Search

PD-L1 immunohistochemistry (IHC) currently has the most Food and Drug Administration (FDA) approvals as a companion diagnostic (CDx) for immunotherapies in specific tumor types; however, multiple other immunotherapy biomarkers exist. We performed this study to examine and report the prevalence of PD-L1 expression in a wide variety of tumor types and examine its relationship to microsatellite instability (MSI), tumor mutational burden (TMB), and CD274 (PD-L1) gene amplification. We performed a retrospective analysis of all cases in which both PD-L1 IHC (using the DAKO 22C3 IHC assay with either tumor proportion score (TPS) or combined positive score (CPS); or the VENTANA SP142 assay with infiltrating immune cell score (IC)) and comprehensive genomic profiling (CGP) were tested at Foundation Medicine between January 2016 and November 2019. Of note, PD-L1 positivity is defined per the CDx indication and tumor proportion score (TPS ≥ 1) for indications without a CDx claim; and TMB positivity is defined as ≥10 mutations/Mb. A total of 48,782 cases were tested for PD-L1 IHC and CGP. Immune cell expression of PD-L1 was more frequently identified than tumor cell expression of PD-L1. We saw a high correlation between PD-L1 expression and CD274 gene amplification (p

Introduction: ASXL1 mutations are frequently seen across the clinical spectrum of myeloid neoplasia. The most commonly identified ASXL1 mutation represents a single base duplication within an 8-guanine repeat at nucleotide position 1934 (c.1934dupG). Due to technical limitations of sequencing homopolymer regions, the ASXL1 c.1934dupG variant has been identified as potential artifact in some sequencing assays, though modern next generation sequencing assays and bioinformatics pipelines can generally accurately detect this mutation. However, a comprehensive comparison of ASXL1 c.1934dupG mutations versus non-c.1934dupG ASXL1 mutations have not been performed to date. Thus, we sought to explore a large dataset to determine if any biologic differences existed between these two groups. Methods: Comprehensive genomic profiling by FoundationOne ®Heme testing was performed on patient samples with known or suspected myeloid neoplasms (MN). All MN patients ≥18 years old with 1 or more mutation were identified by internal database query. Patients were categorized as acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), non-chronic myeloid leukemia myeloproliferative neoplasms (MPN), or MDS/MPN overlap based on mutation and outside clinical and pathology data. Mutations with variant allele fractions (VAF) >1% were included for analysis, except for the ASXL1 c.1934dupG variant, which was only reported if the VAF was ≥15%. Fisher's exact tests were used to evaluate proportional differences between categorical variables, and Mann-Whitney U tests were used for comparisons of continuous variables. Results: Truncating ASXL1 mutations were identified in 1,414 included patients, occurring in 18% of AML and 26% of chronic myeloid neoplasms. Twenty-eight (2%) patients had multiple ASXL1 mutations, and ASXL1 was the sole mutated gene in 52 patients (4%). The most common ASXL1 mutation was c.1934dupG (Figure 1A), and this was the sole or dominant ASXL1 mutation in 520 cases (37%). The remaining 894 patients (63%) had one or more mutations at other sites in the ASXL1 gene (ASXL1other), with p.E635Rfs, p.R693*, and codon 591 mutations being the most common. There were no significant differences in age, sex, or ancestry signatures between ASXL1c.1934dupG and ASXL1other. We noted slightly fewer ASXL1c.1934dupG mutations in patients with MDS (ASXL1c.1934dupG: ASXL1other 0.48:1) compared to AML (0.65:1, p = 0.03) and MPN (0.60:1, p = 0.01) and those in whom ASXL1 was the sole mutation (Figure 1B). However, these trends may have been due to VAF-based reporting thresholds, as ASXL1 VAFs were lower in singly mutated patients and those with an MDS diagnosis classification. Comparison of co-mutated genes with VAFs ≥15% between ASXL1c.1934dupG and ASXL1other revealed no significant difference in median non-ASXL1 mutations (each median 4, IQR 2-5, p = 0.74). When individual genes were assessed, co-mutation rates of STAG2 (p = 0.01) and KMT2A (p = 0.02) were higher in ASXL1c.1934dupG MNs, while SETBP1 (p = 0.01) mutations were more common with ASXL1other. In all MNs, the absolute differences in the frequency of mutations in ASXL1c.1934dupG versus ASXL1other were small. However, some differences emerged within phenotypic subgroups (Figure 1C). For instance, KMT2A rearrangements and STAG2 mutations were strongly associated with ASXL1c.1934dupG in MDS/MPN and MPN, with ASXL1c.1934dupG: ASXL1other ratios of 5:1 (p = 0.03) and 9:1 (p < 0.001), respectively. In contrast, AML patients with TP53 or SETBP1 mutations had a significantly higher mutation rate in ASXL1other (TP53: 11% vs. 3% in ASXL1c.1934dupG, p < 0.01; SETBP1: 14% vs. 7%, p=0.04). We further identified that other specific ASXL1 mutations were more commonly co-mutated in AML with TP53 (ASXL1 p.R693*, p < 0.001) or SETBP1 (ASXL1 p.R404*, p < 0.001). Conclusion: Our results confirm the ASXL1 c.1934dupG variant occurs in a similar patient population to other ASXL1 mutations, and further supports its pathogenicity in myeloid neoplasia. Subset analysis suggests that ASXL1c.1934dupG and ASXL1other may be associated with certain phenotypic and co-mutational tendencies. Thus, ASXL1 mutation site may be an important variable in some patients and should be considered in future mechanistic and clinical studies. Further study is warranted to determine whether clinical outcomes are affected by different ASXL1 mutations. Figure 1 Figure 1. Disclosures Haberberger: Foundation Medicine, Inc.: Current Employment. Ferguson: Foundation Medicine Inc: Current Employment, Other: ownership.

