Faculty Bibliography

BACKGROUND:Malignant angiomyolipoma is an uncommon tumor of the class of perivasciular epithelioid cell neoplasms (PEComas). These tumors are characteristically driven by deleterious mutations in the tumor suppressors TSC1 and TSC2, whose gene products typically act to inhibit mTOR. There are several cases of malignant angiomyolipoma which exhibit transient responses to mTOR inhibitors, forming the basis of current practice guidelines in malignant PEComa. However the tumors ultimately acquire resistance, and there is no well-established second-line option. Despite the increasing prevalence of immunotherapy across a wide range of solid tumors, little is known about the immune infiltrate and PD-L1 expression of angiomyolipoma. Furthermore, there is no reported case on the treatment of malignant angiomyolipoma with an immune checkpoint inhibitor. CASE PRESENTATION/METHODS:A 38 year-old man presented with gross hematuria and was diagnosed with renal epithelioid angiomyolipoma. Despite surgical resection, the tumor recurred and metastasized. Targeted genomic sequencing revealed a deleterious mutation in TSC2, and the patient was treated with the mTOR inihbitor everolimus. The patient went on to have a partial response but ultimately progressed. He was then treated with the anti-PD-1 immune checkpoint inhibitor nivolumab, and achieved a durable near-complete response which is ongoing after two years of treatment. Immunohistochemical staining of tumor tissue revealed strong PD-L1 expression and a brisk T-cell infiltrate. CONCLUSIONS:We report on the first durable systemic treatment of malignant epithelioid angiomyolipoima with the use of PD-1 antibody nivolumab. Given the absence of prospective clinical trials in this exceedingly rare disease, particularly in the second-line setting, immune checkpoint inhibitors like nivolumab should be considered.

For many pathologies, early structural tissue changes occur at the cellular level, on the scale of micrometers or tens of micrometers. Magnetic resonance imaging (MRI) is a powerful non-invasive imaging tool used for medical diagnosis, but its clinical hardware is incapable of reaching the cellular length scale directly. In spite of this limitation, microscopic tissue changes in pathology can potentially be captured indirectly, from macroscopic imaging characteristics, by studying water diffusion. Here we focus on water diffusion and NMR relaxation in the human prostate, a highly heterogeneous organ at the cellular level. We present a physical picture of water diffusion and NMR relaxation in the prostate tissue, that is comprised of a densely-packed cellular compartment (composed of stroma and epithelium), and a luminal compartment with almost unrestricted water diffusion. Transverse NMR relaxation is used to identify fast and slow T

Background: Traditional pathology reports of prostate biopsy mainly focus on presence of carcinoma but ignore other pathologic findings such as inflammation or hyperplasia. In the era of MRI-ultrasound fusion-targeted prostate biopsy (MRF-TB), where specific MRI regions of interest (ROI) are targeted for biopsy, these benign findings should be reported as they may guide decisions on when to repeat imaging or prostate biopsy. In this study, we reviewed MRF-TB prostate biopsies reported as negative for carcinoma to identify pathologic correlates to visible ROI on prostate MRI. Design: From 2012 to2016, 1595 men underwent a total of 1813 prebiopsy prostate MRI, followed by MRF-TB at our institution. We rereviewed the prostate biopsy cores for all patients with PI-RADS 4 or 5 (PI-RADS 4/5) ROI but had no cancer detected on MRF-TB. Pathologic findings were separated into two groups: significant pathologic findings (SPF, such as inflammation, hyperplasia, ASAP/HGPIN) and no significant pathologic findings (NSPF) with or without cancer in same/adjacent site on systematic biopsy (SB). Patients with repeat MRI and follow-up MRF-TB evaluation. Results: 497 men had PI-RADS 4/5 lesions out of 1595 initial biopsies. Of these 497 men, 101 (20%) had MRF-TB negative for carcinoma. Upon review, 54 had SPF and 47 had NSPF on MRF-TB. Of 54 men with SPF on initial MRF-TB, 31 had repeat MRI, 23 of 31 men downgraded in which 16 had repeat MRF-TB with 1 had cancer detect. The other 8 of 31 men had persistent PI-RADS 4/5 lesions, 3 were detected cancer on repeat MRF-TB. Of 47 men with NSPF on initial MRF-TB, 19 had PCa in the same/ adjacent site on SB and were considered as missed on MRF-TB; of the other 28, 13 underwent repeat MRI. 8 of 13 downgraded with 0 had PCa in the repeat MRF-TB and 5 of 13 men with persistent PI-RADS 4/5 lesions, 3 had PCa detect on repeat MRF-TB. Altogether, 22/47 (47%) of the cases with NSPF in the initial MRF-TB were missed cancer. Conclusions: 1/5 of the target biopsy cases on PI-RADS 4/5 ROI had negative cancer detection. Inflammation, nodular hyperplasia and HGPIN can account for some of the cases, and those were downgraded in followup MRI usually had a negative repeat biopsy. Cases with NSPF on MRF-TB for PI-RADS 4/5 lesions are likely (47%) missed PCa, high likelihood of persistent PI-RADS 4/5 ROI on repeat MRI and PCa detection on repeat biopsy. We suggest pathology findings beside cancer should be reported on MRF-TB biopsy as they can guide decisions on repeat imagine and biopsy

