Faculty Bibliography

Background: The NCCN currently recommends microsatellite instability (MSI) or mismatch repair (MMR) immunohistochemistry (IHC) for patients with endometrial cancer (EC) as a screen for Lynch Syndrome and criteria for immune checkpoint inhibitor therapy. Recently, MSI testing has become available through next generation sequencing (NGS). We sought to compare the concordance of MMR IHC and MSI status obtained via NGS in EC.
Design(s): Patients with newly diagnosed primary EC were prospectively consented to an IRB-approved protocol. ECs were subjected to MMR IHC and targeted NGS with a panel covering over 400 cancer-related genes. MSIsensor was used to bioinformatically infer the MSI status, with an MSIsensor score >=10 deemed MSI-high. Tumor cell content was inferred bioinformatically using ABSOLUTE. Concordance between IHC and sequencing results was defined using Cohen's Kappa.
Result(s): 175 ECs were included (116 endometrioid, 17 serous, 11 carcinosarcoma, 13 mixed, 6 clear cell, 7 dedifferentiated, 5 other). 50 (29%) were considered MMRd based on the loss of at least one MMR protein by IHC, of which 30 (60%) ECs were classified as MSI-high by MSIsensor (Table). Of the 125 ECs with retained MMR expression, the vast majority (123; 98%) was molecularly concordant and was not MSI-high (Table). The overall agreement between MMRd and MSI-high by MSIsensor was 153/175 (87%), with a Kappa of 0.749 (good agreement). Of the 20 MMRd classified as non-MSI-high, 9 were MSI indeterminate and 11 microsatellite stable. Of these 11 discrepant ECs (i.e. MMRd and microsatellite stable), 4 displayed loss of MLH1 and PMS2 expression, 1 PMS2 loss and 6 MSH6 loss of expression (Table). The tumor purity was significantly lower in the 20 MMRd non-MSI-high ECs (median 26%, range 19-60%) compared to the 30 MMRd MSI-high concordant ECs (median 49%, range 20-95%; p<0.001). (Table presented)
Conclusion(s): Our findings revealed a good agreement between MMR IHC and MSI status inferred from NGS for EC. Tumor purity may falsely decrease the degree of MSI in EC. However, in addition to MSI status, multi-gene sequencing assays provide information on specific somatic/ germline mutations, copy number alterations and tumor mutational burden in a single assay

Background: Clear cell papillary renal cell carcinoma (ccpRCC) is a relatively new entity and it has been described as an indolent renal neoplasm. ccpRCC shares features of clear cell RCC (ccRCC) and papillary RCC (pRCC) morphologically and immunohistochemically, even though it carries a very different prognostic potential. Epigenetic alterations play a significant role in the development and progression of human tumors. A large scale sequencing efforts demonstrated that hypermethylation in RCC tumors is associated with poor prognosis and may dictate treatment options. However, ccpRCC has not been included and further molecular elucidation is to be done. In this study, we attempt to investigate the methylation patterns in ccpRCC and test if differential patterns can be correlated with histologic subtypes and known biological behaviors.
Design(s): Nephrectomy specimens from our institution were reviewed and 8 cases from 2017-2019 were selected and confirmed as ccpRCC by three pathologists. Tumors were microdissected and DNA from FFPE was extracted and profiled using the Illumina MethylationEPIC array. Methylation data were analyzed with the R Bioconductor package minfi, including quality control, data normalization and differentially methylated CpG site analysis. Subsequent filtering was performed using a p-value cutoff = 0.01 and a minimum mean difference of the Beta-value of 0.1. Clustering was performed using tSNE analysis. Copy numbers were analyzed using conumee package. The methylation data were compared to ccRCC and pRCC from publicly available TCGA dataset.
Result(s): All 8 cases passed QC metrics on bisulfite conversion, hybridization and signal intensity confirming that the DNA quality was optimal for methylation study. ccpRCC clustered very tightly together illustrating that they represented a homogeneous group of tumors. When this cluster was compared to the TCGA dataset, ccpRCC was found to be located in between ccRCC and pRCC which might explain the characteristic features of this tumor subtype (figure 1). (Figure presented)
Conclusion(s): 1. DNA methylation is a useful molecular hallmark of many cancers including renal cancers and is increasingly utilized in diagnosis, prognostication, and clinical trials (epigenetic therapy). 2. Morphologically and immunohistochemically confirmed ccpRCC forms a tight cluster validating the use of methylation assay for future studies. 3. Promotor and pathway specific methylation patterns will be further studied which can distinguish the indolent clinical behavior

