Matthew Lee

Prepubertal Malignant Large Cell Calcifying Sertoli Cell Tumor of the Testis

An otherwise healthy 7-year-old boy was diagnosed with malignant large cell calcifying Sertoli cell tumor (LCCSCT) of the testis. He underwent attempted partial orchiectomy with conversion to radical orchiectomy due to suspected malignancy on intraoperative frozen section. There was no lymph node or visceral metastases. To our knowledge, this is the first report of malignant LCCSCT in the prepubertal population. LCCSCT of the testis is an extremely rare neoplasm, with low malignant potential. Malignant cases are exclusively reported previously in the adult population. We report the first case of malignant LCCSCT in a pediatric patient. We review the literatures and discuss the clinical, pathologic features and treatments of malignant LCCSCT.

Abstract B036: Comprehensive transcriptomic profiling challenges the robustness of prostate cancer prognostic signatures to multifocality

Purpose: Imaging and tissue-based biomarkers are increasingly utilized in men diagnosed with low-grade prostate cancer (PCa) to guide definitive management vs. active surveillance. PCa is uniquely multifocal, with multiple clonally distinct tumor foci present in the majority of men. A single tumor focus may also be composed of both low- and high-grade components. Hence, an ideal prognostic biomarker should be robust to both undersampling of a high-grade component, as well as a not sampled multifocal high-grade tumor focus. Materials and Methods: To assess the robustness of prognostic biomarkers to multifocality, we designed a comprehensive multiplexed targeted RNA sequencing assay (mxRNAseq) capable of assessing multiple classes of transcriptional alterations and deriving available prognostic signature scores (e.g., Prolaris CCP and OncotypeDX GPS). We applied this assay to a retrospective, multi-institution cohort of over 150 formalin-fixed, paraffin-embedded tissue samples representing the range of prostate cancer progression. Single candidate biomarkers and derived prognostic signatures were compared between low-grade foci in cases with and without high-grade foci, all-low-grade foci vs. all-high-grade foci, and cases with extremes of grade differences in multifocal tumors. Frequency of clinically relevant extreme multifocality was determined by identifying patients in a single-institution consecutive prostatectomy database over 9 years who had exclusively low-grade cancer on biopsy but very-high-grade disease on prostatectomy. Results: Our mxRNAseq assay robustly detected known coding gene/lncRNA expression, gene fusions and splice variants, and expressed mutations and germline variants. Supervised hierarchical clustering of target gene expression confirmed expected transcriptional module deregulation and derived prognostic signatures across prostate cancer progression. Prognostic biomarkers (including derived signatures) showed no significant differences in expression between low-grade foci from prostates with and without high-grade multifocal tumors and were uniformly higher in high-grade foci vs. low-grade foci from the same case. In four cases of extreme multifocality (Gleason score 6 vs. >=8 foci), prognostic signatures were significantly lower in low-grade foci vs. high-grade foci. In 1,418 men with biopsy Gleason score 3+3=6 or 3+4=7, 21 (1.5%) had Gleason score ≥ 4+4=8, where the biopsy almost certainly did not sample the most clinically relevant focus. Conclusions: Using a novel comprehensive mxRNAseq assay, our results challenge the robustness of prognostic biomarkers between multifocal low- and high-grade prostate cancer foci. Tissue-based prostate cancer prognostic biomarkers should be specifically validated using an extreme multifocality design to support utility in predicting the presence of an unsampled, aggressive multifocal tumor focus. Citation Format: Simpa Salami, Daniel H. Hovelson, Jeremy B. Kaplan, Romain Mathieu, Aaron M. Udager, Nicole Curci, Matthew Lee, Lorena Lazo de la Vega, Martin Susani, Nathalie Rioux-Leclercq, Daniel E. Spratt, Todd M. Morgan, Matthew S. Davenport, Mark A. Rubin, Shariat F. Shahrokh, Scott A. Tomlins, Ganesh S. Palapattu. Comprehensive transcriptomic profiling challenges the robustness of prostate cancer prognostic signatures to multifocality [abstract]. In: Proceedings of the AACR Special Conference: Prostate Cancer: Advances in Basic, Translational, and Clinical Research; 2017 Dec 2-5; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(16 Suppl):Abstract nr B036.

MP32-07 ASSOCIATION BETWEEN HOSPITAL ACCOUNTABLE CARE ORGANIZATION STATUS AND READMISSION FOLLOWING CYSTECTOMY AND OTHER MAJOR SURGERY

METHODS: We built a cohort of aged Medicare beneficiaries from 2010 through 2013 comprising 17,779,120 person-years and 9,201,163 person-years before and after ACO enrollment, respectively. We characterized our exposure of interest, MSSP ACO enrollment, by identifying all MSSP-enrolled primary care providers and recapitulating published attribution strategies, and identified our outcome of interest, PSA screening, through relevant Medicare claims. We subsequently performed differences-in-differences analysis specifically evaluating the interaction between ACO enrollment and the period following ACO intervention to characterize the effect of MSSP ACO enrollment on the prevalence of prostate cancer screening. RESULTS: Medicare beneficiaries attributed to MSSP ACOenrolled primary care providers were 2.5% more likely to undergo PSA screening than those attributed to non-ACO primary care providers prior to deployment of the MSSP (p<0.0001). We observed significant reductions in the prevalence of PSA screening throughout the study period among both ACO and non-ACO groups. Nonetheless, the rate of decline in the ACO group outpaced that of the non-ACO group, with an observed difference-in-difference of 0.742% (p<0.0001). CONCLUSIONS: ACO enrollment appears to significantly mediate observed reductions in the prevalence of PSA screening among Medicare beneficiaries, although the absolute magnitude of this effect is relatively small. Characterizing the potential additive effects of guideline modifications and payment reforms will be essential to predict the future landscape of prostate cancer epidemiology.