Background: Urothelial carcinoma of the upper tract (UTUC) is relatively rare. Urine cytology has been used in the diagnosis and surveillance of UTUC but its utility is not well established. Another method of UTUC detection is via endoscopic biopsy which itself can be limited due to inadequate sampling. Our study aims to explore the efficacy of urine cytology alone and in combination with endoscopic biopsy results in detecting UTUC. Design: We searched our pathology database for cases with both urine cytology specimen and subsequent histologic followup including endoscopic biopsies, ureterectomy, nephrectomy or nephroureterectomy over a 10 year period. Urine cytology and biopsies done concurrently or within 6 months follow-up are included in the study and the highest degree of abnormality was selected if multiple specimen available. For cases with both biopsies and surgical specimen diagnoses, the final diagnosis from surgery was used. Results: 154 cases of confirmed UTUC were included in the study. Among them, urine diagnoses of suspicious or positive for malignancy were made in 106 cases (69%): 7 cases were low grade UC (LGUC) and 99 cases were high grade UC (HGUC) upon surgical pathology evaluation. 32/154 cases (21%) were atypical urothelial cells (AUC) on cytology, among which 27 had confirmed HGUC and 5 cases had LGUC on surgical pathology. 16/154 urine cytology cases (10%) were negative: 11/16 had HGUC and 5/16 cases had LGUC in the concurrent or follow-up biopsies or surgery. The sensitivity for abnormal urine cytology was 89.6% for detecting UTUC. We also compared the urine cytology and endoscopic biopsies (within 6 months) in detecting UTUC. There are 62 cases are positive for both methods. Four cases show positive urine but negative biopsy, and 1 with negative urine but positive biopsy (HGUC). The sensitivity for combined urine cytology and biopsy is slightly higher for one method alone. Conclusions: Urine cytology is highly sensitive for detecting UTUCespecially for HGUC. Combining urine cytology and endoscopic biopsy results increases the sensitivity for detecting UTUC and should therefore be recommended for clinical practice

Background: Renal cell carcinoma (RCC) with leiomyomatous stroma constitutes a group of uncommon renal epithelial tumors with clear cells and prominent fibromuscular stroma. It is unclear whether it is a distinct type of RCC with distinct genomic changes, or whether it is related to clear cell RCC (CCRCC) or clear cell papillary RCC (CCPRCC). Design: 11 RCC with leiomyomatous stroma were selected based on morphology (clear cells with prominent fibromuscular capsule and stroma and positive CK7/CD10/CA9). Three of these tumors along with 13 CCRCC and 11 CCPRCC were subject to array-based comparative genomic hybridization. Genomic DNA was extracted from formalinfixed, paraffin-embedded tumor and adjacent normal tissue and hybridized to Affymetrix SNP 6.0 Array chips, which contained 1.8 million genetic markers, including 906,600 single nucleotide polymorphisms (SNPs) and 946,000 probes for copy number variation (CNV) detection. Selected immunohistochemical stains were also performed. Results: For patients with RCC with leiomyomatous stroma, mean age was 50.3 (48-84) years. Mean tumor size was 1.9 (0.9-3.5) cm. Grade was 1, 2 and 3 in 2, 5 and 4 cases, respectively. Pathologic stage was T1a in all cases. The dominant morphological feature was clear cells forming arborizing glands and acini with branching contour and basally oriented nuclei and abundant apical cytoplasm with blebs. All tumors had prominent fibromuscular capsule and stroma with various amounts of smooth muscle fibers. The epithelial cells were positive for CK7, CD10 and CA9. The following genomic changes were present in at least 2 of 3 RCC with leiomyomatous stroma: loss- 8p12-q21.11 (containing TCEB1 gene), 7q11.21-22, 19p12-13; gain- 18q21.1-22.1. None of these changes was, however, present in CCRCC and CCPRCC. In 13 CCRCC, the dominant genomic changes (in >7/13 tumors) included loss of 3p, 9p, Yp and gain of 11q. In 11 CCPRCC, the dominant genomic changes (in >6/11 tumors) were gain of 3p Conclusions: This study provides evidence that RCC with leiomyomatous stroma, CCRCC and CCPRCC are distinct tumors in their morphological, immunohistochemical and genomic features. RCC with leiomyomatous stroma comprises CK7/CD10/CA9 positive clear cells forming arborizing glands and acini with branching contour and basally oriented nuclei and abundant apical cytoplasm with blebs, and prominent fibromuscular capsule and stroma. Loss of 8p12-q21.11 (containing TCEB1), 7q11.21-22, 19p12-13, and gain of18q21.1-22.1 are seen in 2 of 3 tumors

