor specimens identified evolutionary phylogenies over space and time representing intrinsic or clonal selection of pre-existing cells. Amongst responding patients, clonal extinction was confirmed by sequencing and immunostaining of surgical specimens. Results: Gleason score at baseline did not differentiate responding and nonresponding patients. Focal ERG staining in biopsies was absent in 100% of responders and present in 60% of responders. Intraductal carcinoma architecture was also absent in all responding tumors. Focal PTEN loss was observed in all nonresponders at baseline. Responding tumor foci were enriched for deletion of chromosome 6q. Comparison of baseline mpMRI sequences indicated a trend towards lower variance on dynamic contract enhanced (DCE) MR and lower local texture heterogeneity metrics in responding lesions. Conclusions: Response to neoadjuvant intense ADT correlates with distinctive pathologic, molecular, and imaging features that can be observed prior to treatment. Selection of patients based on these parameters may improve overall responses to treatment in subsequent clinical trials. Clinical trial information: NCT02430480.

5082Background: Neoadjuvant androgen deprivation therapy (nADT) with abiraterone and prednisone for localized disease leads to pathological complete response (pCR) rates of up to 10%. In this study...

121 Background: As the use of MRI in the diagnosis of prostate cancer (PCa) becomes more established, knowledge of its utility as a prognostic tool is needed to better counsel patients. Here, we aim to assess the ability of mpMRI to predict adverse pathologic factors, such as extraprostatic extension (EPE), seminal vesicle invasion (SVI) or lymph node involvement (LNI) on final pathology after radical prostatectomy (RP). Methods: A prospectively maintained database was queried for all men who underwent mpMRI with current PI-RADS v2 scoring system, MRI/TRUS fusion biopsy (FBx), and RP at our institution (n = 112). Patient demographics, clinical data, imaging and pathology were recorded. Patients were stratified by presence of adverse pathology (AP), defined as EPE, SVI or LNI. Statistical analyses were used to identify mpMRI characteristics predictive of AP using STATA 13. Results: Patients with adverse pathology were middle aged, predominantly caucasian, and significantly higher PSA at baseline (Table 1). Having a PI-RADS 5 lesion was positively associated with AP (adjusted OR = 4.37, 95% CI = 1.20-15.87, p = 0.02). Mean lesion diameter on mpMRI was greater in patients with AP (p < 0.01). However, number of lesions (p = 0.12) and prostate volume on mpMRI (p = 0.88) were not associated with AP. Conclusions: The results of this study demonstrate that features on mpMRI, particularly a PI-RADS score of 5 and large lesion diameter, are predictive of adverse pathology. This information will prove valuable in developing an imaging-based prognostic model in the era of MRI and FBx. This research was supported by the Intramural Research Program of the National Cancer Institute, NIH and NIH Medical Research Scholars Program[Table: see text]

117 Background: Extended sextant systematic prostate biopsies have the inherent risk of under-sampling prostate cancer. Fusion guided multiparametric magnetic resonance imaging (mpMRI) biopsies have been employed to better represent the disease and guide treatment. We sought to determine if due to heightened suspicion of cancer and/or visualization of mpMRI there were any discrepancies between systematic biopsies done with a Fusion System as compared to those done without (TRUS alone). Methods: From a prospectively collected database, we performed a review collecting age, race, clinical stage, PSA, and time until repeat systematic biopsy as part of fusion guided biopsy (IB). We also collected pathology results reported as Gleason Score (GS), for both the patients’ OB and our IB. Patients were stratified into groups based on time between OB and IB/fusion biopsy ( < 6 months, < 1 year and < 2 years). Results: 69 patients with a previous OB underwent combined fusion and IB within our designated time intervals. Cancer detection rates between the OB and IB results were similar at 6 months, 1 year and 2 years (80 vs 90%, 87.5 vs 87.5% and 65 vs 69%). Detection rates of GS ≥ 3+4 were higher with IB within 12 months compared with IB from 12-24 months (72.7 vs 40.9%, p = 0.03 OR 3.85 (1.09-13.66). Of the patients who were upgraded (n = 24), 54.2% (n = 13) went from benign pathology to a diagnosis of prostate adenocarcinoma. Of all OB GS 3+3 (n = 31), 29% were restaged to higher risk disease on IB. Rates of IB upgrading were similar within 6 months, 1 year and 2 year, 40%, 33.33% and 31.91%). Patients who were upgraded on IB compared to those who were not upgraded were of similar age (67.0 ± 6.53 vs 66.50 ± 6.47), race (17.4% African-American vs 13%), PSA (7.60 ± 5.61 vs 7.50 ± 4.39) and prostate volume on MRI (51.30 ± 26.87 vs 59.26 ± 38.52). Conclusions: A systematic biopsy at our referral center during a mpMRI fusion biopsy was over 3.5 times more likely to detect GS ≥ 3+4 when done within 1 year of the outside biopsy. There continued to be a risk, 34.7% overall, of disease upgrading in all time periods. This research was supported by the Intramural Research Program of the National Cancer Institute, NIH

124 Background: MRI−TRUS fusion biopsy (FBx) use in the diagnosis of prostate cancer (PCa) results in a more accurate assessment of disease burden and has increasingly been incorporated into urologic practice. In addition, with more men choosing active surveillance (AS) and the reports of increased PCa aggressiveness with obesity, we wanted to study the impact of obesity on the risk of PCa progression in men on AS diagnosed and followed by MRI and MRI−TRUS FBx. Methods: A retrospective review was performed on a prospectively maintained database of all men who underwent MRI−TRUS FBx at our institution from January 2007 to May 2015. Patient demographics, clinical data, imaging, pathology, treatment and outcomes were recorded. Patients who enrolled on AS were stratified by BMI into normal weight (BMI 18.5−24.9), overweight (BMI 25.0−29.9), and obese (BMI ≥ 30.0). Statistical analysis was performed using SPSS software. Results: 204 men were enrolled in AS. Within the AS cohort, 51 (25%) had a normal weight, 101 (49.5%) were overweight, and 52 (25.5%) were obese. Age, BMI, PSA and mean estimated progression free survival time are described for each of these groups in Table 1. The overall rate of progression was 32.8%. Of the patients who progressed, 18 (26.9%) were normal weight, 32 (15.7%) were overweight and 17 (25.4%) were obese. On multivariate analysis, BMI was not a risk factor for AS progression, HR = 1.00 (p = 0.99, 95% CI = 0.95−1.06). Conclusions: There is evidence of increased risk of aggressive PCa specific death in obese patients. However, we demonstrate that in patients diagnosed by FBx, obesity does not confer an additional risk of progression on AS. This may be due to the improved characterization of cancer volume and grade by MRI−TRUS fusion biopsy. Further study is required to determine risk factors for AS progression in patients undergoing FBx. This research was supported by the Intramural Research Program of the National Cancer Institute, NIH, Medical Research Scholars Program.