Background: Recent clinical guidelines for management of prostate adenocarcinoma are aimed at expanding active surveillance (AS) to include men with intermediate-risk (Gleason score 3+4=7) disease on needle biopsy (NB). However, studies reported a large portion of men with Gleason 3+4=7 prostate cancer on biopsy, that harbored adverse surgical pathologic findings. It remains unclear which subset of intermediate-risk patients with Gleason score 7 cancers can be safely treated with AS. In this study we investigate whether the percentage of Gleason pattern 4 in NB with Gleason score 7 cancers is an indicator for stratifying risk.
Design(s): We retrospectively reviewed our electronic record database for patients that underwent core NB over a 6-year period. We included NB with Gleason score 3+3=6 (G336), 3+4=7 with <5% Gleason pattern 4 (G4%<5) and 3+4=7 with 6-49% maximum Gleason pattern 4 (G4%6-49); all cases had corresponding radical prostatectomy (RP) within 6 months of the biopsy. We defined adverse pathology (AP) as any RP with Gleason score equal to or greater than 4+3=7 and/or stage T3 or higher. We compare AP outcomes in final follow-up RP of three NB groups: G336, G4%<5 and G4%6-49.
Result(s): A total of 289 NB with corresponding radical prostatectomies were identified. The breakdown of Gleason groups is shown in Table 1. GS336 has an AP rate of 26.6%, while G4<5% an AP rate of 20%. In comparison, the group of patients with G4%6-49 exhibited a 42% rate of AP (Table 1). A Chi-square test performed comparing AP of G4%<5 and G%6-49 is statistically significant p= .0237 (Figure 1). Conversely, there is no statistical difference between the rate of AP in G4<5% and G336, p= 0.46. G336 and G4%<5 were aggregated into a new group G%0-5 with an AP rate of 25.2% compared to G%6-10 AP rate 39.6%, p= 0.0576 (Figure 2). G4%6-10 and G4%11-49 had comparable rates of AP, 39.6% and 43.1%, respectively (p=0.681). (Table presented)
Conclusion(s): Currently there is a paradigm shift amongst pathologist and the significance of minimal percentage pattern 4 on prostate biopsies. Our current data supports the recent literature publications that <5% maximal Gleason pattern 4 on a single core has a similar rate of adverse pathological outcomes as Gleason score 3+3=6 and can be considered for AS. Although we did not detect a statistical difference between the rate of AP between G4%0-5 and G4%6-10, the data is beginning to approach statistical significance and warrants further risk stratification

Background: Low risk prostate cancers are amenable to active surveillance which can reduce harmful overtreatment of indolent disease. However there is great variability of criteria in urological practice for determination of active surveillance candidacy. The quantity of Gleason pattern 4 is an important prognostic parameter and may influence treatment decisions. A genomic classifier, the Oncotype DX Genomic Prostate Score has been used to predict both clinical risk and tumor aggressiveness in patients diagnosed on biopsy with low risk prostate cancer (Grade group (GG) 1 and 2). This study investigated whether Genomic Prostate Score (GPS) can predict unfavorable disease and correlates with percentage of Gleason pattern 4 in low grade prostate cancer.
Design(s): We searched our surgical pathology database for prostate biopsies with Oncotype DX Genomic Prostate Score reports (2016- 2019). Oncotype Dx was performed on the single core with worst disease (core with longest tumor and/or maximum percentage of pattern 4). Biopsy results including Gleason Score and length of the tumor and percentage of pattern 4 from the core submitted for Oncotype test were recorded. Follow-up repeat biopsy or prostatectomy, if performed, were also retrieved for review of Gleason Score. Oncotype GPS score and related clinical information were analyzed in the study.
Result(s): 306 prostate biopsy cases with Oncotype DX test report were included in the study. Among these cases, 124 cases were originally diagnosed as GG1 (Gleason Score 3 + 3) and 182 cases were GG2 (3 + 4). The average GPS in GG 2 is significantly higher than GG1 (28.52 +/- 11.80 vs 17.88 +/- 9.35, p < 0.0001). Forty cases in GG1 had follow up repeat biopsy or prostatectomy. Twenty cases were upgraded to GG2. Cases with higher GPS score are more likely to upgrade to GG2 (23.10 +/- 10.13 vs 16.55 +/- 8.22, p < 0.05) in follow up repeat biopsy or prostatectomy. In GG1 group, GPS score correlated with the maximum tumor length and tumor percentage (p < 0.01). In GG2, patients with Gleason pattern 4 greater than 30% received higher GPS score than patients with pattern 4 less than 30% (p < 0.05). GPS significantly correlated with percentage of Gleason pattern 4 but not length of pattern 4, length of tumor, PSA level or PSA density.
Conclusion(s): In GG1, Oncotype Dx GPS can predict the likelihood of unfavorable disease at follow up repeat biopsy or prostatectomy. In GG2, GPS score correlated with percentage of pattern 4, supporting its role as an auxiliary tool for clinical risk classification