OBJECTIVE: To determine if multiparametric MRI (mpMRI) identifies significant apical disease, thereby informing decisions regarding preservation of the membranous urethra. MATERIALS AND METHODS: Men undergoing radical prostatectomy between January 2012 and June 2016 who underwent a 12-core transrectal-ultrasound guided systematic biopsy, preoperative 3-T MRI, and sectioning of the prostate specimen with tumor foci mapping were extracted from a single surgeon's prospective longitudinal outcomes database. Apical systematic biopsy vs. mpMRI lesion were compared for predicting aggressive tumor in the prostatic apex defined as Prostate Cancer Grade Group >1. RESULTS: Of the 100 men who met eligibility criteria, 43 (43%) exhibited aggressive prostate cancer in the distal 5mm of the apex. A Likert score > 2 in the apical one-third of the prostate was found to be more reliable than any cancer found on apical systematic biopsy at detecting aggressive cancer in the apex. On multivariate regression that included Likert score in the apex, age, PSA, prostate size, and presence of any cancer on apical biopsy, only Likert score (p=.005) and PSA (p=.025) were significant and independent predictors of aggressive cancer in the distal apex. CONCLUSION: MRI is superior to systematic biopsy at identifying aggressive prostate cancer within the distal prostatic apex and may be useful for planning the extent of apical preservation during prostatectomy.

OBJECTIVE: To provide insights into the role of mpMRI for predicting oncological control following two focal ablation (FA) templates for selective cases of prostate cancer (PCa). MATERIALS AND METHODS: 59 radical prostatectomies were performed between 2012 and 2016 on cases that fulfilled criteria for FA. The Gleason score (GS), extent of Gleason pattern (GP) 4, maximum linear cross sectional length (MLCSL) and location of tumor foci were recorded and related to scale on corresponding 3mm transverse slice prostate maps. Gleason pattern 4 extra-focal disease (GP4EFD) was defined as PCa with any GP 4 not detected by mpMRI and transrectal ultrasound systematic biopsy observed outside a specified ablation zone. The location of these GP4EFD relative to the MRI lesion (MRI-L) (contralateral or ipsilateral) was recorded and used to predict oncological control following a hypothetical margin and ipsilateral hemi-ablation templates. RESULTS: Overall, 15/59 (25.4%) of the prostate specimens had at least one GP4EFD. Of the total 20 GP4EFD, 7/20 (35%) were ipsilateral and 13/20 (65%) were contralateral to the MRI-L. Of the GP4EFD, 16/20 (80%), 2/20 (10%), and 2/20 (10%) were GS 3+4, GS 4+3, and GS 4+4, respectively. Of these GP4EFD, 10/20 (50%) exhibited a MLCSL < 5mm. Ablating only the MRI-L+10mm or performing a ipsilateral hemi-ablation would leave residual GP4 in 14/59 (23.7%) and 11/59 (18.6%) of cases, respectively. CONCLUSIONS: Since a significant proportion of candidates for FA based on mpMRI and systematic biopsy will have pre-existing GP4EFD outside ablation templates, active surveillance of the untreated prostate is mandatory.

Distinguishing between indolent and aggressive prostate adenocarcinoma (PCa) remains a priority to accurately identify patients who need therapeutic intervention. SIN3B has been implicated in the initiation of senescence in vitro Here we show that in a mouse model of prostate cancer, SIN3B provides a barrier to malignant progression. SIN3B was required for PTEN-induced cellular senescence and prevented progression to invasive PCa. Furthermore, SIN3B was downregulated in human PCa correlating with upregulation of its target genes. Our results suggest a tumor suppressor function for SIN3B that limits PCa progression, with potential implications for the use of SIN3B and its target genes as candidate diagnostic markers to distinguish indolent from aggressive disease.