94 Background: Intensive androgen-targeted therapy improves survival for patients with androgen naive metastatic prostate cancer. Neoadjuvant androgen deprivation therapy (nADT) with abiraterone for localized disease leads to complete responses (pCR) rates of up to 10%. In this study we tested the efficacy of nADT with enzalutamide for patients with high risk, localized disease and the ability of post-treatment multiparametric MRI (mpMRI) to evaluate response. Methods: A single institution trial (NCT02430480) evaluated nADT and enzalutamide 160 mg daily in patients with high risk non-metastatic prostate cancer. Patients underwent MRI-TRUS-guided fusion biopsy at screening and repeat mpMRI followed by radical prostatectomy (RP) after completing six months of therapy. Results: Twenty patients have completed therapy and undergone radical prostatectomy. Clinical characteristics were typical of those with high risk disease: median PSA 8.96 ng/ml (range 2.07-985.9) and clinical stage by mpMRI cT2 (one patient), cT3 (17 patients), or cT4 (2 patients); 7 patients had enlarged pelvic lymph nodes. Following 6 months of therapy, median pre-RP PSA was < 0.02 ng/ml (range < 0.02 – 0.35). On final pathology, 0 patients were upstaged, 5 (25%) were unchanged, and 15 (75%) were downstaged, including 3 (16%) pCR. Four patients were downstaged from N1 to N0. There was little correlation between post-treatment mpMRI tumor volume and pathologic tumor volume. Two patients had radiographic complete responses, one of whom had extensive residual disease. Conclusions: Six months of nADT with enzalutamide has activity in high risk, localized prostate cancer, with a small number of patients having exceptional responses. Standard analysis of mpMRI does not reliably characterize depth of response. Evaluation of molecular characteristics that predict exceptional response or intrinsic resistance is on-going. Patients are being followed for recurrence. Clinical trial information: NCT02430480.

108 Background: Many patients with low-risk prostate cancer are encouraged by their physicians to pursue active surveillance (AS). AS has increasingly been utilized, however there remains anxiety by patients and their physicians that more aggressive disease will be missed and allowed to progress. African-American (AA) patients may present with more aggressive disease and higher rates of upgrading at the time of radical prostatectomy. Due to these factors, physicians may be hesitant to recommend AS to AA patients. We examined the role of AS in these patients in the era of MRI targeted biopsies. Methods: A prospectively maintained database was queried for all patients who underwent an MRI guided fusion biopsy from 2007 to 2016 and chose AS as their primary management strategy. Patents with Gleason Group (GG) 1 or 2 were eligible. Patients were then followed with yearly PSA, exam, MRI and biopsy if warranted. MRI Fusion biopsies were reviewed to determine any GG progression. Results: A total of 19 AA and 143 non-AA patients were reviewed with median follow up times of 31.63 (15.42 -89.50) and 30.87 (3.45 – 99.85) months, respectively. AA and non-AA patients had similar baseline PSA values (6.08 ± 2.93 vs. 5.89 ± 4.23, p = 0.85), proportion of GG 1 (15.89% vs 21.68%, p = 0.55) and PSA density (0.103 ± 0.041 vs. 0.123 ± 0.041, p = 0.36. However, AA patients did present at an earlier age (58.89 ± 6.64 vs. 63.69 ± 6.64, p = 0.004). A total of 8/19 (42.1%) AA and 46/143 (32.2%) non-AA had GG upgrading while on AS, p = 0.34. The median time until progression for AA and non-AA patients was 60.76 and 77.42 months, p = 0.68. Conclusions: In our study, AA men did begin AS at an earlier age than non-AA men. While both groups had statistically similar rates of progression, the relative risk of progression was higher in the AA cohort during this time period. Therefore, in the era of MRI and fusion biopsies we are better able to detect upgrading and somewhat mitigate the the risks associated with upgrading during AS irrespective of race but larger studies are needed to determine whether there are meaningful differences in the rates of progression between AA and non-AA men. This research was supported by the Intramural Research Program of the National Cancer Institute, NIH.

34 Background: AR directed therapies are standard of care for metastatic prostate cancer, and their use for high-risk localized disease may improve survival. Previous trials showed low rates of complete pathologic responses to neoadjuvant ADT. The underlying biology and molecular features of tumors that fail to respond remain unknown. Methods: Men with localized prostate cancer with high risk features received 6 months of GnRH agonist plus enzalutamide. mpMRI was performed at baseline and prior to RP. MRI-US fusion guided biopsies were performed at study enrollment. RP specimens were sectioned in the same plane as mpMRI. Imaging and anatomical landmarks on pre- and post-treatment mpMRI were used to match targeted biopsies to RPs. IHC stains including AR, GR, and SYP were performed to guide laser capture microdissection (LCM) and characterization of residual disease. DNA was extracted from LCM biopsy and RP tissues. Whole exome sequencing with somatic mutation calling and copy number analysis was performed. Results: Most patients showed incomplete pathologic responses despite reductions in tumor volume. Several patients harbored multiple tumor clones prior to treatment, distinguished by distinct copy number alterations and mutations in biopsies. Alignment of biopsies to pre- and post-treatment mpMRI correlated with imaging response and histopathology of matched RP sections. Minimal genomic divergence from untreated clones on biopsy suggested intrinsic resistance, while extensive divergence indicated selection for pre-existing subclones. Most residual tumor was AR- and GR-positive adenocarcinoma, with some regions of AR-negative Panneth cell-like differentiation positive for synaptophysin. AR-V7 was rare in residual tumor and < 1% cells were positive for KI-67. The most common alteration in all samples was single-copy loss of chromosome 16q overlapping with ZFHX3. Conclusions: Our approach of repeated sampling by matching targeted biopsies to mpMRI demonstrates feasibility in comparing mpMRI response, pathologic treatment response, and the underlying genomic features that drive tumor development versus treatment resistance. Clinical trial information: NCT02430480.