PD33-05 TARGETED NEXT GENERATION SEQUENCING TO CHARACTERIZE MAGNETIC RESONANCE IMAGING VISIBLE AND INVISIBLE PROSTATE CANCER: BIOLOGICAL INSIGHTS AND THERAPEUTIC IMPLICATIONS

METHODS: Formalin-fixed paraffin embedded (FFPE) primary prostate cancer samples were obtained from the Welsh Cancer Bank. Targeted-NGS was performed using the Life Technologies Ion Torrent: Ion AmpliSeq Cancer Hotspot Panel v2 and the Ion Personal Genome Machine sequencer. The hotspot panel covers ~2800 COSMIC mutations of 50 oncogenes and tumour suppressor genes. Standard IHC techniques were also used concentrating on markers of the Wnt, PI3Kinase (PI3K) and MAP-Kinase (MAPK) oncogenic signalling pathways. RESULTS: 61 primary prostate cancer samples were sequenced, 58 from radical retropubic prostatectomy (RRP) specimens and 3 from transurethral resection of prostate (TURP) sections, with a range of Gleason Scores (GS). 21/61 (34.4%) samples harboured a mutation in a cell cycle pathway gene such as TP53 or RB1 and 3/61 (4.9%) in a DNA repair gene such as ATM. 10/61 (16.5%) of samples harboured a mutation in a gene associated with the Wnt pathway such as APC or CTNNB1. 14/61 (23.0%) of samples analysed had a mutation in a gene commonly associated with the PI3K pathway such as PTEN or AKT1. 5/61 (8.2%) had a mutation in a gene associated with the MAPK pathway such as KRAS. IHC profiles were analysed on 317 prostate samples: 73 normal and 244 cancers. There was greater expression of markers associated with Wnt, PI3K and MAPK signalling pathways in prostate cancer samples when compared to normal samples. There was greater expression in high-risk GSs with some markers associated with biochemical recurrence following RRP. Furthermore, we were able to separate lowand high-risk GS samples based on molecular profiles using markers of the Wnt, PI3K and MAPK and principle components analysis. CONCLUSIONS: Targeted NGS and IHC can identify recurrent mutations and signalling pathway aberrations within primary prostate cancer samples, which have potential to be targeted and used in routine clinical practice. In addition, the molecular signatures of lowand highrisk are different and can be separated using a combination of markers and IHC. This finding could explain the marked difference in the behaviours of these tumours types.

MP08-12 MULTI-INSTITUTIONAL EVALUATION OF MRI AND FUSION BIOPSY IN CONFIRMATORY BIOPSY FOR ACTIVE SURVEILLANCE

INTRODUCTION AND OBJECTIVES: To provide standardization as prostate MRI becomes increasingly utilized, the Prostate Imaging-Reporting and Data System (PIRADS) was developed and has been modified to its latest version (v2). Using biopsy outcome as the standard, we examined the predictive accuracy of a PIRADS 4 or 5 read for clinically significant (Gleason 7+) PCa in a blinded fashion. METHODS: We reviewed our prospectively maintained database of consecutive men who underwent prostate MRI prior to biopsy between September 2014 and December 2015. A proportionally representative sample (based on the original clinical PIRADS v2 interpretation) was selected for re-examination (n1⁄432). The prostate MRIs for these patients were de-identified and were loaded by a blinded third party. Four radiologists of varying levels of experience independently interpreted all prostate MRI, blinded to all clinical information. An 00overread00 was defined as a PIRADS 4 or 5 read with biopsy result of benign prostate or Gleason 6 PCa. An 00under-read00 was defined as a PIRADS 1-3 read with resulting biopsy result of Gleason 7+ PCa. RESULTS: The distribution of accuracy is provided in Table 1. Accurate interpretation ranged from 56% (18/32) to 75% (24/32), and the differences among the radiologists were not significant (p1⁄40.48). The improvement of accuracy with a 00majority read00, as defined by two or more accurate radiologists0 blinded interpretations, over the original clinical read trends toward significance (p1⁄40.16). No clinical variable was predictive of an incorrect 00majority read00, including age, PSA, family history, use of 5-alpha reductase inhibitors, prostate volume, or previous biopsy history. CONCLUSIONS: In a blinded assessment of radiologists at our institution, we find that the predictive accuracy of PIRADS 4 or 5 for clinically significant PCa varies among radiologists independent of experience level. A 00majority read00 performed better than the original clinical interpretation, suggesting that consensus interpretation of prostate MRI may improve predictive accuracy.