Background Carcinoma of the urethra (UrthCa) is an uncommon Genitourinary (GU) malignancy that can progress to advanced metastatic disease. Methods One hundred twenty-seven metastatic UrthCa underwent hybrid capture-based comprehensive genomic profiling to evaluate all classes of genomic alterations (GA). Tumor mutational burden was determined on up to 1.1 Mbp of sequenced DNA, and microsatellite instability was determined on 114 loci. PD-L1 expression was determined by IHC (Dako 22C3). Results Forty-nine (39%) urothelial (UrthUC), 31 (24%) squamous (UrthSCC), 24 (19%) adenocarcinomas NOS (UrthAC), and 12 (9%) clear cell (UrthCC) were evaluated. UrthUC and UrthSCC are more common in men; UrthAC and UrthCC are more common in women. Ages were similar in all 4 groups. GA in PIK3CA were the most frequent potentially targetable GA; mTOR pathway GA in PTEN were also identified. GA in other potentially targetable genes were also identified including ERBB2 (6% in UrthUC, 3% in UrthSCC, and 12% in UrthAC), FGFR1-3 (3% in UrthSCC), BRAF (3% in UrthAC), PTCH1 (8% in UrthCC), and MET (8% in UrthCC). Possibly reflecting their higher GA/tumor status, potential for immunotherapy benefit associated with higher tumor mutational burden and PD-L1 staining levels were seen in UrthUC and UrthSCC compared to UrthAC and UrthCC. Microsatellite instability high status was absent throughout. Conclusions Comprehensive genomic profiling reveals GA that may be predictive of both targeted and immunotherapy benefit in patients with advanced UrthCa and that could potentially be used in future adjuvant, neoadjuvant, and metastatic disease trials.