116 Background: Many urologists are now using active surveillance (AS) to manage patients with low-risk prostate cancer. A variety of criteria have been proposed to select patients for AS. A key feature of these criteria is Gleason Group as determined by systematic or “random” biopsies (SB). In this study, we examined the progression of Gleason Group (GG) detected only on SB compared to patients with positive MRI-fusion or “targeted” biopsies (FB). Methods: A prospectively maintained database was queried for all patients who underwent an MRI guided FB and SB from 2007 to 2016 and chose AS as their primary management strategy. Patents with GG 1 or 2 were eligible. Patients were then followed with yearly PSA, exam, MRI and biopsy if warranted. FBs were reviewed to determine GG progression. We divided patients into those who had cancer detected on SB only and those who had positive FBs with or without a positive SB. Results: A total of 162 patients met our criteria with 73 having an initial positive FB and 89 having cancer only on SB and a negative FB. Median follow-up was 27.62 (4.14 – 96.56) and 33.83 (3.45-99.84) months, respectively. The two groups were similar in age (64.49 ± 6.57 vs. 62.08 ± 6.92 years, p = 0.03) and PSA (6.23 ± 4.23 vs. 5.67 ± 3.98 ng/ml, p = 0.39). The median time to pathologic progression for GG 1 patients with initial FB positive was 53.88 months and those with GG 1 on initial SB was 80.81 months (p = 0.003). The median time to pathologic progression for GG 2 patients with initial FB positive was 36.26 months and those with GG 2 on initial SB was 76.01 months (p = 0.099). Conclusions: Previously accepted AS criteria have been based on SBs. However, those patients placed on AS after a positive FB progressed fast than cancer detected only on SB. With MRI and FBs more routinely used in clinical practice, clinicians could use this information from FBs to better stratify and closely monitor patients during AS. This research was supported by the Intramural Research Program of the National Cancer Institute, NIH.

113 Background: Prostate cancer (PCa) can show heterogeneous histology within lesions. MRI-targeted biopsy (Tbx) of the prostate improves PCa detection, but sampling within lesions has yet to be standardized. Furthermore, Tbx results are often heterogeneous as evidenced by differing histologic grades of Tbx cores within the same lesion. This introduces potential variability in biopsy results, on which clinical decisions are made. Here we aim to characterize lesion heterogeneity and identify predictive multiparametric MRI (mpMRI) features. Methods: A cohort of men who underwent mpMRI and Tbx between 2014-2017 were selected for analysis from a prospectively maintained database. To characterize lesion heterogeneity, only men with ≥2 positive Tbx cores were included. Histologic grades were scored according to International Society of Urological Pathology (ISUP) grades. Lesion heterogeneity, reported as a heterogeneity index (HI), was calculated as the difference of the average ISUP grades of Tbx cores per lesion from the maximum sampled ISUP grade of that lesion. Statistical analyses identified associations between imaging features and lesion heterogeneity. Results: 157 lesions in 114 patients met inclusion criteria. Maximum ISUP grade ranged from 1 to 5, with a median ISUP grade of 2. Higher ISUP grades were associated with greater lesion heterogeneity, HI for ISUP grade ≥3 = 0.58±0.11 vs

59 Background: Pathologic progression is identified in > 25% of prostate cancer (CaP) patients on active surveillance (AS). Yet, identifying patients at risk for progression is limited to PSA based biomarkers with variable utility. Multiparametric MRI (mpMRI) with fusion-guided prostate biopsy (FBx) has shown utility in risk stratification for patients considering AS. We compared mpMRI characteristics with PSA kinetics for the prediction of pathologic progression in AS patients. Methods: A review of men on AS with serial mpMRI and 2 or more FBx sessions was performed. FBx sessions consisted of targeted biopsies and a 12 core systematic biopsy. Men who met NIH Expanded AS criteria included those with low and intermediate risk CaP, Gleason score ≤ 3+4 = 7 with no restriction on percent core involvement or number of cores positive. Progression was defined by patients with initial Gleason 3+3 = 6 to any Gleason 4, and Gleason 3+4 = 7 to a primary Gleason 4 or higher. MRI progression was defined as increase in suspicion score, size, or new lesion on follow-up. PSA velocity (PSAV) > 0.75ng/ml/yr, PSA doubling time (PSAdt) < 3 yrs, and imaging characteristics were examined for association with pathologic progression at surveillance biopsy. Results: A total of 164 men, median age of 63 yrs [58-67], were included. Median length of follow up was 19.4 months [IQR 14.3-30.0]. Median PSA, and prostate volume of our cohort were 4.9ng/ml [3.3-7.3] and 47.0ml [36.5-59.8]. The sensitivity and specificity of predicting pathologic progression by mpMRI, PSAV and PSAdt were 45% and 65%, 30% and 76%, and 17% and 86% respectively. A combination of MRI, and PSAV or PSAdt yielded a sensitivity and specificity of 61% and 49% or 54% and 56% respectively. Using a decision curve analysis, mpMRI offers minimal benefit for predicting pathologic progression of CaP. Conclusions: MpMRI alone marginally outperforms PSA kinetics for predicting pathologic progression in men on AS for CaP. The combination of mpMRI along with PSA parameters increase the sensitivity of prostate imaging in identifying progression in AS patients. Research in combinations of imaging with other biomarkers will be needed to more accurately risk stratify AS patients.

