Sokol, Ethan S.

PURPOSE Poly ADP-ribose polymerase inhibitors (PARPi) are approved for patients with human epidermal growth factor receptor 2–negative metastatic breast cancer (mBC) and germline pathogenic/likely pathogenic variant (hereafter mutation) in the BRCA1/ 2 genes (g BRCA); however, clinical benefit has also been demonstrated in mBC with somatic BRCA1/ 2 mutations (s BRCA) or germline PALB2 mutations (g PALB2). This study aims to describe the genomic landscape of homologous recombination repair (HRR) gene alterations in mBC and assess PARPi treatment outcomes for patients with g BRCA compared with other HRR genes and by status of a novel homologous recombination deficiency signature (HRDsig). METHODS A real-world (RW) clinico-genomic database (CGDB) of comprehensive genomic profiling (CGP) linked to deidentified, electronic health record–derived clinical data was used. CGP was analyzed for HRR genes and HRDsig. The CGDB enabled cohort characterization and outcomes analyses of 177 patients exposed to PARPi. RW progression-free survival (rwPFS) and RW overall survival (rwOS) were compared. RESULTS Of 28,920 patients with mBC, g BRCA was detected in 3.4%, whereas the population with any BRCA alteration or g PALB2 increased to 9.5%. HRDsig+ represented 21% of patients with mBC. BRCA and g PALB2 had higher levels of biallelic loss and HRDsig+ than other HRR alterations. Outcomes on PARPi were assessed for 177 patients, and g BRCA and s BRCA/ gPALB2 cohorts were similar: g BRCA versus s BRCA/g PALB2 rwPFS was 6.3 versus 5.4 months (hazard ratio [HR], 1.37 [0.77-2.43]); rwOS was 16.2 versus 21.2 months (HR, 1.45 [0.74-2.86]). Additionally, patients with HRDsig+ versus HRDsig– had longer rwPFS (6.3 v 2.8 months; HR, 0.62 [0.42-0.92]) and numerically longer rwOS (17.8 v 13.0 months; HR, 0.72 [0.46-1.14]). CONCLUSION Patients with s BRCA and g PALB2 derive similar benefit from PARPi as those with g BRCA alterations. In combination, HRDsig+, s BRCA, and g PALB2 represent an additional 19% of mBC that can potentially benefit from PARPi. Randomized trials exploring a more inclusive biomarker such as HRDsig are warranted.

PURPOSE APOBEC mutagenesis underlies somatic evolution and accounts for tumor heterogeneity in several cancers, including breast cancer (BC). In this study, we evaluated the characteristics of a real-world cohort for time-to-treatment discontinuation (TTD) and overall survival on CDK4/6 inhibitors (CDK4/6i) plus endocrine therapy (ET) and immune checkpoint inhibitors. METHODS Comprehensive genomic profiling results from 29,833 BC samples were analyzed for tumor mutational burden and APOBEC signatures. For clinical outcomes, a deidentified nationwide (United States–based) BC Clinico-Genomic Database (CGDB) was evaluated with log-rank and Cox models. Patients with hormone receptor–positive (HR+) human epidermal growth factor receptor 2–negative (HER2–) BC who received first-line ET and CDK4/6i were included. Eligible patients from Mayo Clinic and Duke University were HR+ HER2– BC with sequencing data between September 2013 and July 2020. RESULTS Of 29,833 samples sequenced, 7.9% were APOBEC+ with a high rate in invasive lobular carcinoma (16.7%) and in metastatic tumors (9.7%) relative to locally biopsied BC (4.3%; P < .001). In CGDB, 857 patients with HR+ HER2– BC received ET plus CDK4/6i in the first line. APOBEC+ patients had significantly shorter TTD on ET plus CDK4/6i than APOBEC– patients, 7.8 (95% CI, 4.3 to 14.6) versus 12.4 months (95% CI, 11.2 to 14.1; hazard ratio, 1.6; 95% CI, 1.03 to 2.39; P = .0036). Clinical benefit to immune checkpoint inhibitors was observed in HR+ HER2–, APOBEC+, tumor mutational burden–high patients, with four of nine CGDB patients (TTD 0.3-11.3 months) and four of six patients in Duke/Mayo cohorts (TTD 0.9-40.5 months) with a TTD of ≥ 3 months. CONCLUSION APOBEC+ HR+ HER2– patients had shorter TTD on first-line ET plus CDK4/6i relative to APOBEC– patients. Further research is needed to optimize the treatment of APOBEC+ HR+ HER2– BC and to investigate the efficacy of immunotherapeutic strategies in this population.