421 Background: PDAC has a propensity for early systemic dissemination and most pts with primary tumors radiographically confined to the pancreas harbor micro-metastases. cfDNA based CGP offers a non-invasive mechanism to identify actionable genomic targets in PDAC and account for serial clonal evolution in response to therapeutic selection pressure. We interrogated the Foundation Medicine database to characterize cfDNA based CGP in pts with PDAC. Methods: We performed a retrospective analysis of pts with PDAC who underwent cfDNA based CGP between May 2016 and February 2020. cfDNA based CGP was done utilizing either the FoundationOne Liquid (F1-Liquid) or the Foundation-ACT (F-ACT) testing platform. Samples were interrogated for exonic and select intronic regions of cancer-related genes for both F1-Liquid (Exons: 70 genes; Select Introns: 7 genes) and F-ACT (Exons: 59 genes; Select Introns: 6 genes). Variant zygosity and somatic/germline status (SGZ) for short variant mutations was computationally predicted without matched normal tissue using an investigational method. A Gaussian Mixture Model (GMM) was used to identify latent molecular classes in samples harboring somatic, pathogenic variants (n=613); the Bayesian Information Criteria (BIC) was used for model selection. Results: We identified 1,009 pts with cfDNA based CGP - median age was 65, 55% were male, and median cfDNA concentration was 24 ng/µL. cfDNA yield had an influence on the detection of somatic alterations (p

422 Background: Early systemic dissemination is a salient feature of PDAC and the ability to reliably detect as well as monitor cfDNA based alterations may have predictive and/or prognostic value. However, a clear correlation between the clonality of the primary tumor and micro-metastatic disease has not been established in PDAC. We examined the correlation between T-CGP and cfDNA based CGP in pts with PDAC. Methods: We performed a retrospective analysis of PDAC pts who had both T-CGP (October 2010 – March 2020) and cfDNA based CGP (May 2016 – March 2020), with an average of 645 days between tests (Range: 21 – 2101 days) For T-CGP, exonic regions and select intronic regions of cancer related genes were sequenced to a high uniform depth (>500X median coverage) using FoundationOne Heme (Exons: 315 genes; Select Introns: 28 genes) and FoundationOne CDx (Exons: 310 genes; Select Introns: 34 genes). cfDNA based CGP, was performed using either FoundationOne Liquid (Exons: 70 genes; Select Introns: 7 genes) or Foundation-ACT (Exons: 59 genes; Select Introns: 6 genes) platform. Results: Eighty-one pts were identified with both T-CGP and cfDNA based CGP - median age was 61 years, 59.3% (48/81) were male and 66.2% (49/74) had reported metastatic disease. The median tumor DNA content available for T-CGP analysis was 112.3 ng. The median cfDNA concentration was 1.7 ng/µL, and the number of reported ct-DNA based somatic alterations correlated significantly with the concentration of ct-DNA (Spearman’s R = 0.19, p < 10-8). 75/81 (92.6%) and 41/81 (50.6%) of pts had at least one pathogenic somatic alteration detected by T-CGP and cfDNA based CGP, respectively. KRAS, TP53 and CDKN2A were the most frequently altered genes in both T-CGP and cfDNA based CGP (Table1). Homologous Recombination DNA Damage Repair (HR-DDR) gene alterations [ BRCA2, ATM, and CHEK2 alterations] were detected in 7.4% (6/81) of T-CGP and 7.4% (6/81) cfDNA based CGP, respectively. Two BRAF fusions were detected from samples tested with T-CGP, however these were not detected with the cfDNA assay. Conversely, an ALK-EML4 fusion was detected using both platforms. Conclusions: The concordance between T-CGP and cfDNA based CGP in HR-DDR gene alterations may be indicative of the germline status of some of these alterations; nevertheless, it suggests the predictive utility of cfDNA based CGP in PDAC. The differential prevalence of other gene alterations between the two modalities is likely reflective of the cfDNA yield available for CGP and to a lesser extent, clonal evolution. cfDNA based CGP holds promise as a complimentary predictive tool to T-CGP. [Table: see text]