30 Background: Endorectal(er) MRI is an emerging tool in assessing intraprostatic tumors. Immunotherapy development in prostate cancer has been limited due to the lack of intermediate (bio)markers of response. Methods: Untreatedpts with high-risk prostate cancer were randomized in a trial (NCT01496131) of standard ADT+Radiation + an immunotherapy targeting MUC1 (tecemotide, aka L-BLP25). Pts had erMRI at baseline and after 2 months of ADT+/- biweekly immunotherapy. Low dose (300 mg/m2, maximum 600 mg) cyclophosphamide for regulatory T-cell depletion preceded first immunotherapy. Multiparametric MRI included evaluation of apparent diffusion coefficient (ADC) maps from diffusion-weighted MRI. Monthly peripheral blood assessments utilized flow cytometry to evaluate immune cell subsets. This analysis focuses on the 2 month neoadjuvant period of ADT+/-immunotherapy before radiation. Results: 28 pts (n = 14/arm) with high risk prostate cancer (Gleason 8-10, PSA > 20, or stage T3) were enrolled. PSA declined in all pts 2 months after ADT. erMRI findings at 2 months indicated greater improvements in ADC values in pts receiving immunotherapy+ADT vs. ADT alone. Improved ADC on MRI suggests improvements in intratumoral diffusion and has been associated with decreased tumor density. This relative improvement between the groups occurred both per patient (p = 0.16) and per lesion (p = 0.031). Relative to baseline, pts receiving immunotherapy+ADT had increases in CTLA4+ CD8+ T-cells consistent with immune activation (p = 0.0134) and decreases in myeloid derived suppressor cells (MDSCs; p = 0.0353) during the neoadjuvant period corresponding to the erMRI changes. These immune findings were not seen in the ADT alone group. Conclusions: Pts who received immunotherapy+ADT for 2 months had greater improvements in ADC values on erMRI, consistent with decreased tumor density, relative to pts receiving ADT alone. Corresponding increases in activated CD8+ T-cells and decreases in MDSCs were seen in pts receiving vaccine+ADT. These preliminary findings suggest that ADC on MRI may be useful in assessing immunologic impact. Further study is warranted. Clinical trial information: NCT01496131.

15 Background: To determine the effect of learning curves and changes in fusion platform during 9 years of NCI’s experience with multiparametric MRI (mpMRI)/TRUS fusion biopsy. Methods: A review was performed of a prospectively maintained database of patients undergoing mpMRI followed by fusion biopsy (Fbx) and systematic biopsy (Sbx) from 2007−2016. The patients were stratified based on the timing of first biopsy in 3 groups. Cohort 1 included patients biopsied between 7/2007−12/2010, accounting for learning curve at our institution. Cohort 2 included patients biopsied from 1/2011 up to the debut of UroNav (Invivo) platform in 5/2013. Cohort 3 included patients biopsied after 5/2013. Clinically significant (CS) disease was defined as Gleason 7 (3+4) or higher. Cancer detection rates (CDR) between Sbx and Fbx during different time periods were compared using McNemar’s test. Age and PSA standardized CDRs were calculated for comparison between 3 cohorts. Results: 1528 patients were included in the study with 219, 549 and 761 patients included in 3 respective cohorts. Mean age, PSA and race distribution were similar across 3 cohorts. In cohort 1 there was no significant difference between CDR of CS disease by Fbx (24.7%) vs Sbx (21.5%), p = 0.377. Fbx was significantly better than Sbx in detection of CS disease in cohort 2 and cohort 3 (31.5% vs 25.3%, p = 0.001; 36.5% vs 30.2%, p < 0.001, respectively). There was significant decline in detection of low risk disease by Fbx compared to Sbx in the same period (cohort 2: 14.2% vs 20.9%, p < 0.001; cohort 3: 12.5% vs 19.5%, p < 0.001). Age and PSA standardized CDR of CS cancer by Fbx increased significantly between each successive cohort (cohort 1 and 2: 5.2%, 95% CI [2.1-8.5]), 2 and 3 (5.2%, 95% CI [1.8-8.6]). Conclusions: Our results show that after an early learning period using Fbx, CS prostate cancer was detected at significantly higher rates with Fbx than with Sbx, and low risk disease was detected at lower rates. Advances in software allowed for even greater detection of CS disease in the last cohort. This study shows that accuracy of Fbx is dependent on multiple factors; surgeon/radiologist experience and software improvements together produce improved accuracy.

32 Background: Several studies have demonstrated an association in prostate cancer patients between the tumor volume of the index lesion calculated from whole mount pathology after radical prostatectomy and biochemical recurrence (BCR). We aimed to evaluate whether the index lesion tumor volume (ILTV) measured on MRI would have similar predictive ability. Methods: Clinical and pathologic data from a prospectively maintained, single-institution database was analyzed for 501 patients who underwent RARP (without prior therapy) between 2007 and 2016. Patients were excluded if they were found to have metastatic disease on pathology or if postoperative PSA ≥ 0.2ng/ml after RARP. ILTV was calculated from T2W MRI by multiplying length by width by depth by 0.52. Tumor volume was stratified into two categories based on the cutoff of 2cm3. BCR was defined as 2 PSA values ≥ 0.2 ng/ml, 1 PSA value ≥ 0.4 ng/ml, or receipt of radiation therapy/hormonal therapy after RARP. Cox proportional hazard regression was used to assess associations of clinical, imaging, and histopathological variables with BCR. Results: The median age of our cohort was 60 years (IQR 10) and median PSA was 6.20 ng/ml (IQR 6). Of the 493 patients who met our inclusion criteria, 55 (11.2%) experienced BCR. Patients with ILTV > 2cc had a higher incidence of BCR compared to those with ILTV < 2cc (18.0% vs 8.8%, p = 0.008). On univariate analysis, age (HR 0.96, p = 0.026), PSA (HR 1.04, p < 0.001), black race (HR 1.816, p = 0.047), Gleason 8-10 on biopsy (HR 3.725, p < 0.001), and ILTV (HR 2.496, p = 0.002) were significant for prediction of BCR. Although age (HR 0.955, p = 0.021), PSA (HR 1.029, p = 0.013), and Gleason 8-10 on biopsy (HR3.367, p < 0.001) were found to be independent predictors of BCR, tumor volume was not found to be an independent predictor (HR 0.960, p = 0.642). Conclusions: Although ILTV alone was predictive of BCR, ILTV was not found to be an independent predictor of BCR, indicating that at this time, MRI cannot yet measure ILTV with enough accuracy to predict oncologic outcomes. As imaging technology improves, ILVT may provide additional prognostic information.

31 Background: To determine the accuracy of multiparametric MRI (mpMRI) and fusion biopsy in predicting focal therapy candidates by correlating agreement between patients’ pre-prostatectomy MRI and biopsy findings with whole mount pathology. Methods: Patients who underwent mpMRI and biopsy followed by prostatectomy between 2010-2015 were included in the study. Patients were considered good candidates for focal therapy if they had intermediate risk cancer (max Gleason 3+4 or 4+3 on biopsy) in only one hemisphere of the prostate, and if they had an mpMRI visible lesion(s) on one hemisphere that did not cross midline. Patients were considered poor candidates for focal therapy if they had either low risk cancer (Gleason 6), high risk cancer (Gleason 8, 9 and 10), intermediate risk cancer bilaterally, MRI visible lesions that cross the midline, or contained extracapsular extension or seminal vesicle invasion. Whole mounts of patients considered good candidates were then examined histologically for maximum tumor involvement. Results: 501 patients who underwent prostatectomy between 2010 and 2015 were included in the study, with mean age 60.0 (±10.0) and PSA 6.21 (±6.0). 395 patients were considered poor focal candidates, and 106 were considered good focal candidates. Of these 106, 66 patients had whole mounts available for analysis (14 random biopsy, 52 fusion biopsy). 42/66 (63.6%) patients were confirmed as candidates (6/14 (42.9%) who had random biopsy and 36/52 (68.0%) who had fusion). Candidates failed prediction due to Gleason upgrade in final histopathology in 20/24 (83.3%), and for significant Gleason detection bilaterally in 4/24 (16.7%). Conclusions: Our results suggest that application of mpMRI and biopsy criteria to predict focal therapy candidates may result in undertreatment of a large portion of patients with significant cancer. Results from mpMRI guided targeted biopsy were shown to be a more accurate tool for predicting candidates for focal therapy than results from random biopsy alone. More accurate predictive capability is needed before focal therapy can be widely accepted.

e540 Background: Magnetic resonance imaging (MRI)/transrectal ultrasound (TRUS) fusion-guided prostate biopsy (FBx) has rapidly proliferated in prostate cancer evaluation. FBx success may be dependent on multiple factors, including skills learned with user experience and the type of ultrasound (US) probe used. Our objective of this study was to compare the accuracy of the MRI/TRUS FBx technique utilizing an end fire versus a side fire US probe on an FBx platform. Methods: The participants of the study consisted of three different experience levels (expert ( > 1000 FBx), Urology fellows with none to minimal FBx experience, Urology residents with no FBx experience). Each participant performed three trials of the FBx with the UroNav System in separate phantom models of the prostate. Each trial involved targeting the same set of fiducials through FBx platform and then using US only; first using an end fire US probe, and then using a side fire US probe. Fusion registration error (FRE) was defined as the distance between MR target and transformed core location from the US only biopsy. Results: Six users with 3 different experience levels performed the FBx. Mean FRE with the end fire probe for the residents, fellows, expert was 7.42(±0.11), 5.10(±0.73) & 3.75(±0.75) respectively. Mean FRE with the side fire probe for the residents, fellows, expert was 6.08(±0.09), 4.76(±0.68) & 3.46(±0.08) respectively. There was no significant difference between the FRE averages for end fire vs. side fire. There was an inverse relationship of mean FRE with user experience level (corr. coefficient (r) = -0.79, p = 0.011 for end fire; r = -0.84, p = 0.004 for side-fire). Conclusions: We report the results of a pilot study comparing the efficacy of side fire and end fire US probes in MRI/TRUS FBx. Our preliminary results suggest the type of US probe used does not significantly affect FRE and FBx performance. As expected, FRE decreases with increased user experience. A larger study with more participants will be needed to compare these probes with adequately powered analysis and will dictate the practice in clinical settings as minimization of FRE will improve diagnostic accuracy of this technique in detecting clinically significant cancer.

231 Background: Conventional imaging of advanced prostate cancer (computerized tomography and nuclear bone scintigraphy) is limited and indicates a need for a more specific molecular imaging probe. DCFBC is a radiolabeled PET agent that binds with high affinity to prostate specific membrane antigen (PSMA), which is overexpressed in almost all prostate cancers and through whole-body non-invasive functional imaging, may provide new information on the expression of PSMA. We compare the uptake of DCFBC in bone with respect to NaF PET/CT in metastatic prostate cancer patients. DCFBC has added capability to detect soft tissue metastasis whereas NaF is confined to secondary effects of bone disease. Methods: Subjects with known or suspected prostate cancer metastasis underwent DCFBC PET/CT imaging performed at 1 hour and 2 hours after IV bolus injection of 8 mCi of DCFBC. Patients also underwent a whole body NaF PET/CT scan within 3 weeks of DCFBC PET/CT to assess for bone metastases. Patients received 3 mCi of NaF IV bolus and then were imaged 1hour post injection. PSA levels and antiandrogen therapy status were obtained at the time of DCFBC imaging. Results: Fifteen patients have been preliminarily analyzed. PSA ranged from < .01 to 4379 ng/mL. NaF identified bone lesions in 10 patients but matching focal DCFBC uptake was only seen in 3 patients. DCFBC additionally showed lymph node metastasis in 1of these 3 patient. There were 5 patients without focal abnormal bone uptake on NaF or DCFBC. In this group, 4 of 5 patients had focal DCFBC uptake in lymph nodes or soft tissue lesions. Ten patients were on some form of androgen deprivation therapy (ADT). For those on ADT, 7 of 10 patients had positive findings on NaF, compared to 2 of 10 patients on DCFBC. Conclusions: DCFBC uptake in bone metastasis does not routinely correspond to focal NaF uptake which could be due to distinct mechanisms of tracer uptake and tumor biology. There is an inverse association in focal bone findings when comparing each tracer on antiandrogen therapy. Through whole-body non-invasive functional imaging and further study, DCFBC may prove useful in characterizing prostate cancer based on PSMA expression. Clinical trial information: NCT02190279.

43 Background: Tumor volume measured on postoperative pathology has been shown to be associated with adverse pathologic and oncologic outcomes. Advances in imaging technology may facilitate the prediction of these outcomes prior to surgery. In this study, we evaluated the predictive value of tumor volume of the index lesion (ILTV) calculated from prostate MRI in analyzing adverse pathologic outcomes following radical prostatectomy (RARP). Methods: Clinical and pathologic data from a prospectively maintained, single-institution database was analyzed for patients who underwent 3T MRI prior to RARP (without prior therapy) between 2007 and 2016. ILTV was calculated from T2W MRI by multiplying length by width by depth by 0.52. Adverse pathologic outcomes were determined on whole mount prostatectomy specimens, and defined as positive margins (PM), extracapsular extension (ECE), positive lymph nodes (LNI), and seminal vesicle invasion (SVI). Logistic regression was used to assess associations of clinical, imaging, and histopathological variables with adverse pathologic features. Results: From the 501 patients who underwent RARP, 490 received MRI prior to surgery. The median age of our cohort of was 60 years (IQR 10), and median PSA was 6.21 ng/ml (IQR 6). There were 74/480 patients (15.4%) with PM, 112/474 (23.6%) with ECE, 30/478 (6.4%) with LNI, and 32/482 (6.7%) with SVI. Patients with adverse outcomes had larger ILTV compared to those without (PM: 1.038cc vs 0.807cc, p=0.041; ECE: 1.388cc vs 0.765cc, p

11028Background: Desmoids are rare, locally invasive, slow growing soft tissue tumors that are either sporadic or associated with familial genetic syndromes such as familial adenomatous polyposis. ...

115 Background: Active surveillance (AS) is an established option for men with prostate cancer. Studies have shown that multiparametric-MRI along with MRI-TRUS fusion-guided biopsy (FB) may better assess risk in patients eligible for AS, compared to 12-core biopsy, due to improved detection of clinically significant cancers. The objective is to evaluate the performance of expanded criteria eligibility in men on AS being monitored with MRI-TRUS guided biopsy. Methods: Men on AS were included if they had mp-MRI and pathology data for 2 or more FB sessions. FB procedures consisted of targeted biopsies and random 12 core biopsies. Men participated in AS with low and intermediate risk prostate cancer, Gleason score ≤ 3+4=7 with no restriction on percent core involvement or number of cores positive. Progression was defined by patients with initial Gleason 3+3=6 to any Gleason 4, and Gleason 3+4=7 disease progressing to a primary Gleason 4 or higher. Results: 124 men on AS met study criteria. Low risk men had a mean age of 61.3 years versus intermediate risk men with a mean age of 65.5 years (p=0.0062). Mean PSA levels of the low and intermediate risk groups were 5.8 and 5.76 ng/ml (p=0.95), respectively. The mean length of follow-up was 22.56 months (range: 3.6 – 74.4 mo). Rates of pathologic progression in the intermediate and low risk patients were, 38.5% vs. 28.5% (p=0.33). Intermediate risk men had a mean progression-free survival (PFS) of 2.8 years compared to low risk men of 3.9 years (p=0.27). Patients were stratified according to established AS criteria (Epstein, Toronto, PRIAS) and rates of progression are summarized in the Table. 69% of patients met Epstein criteria for AS of which 29.4% (20/68) progressed compared to 28.5% for the low risk cohort overall. Conclusions: Men in our cohort who met strict criteria for AS had the same rate of progression as the entire expaned criteria low risk cohort, 29.4% vs 28.5%, respectively. Our data suggests that with accurate initial Gleason classification other AS criteria such as percent core or number of cores positive have no added benefit in predicting which men may have reclassification or progression of disease. [Table: see text]

61 Background: Multiparametric MRI (MP-MRI) can visualize prostate tumors. MP-MRI characteristics (extraprostatic extension (ECE), tumor volume, etc.) can be predictive of final pathologic findings such as lymph node (LN) involvement, pathological ECE (pECE), and biochemical recurrence (BCR). These are pivotal in the decision-making process regarding treatment. Tumor contact length (TCL) is defined as the length of cancer in contact with the prostatic capsule. We evaluated the ability of MP-MRI determined TCL in predicting pECE, BCR and LN in patients undergoing radical prostatectomy. Methods: All patients who underwent a 3T MP-MRI at the NCI from 2007 to 2015 were retrospectively classed into no ECE, suspicious ECE (sECE) and frank ECE (fECE) based on MP-MRI findings. sECE was defined as tumor with capsular bulge on MRI while fECE was clear capsular obliteration and tumor extension beyond the prostatic capsule. Demographic data was obtained on patients with fECE and sECE on MP-MRI with the presence of pECE, LN, and BCR status following radical prostatectomy from a single surgeon (PP) experience. Chi-Square and Wilcoxon rank sum test were used to compare proportions and continuous variables respectively. Logistic regression was used to determine the predictive ability of TCL. Statistical significance was defined as p-value ≤0.05. Results: Of all 1,260 patients who underwent MP-MRI, we focused on 146 who had sECE (68) or fECE (78) on MP-MRI. Mean age was 60 years and median prostate specific antigen was 11.7 ng/ml. Logistic regression analysis showed that MP-MRI determined TCL was predictive of ECE (p=0.01), LN status (p=0.0001) on final pathology and BCR (p=0.05) during follow up. Patients with pECE had a longer median MP-MRI TCL (2.8 cm) compared to those without pECE (2.4 cm), p=0.04. When analyzed individually, fECE correlated with pECE (p=0.05) while s ECE did not correlate with pECE (p=0.11). Although, not statistically significant, the median MP-MRI TCL for sECE with pECE was still longer than in sECE with no pECE in the sub-group analysis. Conclusions: Longer TCL on MP-MRI can indicate presence of ECE, LN involvement at final pathology as well as predict BCR on follow up.

e16128 Background: The majority of patients with low risk prostate cancer (CaP) harbor indolent disease and maybe safely observed on active surveillance (AS). However more than 30% of those on AS will ultimately receive definitive treatment. Focal therapy offers a treatment alternative that may provide oncologic control while avoiding treatment-related morbidity. The objective of this study was to determine the safety and feasibility of magnetic resonance (MR) image-guided focal laser ablation (FLA) of biopsy confirmed and MR visible prostate cancer. Methods: MR image-guided FLA was performed for men with Gleason < 7(3+4) when < 3 cores were positive in a standard 12 core biopsy corresponding to locations of MR lesions, or < 4 cores positive on MR image-guided biopsy. MR localization of biopsy-positive lesions were targeted via a transperineal approach and treated with a laser thermal ablation system. Validated quality of life questionnaires were completed pre-procedure and at each subsequent followup vis...

10563 Background: Desmoids are rare, invasive, slow growing soft tissue tumors that are sporadic or associated with familial genetic syndromes, such as familial adenomatous polyposis. Desmoids are ...

52 Background: The subcapsular region is often under-sampled by standard 12-core transrectal ultrasound (TRUS) prostate biopsy. We retrospectively evaluated multi-parametric MRI (mpMRI) features of subcapsular prostate cancers (PCa) and their detection on MRI/TRUS fusion-guided biopsy (FgBx). Methods: All patients referred for 3T mpMRI and subsequent FgBx under an IRB-approved protocol from Jan 2010 to July 2014 were reviewed. Histologically confirmed subcapsular lesions (just inside the prostate capsule with capsular border length >> width) were identified. Impact of the FgBx on clinical management according to current National Comprehensive Cancer Network (NCCN) guidelines was evaluated, with prior status based on outside TRUS biopsy results. Results: Of 992 eligible patients, 33 had subcapsular prostate lesions on mpMRI. In this subpopulation, mean age, PSA, and prostate volume were 63 years (range 52-76), 8.4ng/mL (1.2-63), and 53 mL (12-125). The FgBx, consisting of targeted (TBx) and systematic (SBx) portions, confirmed PCa in 24 of 33 patients. Five cancers (21%) were only detected by TBx. The remaining 19 were diagnosed on both TBx and SBx, but in 5 of these 19 (26%), higher Gleason scores were found on TBx cores. The FgBx results more accurately classified patients’ NCCN risk scores, with 17 prior unknown or low risk PCa upgraded to higher risk scores (Table). Conclusions: Subcapsular lesions on mpMRI are rare but largely malignant. Most are found on 12-core SBx, but ~20% require TBx for diagnosis, and TBx more accurately grades >25% of the lesions. The FgBx may improve clinical management of patients with subcapsular PCa, since more aggressive cancer was found in 17 of 24 patients. [Table: see text] [Table: see text] [Table: see text]

43 Background: Active surveillance (AS) is an established treatment option for men with low risk prostate cancer. Its role in intermediate prostate cancer is still being investigated. Recent studies have shown that multiparametric-MRI (mp-MRI) along with MRI-TRUS fusion-guided biopsy may better assess risk in patients eligible for AS, compared to 12-core biopsy, due to improved detection of clinically significant cancers. The objective is to determine the performance of MRI-TRUS guided biopsy for men on AS with both low and intermediate risk disease. Methods: Between 2007-2014 men on AS were included if they had complete mp-MRI and pathology data for 2 or more MRI-TRUS biopsy sessions. Fusion guided biopsy procedures consisted of MRI identified targeted biopsies as well as random 12 core biopsies. Men were allowed to participate in AS with low and intermediate risk prostate cancer, Gleason score ≤ 3+4=7. Progression was defined by patients with initial Gleason 3+3=6 to any Gleason 4, and Gleason 3+4=7 disease progressing to a primary Gleason 4 or higher. Results: 89 men met our study criteria with an average age of 62 years old (range 45-79). 75 men had low risk Gleason 3+3=6 at the outset of AS by 1st biopsy session with a median PSA 5.1 ng/ml. The other 14 men had intermediate risk prostate cancer Gleason 3+4=7 at the outset of AS and a median PSA 4.6 ng/ml. During follow-up, 25 (33%) low risk men progressed to 3+4 or above at a median of 20.6 months. Of these, 19 were found by targeted biopsy. 6 (43%) of the intermediate risk men progressed to Gleason 4+3=7 at a median of 36.8 months. 4 of these progressed on targeted fusion biopsy. In the intermediate risk men, 84 random biopsy cores were require to detect 1 progression versus 15 targeted biopsy cores to detect 1 progression. Conclusions: The majority of patients on AS who progressed were identified by MRI-TRUS targeted biopsy. Less biopsy cores are required to detect progression with targeted biopsy. These results are preliminary and a larger cohort with longer follow-up would be beneficial.

208 Background: Conventional imaging methods of lymph node staging in prostate cancer are limited. The goal of this study is to determine the utility of ferumoxytol enhanced MRI in lymph node (LN) staging. Methods: This ongoing IRB-approved clinical trial enrolls prostate cancer patients at high risk for LN metastases. Patients first undergo baseline T2 and T2* weighted MRI scans followed by 7.5mg/Kg ferumoxytol injection. Repeat scans are acquired at 24 hr and 48 hr post-injection. The criterion for positive LNs was hyperintense signal indicating failure to take up ferumoxytol. Validation was determined on clinical grounds or by histopathology when available. Results: To date, 11 patients have completed the study. One patient was examined pre-operatively while the other 10 had suspected therapy failure. Median age and PSA were 65 yrs (36-75) and 5.6ng/dL (0.3-201). Of 16 LNs with median size 1.6 x 1.1cm, 10 were true positives, one was false positive and one was false negative with 4 nodes pending validation. The LN status was correctly identified in 8 of 11 patients [Table]. Conclusions: Ferumoxytol enhanced MRI shows promise in detecting malignant LNs with diameter >6mm in prostate cancer paients. This method is performed in a conventional MRI unit with off-label use of an FDA-approved agent and thus could be widely available. However, further validation is necessary before routine use. Clinical trial information: NCT02141490. [Table: see text]

115 Background: 18F-FDG PET/CT is widely used to diagnose malignancy, but is not recommended for localized prostate cancer. This study explores the value of multi-parametric MRI (mpMRI) in characterizing incidentally detected prostate FDG uptake. Methods: Thirty-one patients who underwent FDG PET/CT and prostate MRI were eligible for this study. 14 patients were excluded (n=8 insufficient histopathology, n=6 radical prostatectomy before PET), with final analysis of 17 patients. The mpMRI sequences included T2-weighted, dynamic contrast enhancement (DCE), apparent diffusion coefficient (ADC), and MR spectroscopy (MRS). Nuclear medicine physicians, blinded to clinicopathologic findings, identified suspicious areas and maximum standardized uptake values (SUVmax) on FDG PET/CT. The lesion and sector-based imaging findings were correlated with annotated histopathology from whole-mount or MRI/TRUS fusion biopsy samples. Positive predictive values (PPVs) were estimated using generalized estimating equations with logit link. Results were evaluated with Kruskal-Wallis and Dunn’s multiple comparisons tests. Results: The PPV of FDG PET alone in detecting prostate cancer was 0.56. Combining FDG (base parameter) with mpMRI modalities (T2, DCE, ADC, MRS) increased the sector-based PPV to 0.79, 0.82, 0.80, and 0.89, respectively. All benign lesions had SUVmax < 5, and malignant lesions had higher mean SUVmax values that correlated with Gleason scores [Table]. This relationship between SUVmax and Gleason score was significant, with p=0.012 on the Kruskal-Wallis test and p=0.015 on the Dunn’s multiple comparisons test for Gleason 0 vs Gleason ≥ 4+5. Conclusions: Incidental prostate FDG uptake has low clinical utility alone, but these areas may harbor high-grade prostate cancer, especially if the SUVmax is greater than 5. [Table: see text]

51 Background: Multiparametric MRI (mpMRI) has been shown to improve clinically significant prostate cancer (CaP) detection. Targeted biopsy using MRI/transrectal ultrasonography (TRUS) fusion is a novel diagnostic tool. The negative predictive value (NPV) of an MRI/TRUS fusion targeted biopsy of a suspicious lesion on mpMRI was determined. Methods: 30 of 181 men who underwent prostatectomy from 2008-2014 were retrospectively identified and had at least one lesion on mpMRI negative for cancer on MRI/TRUS fusion biopsy. Whole mount pathology specimens, gold standard for CaP detection, were aligned with MRI to assess true histopathology of all identified targets. Lesions negative for CaP on biopsy and not identified as cancer on pathology were considered true negatives (TN). Lesions biopsied negative but later found to possess foci of CaP on whole mount were considered false negatives (FN). Calculations of NPV were then made per biopsy year, MRI suspicion score, and lesion size on MRI. Results: 48 lesions of a total 81 identified on mpMRI were reported negative for CaP in the 30 patients who underwent fusion biopsy. Of these, 37 lesions were found to be truly negative on histopathology, while 11 lesions had CaP foci on whole mount specimen. Overall NPV was 77% (37/48). The NPV increased over time (Table 1), and was as high as 85.7% most recently. Conclusions: This series demonstrated a NPV of 77% for targeted MRI/TRUS fusion biopsy of lesions seen on mpMRI. The increasing NPV trend noted over time may have further applications to assess the learning curve for this diagnostic method. Not surprisingly, NPV is higher for low and moderately suspicious lesions than for highly suspicious lesions. This data may help physicians interpret the clinical implications of a negative fusion biopsy. [Table: see text]

44 Background: Central gland (CG) prostate cancers (CaP) are reported with lesser incidence and smaller tumor volume compared to the peripheral zone (PZ). Index tumor lesions defined by highest grade may be missed when in the CG. MRI/US fusion-guided biopsy allows targeting of lesions, potentially identifying cancer outside the traditional TRUS biopsy template. Methods: Retrospective review of 1,003 patients who underwent multiparametric MRI (mpMRI) found 2,119 suspicious lesions. Targets were biopsied and stratified by zonal distribution, CG or PZ. Cancer detection rates (CDR) were tabulated by location and correlated with PSA, Gleason score, prostate volume and MRI suspicion. Results: Fusion-guided biopsy targeted lesions in the central (711, 34%) or peripheral (1408, 66%) prostatic zones. CDR was similar between zones: 35.2% CG compared to 33.6% PZ (Table). CDR of clinically significant disease (Gleason >4+3) was similar in the CG and PZ despite higher prostate volume in those with CG lesions. In contrast to TRUS biopsy, upgrading occurred in 18.5% of CG patients versus 13.3% PZ (p=0.024). 36.6% (77/210) of CG lesions represented the highest risk lesion on MRI, translating to 13% (77/592) of the biopsy-proven CaP cohort. Conclusions: CG cancers occur at a similar frequency as PZ CaP. CG lesions were more likely to be upgraded from TRUS biopsy, frequently representing the index lesion. In upgraded patients, CG targets constituted the index lesion in a third of all males. MRI/US fusion-guided biopsy identifies clinically significant disease of the CG not captured on traditional biopsy. [Table: see text]