Abstract Background: APOBEC mutagenesis underlies somatic evolution and accounts for tumor heterogeneity in several cancers, including BC. In our clinical experience, HR+HER2- BC with an APOBEC signature do poorly on standard of care (SOC) first line endocrine therapy (ET) + CDK4/6 inhibitor (CDK4/6i) and need additional treatment options. Here we evaluated the characteristics of a real-world cohort for time-to-treatment discontinuation (TTD) and overall survival (OS) on SOC and immune checkpoint inhibitors (ICI).Methods: Hybrid capture-based comprehensive genomic profiling (CGP) results from 29,833 formalin-fixed paraffin-embedded tissue biopsies of all BC subtypes were analyzed. For outcomes (TTD and OS), this study used a de-identified nationwide (US-based) BC clinico-genomic database (CGDB, ~800 sites of care, 1/2011 - 12/2020). HR+HER2- metastatic BC pts who received first line ET + CDK4/6i were included (CGDB cohort).TTD was defined as the difference between the first and last drug episode within a given line of treatment (LOT). LOT were derived based on FH algorithms. OS was defined as the time from LOT start to the date of death or data cutoff. Log-rank test and Cox model were used to evaluate the difference in outcomes. To reduce the impact of confounding variables (Age at Dx, Stage at Dx, Tumor Type, Metastases sites, TMB group (≥10 vs &lt;10), SOC treatment group, PIK3CA), inverse probability of treatment weighting (IPTW) was used. Eligible pts from Mayo Clinic and Duke University were HR+HER2- mBC with sequencing data from FMI between 9/2013-7/2020. Clinical data were manually extracted from Mayo and Duke EHR. Results: Of all 29,833 BC samples in the CGP cohort, 7.9% were APOBEC+ with high rate in invasive lobular carcinoma (ILC) 16.7% vs. 4.9% in invasive ductal carcinoma (IDC) and metastatic lesions 9.7% vs. 4.3% from breast. APOBEC+ samples had a higher median TMB 12.5 vs. 2.5 mut/Mb. In CGDB, 857 HR+HER2- BC met inclusion criteria; 69 (8%) pts were APOBEC+ and 788 (9.2%) were APOBEC-. APOBEC+ pts had significantly shorter TTD on SOC ET+CDK4/6i than APOBEC- pts, 7.8 (95% CI 4.3-14.6) vs. 12.4 (95% CI 11.2-14.1) months (p=0.0036). APOBEC+ pts also had noticeably shorter OS compared to APOBEC- pts, 32.4 (95%CI 19.8-47.4) vs. 40.5 (95%CI 36.9-45.7) months (p=0.06).Cox regression results indicate that the relative risk of shorter TTD for the APOBEC+ vs the APOBEC- was 1.6 (95%CI 1.03-2.39). Also, APOBEC+ pts had almost twice the risk that APOBEC- pts had of death (HR=1.96, 95%CI 1.2-3.3). In CGDB, there were 10 APOBEC+ pts who received ICI, 9/10 had evaluable TTD data, 4/9 were still on treatment as of Dec 2020. 5/9 received ICI monotherapy, 4 pts received ICI + chemotherapy. TTD ranged from 0.3 to 11.3 mo, 1 pt’s TTD was &gt; 6 mo. In Mayo and Duke cohort, there were 6 pts, 5/6 received ICI + chemotherapy. The TTD was 0.9-40.5 months with longest 2 pts receiving 5-FU plus ICI (11 and 40.5 months). To better understand the ICI treatment landscape, TTD in HR+HER2- hTMB MSS APOBEC- CGDB cohort (N=6) was analyzed: 5/6 had evaluable treatment data, 4/5 finished ICI treatment, 1 pt’s TTD was &gt; 3 mo, 0/5 had TTD &gt; 6 mo. Conclusions: APOBEC+ occurs in ~7% of BC and is more common in ILC and metastatic lesions. APOBEC+ HR+HER2- pts had shorter TTD and OS on SOC ET+CDK4/6i relative to APOBEC- pts. However, TTD on ICI tended to be longer in APOBEC+ pts, but our data is limited, and more research is needed. CGDB APOBEC+ vs. APOBEC- with SOC 1st lineAPOBEC+ (N=69)APOBEC- (N=788)p adjusted (FDR)*Age at Dx, Median (IQR)59.0 (53.0, 65.0)56.0 (47.0, 65.0)0.102Stage at Dx0.067- 0-III52 (75.4%)505 (64.1%)- IV11 (15.9%)245 (31.1%)- Not documented6 (8.7%)38 (4.8%)Metastasis free interval, yrs, Median (IQR)5.2 (3.0, 10.5)5.1 (2.8, 9.3)0.734Tumor Grade0.104- Grade 12 (2.9%)49 (6.2%)- Grade 234 (49.3%)259 (32.9%)- Grade 311 (15.9%)163 (20.7%)- Not documented22 (31.9%)317 (40.2%)Tumor Type0*- IDC7 (10.1%)251 (31.9%)- ILC30 (43.5%)138 (17.5%)- Other /Not documented32 (46.4%)399 (50.6%)Metastases sites0.734- Bone-only17 (24.6%)169 (21.5%)- CNS11 (15.9%)104 (13.2%)- Visceral41 (59.4%)514 (65.3%)TMB, Median (IQR)11.3 (8.8, 18.8)2.5 (1.3, 3.8)0*MSI0.946- MSI-H0 (0.0%)4 (0.5%)- MSI-I0 (0.0%)1 (0.1%)- MSS68 (98.6%)728 (97.2%)- Not documented1 (1.4%)16 (2.1%)BRCA5 (7.2%)45 (5.7%)0.734PIK3CA46 (66.7%)341 (43.3%)0.001*Treatments0.005*- AI + CDK4/6i31 (44.9%)510 (64.7%)- Fulvestrant + CDK4/6i38 (55.1%)278 (35.3%) Citation Format: Saranya Chumsri, Kira Raskina, Sarah Sammons, Laura Alder, Natalie Danziger, Alexa B Schrock, Kim McGregor, Ethan Sokol. APOBEC signature, clinical characteristics, and outcome in hormone receptor-positive (HR+) HER2-negative (HER2-) breast cancer (BC) patients (pts) in real-world data (RWD) [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD14-09.

Abstract Background: FGFR dysregulation is observed in multiple cancers and targeting FGFR is an emerging therapeutic strategy with FDA approved treatments in bladder and cholangiocarcinoma. Here we examined the prevalence of FGFR mutations, fusions, and high-level amplifications in breast cancer, stratified by receptor subtype and local/metastatic status, in both Foundation Medicine (FM) and institutional Mayo Clinic (MC) cohorts. Methods: For the FM cohort, comprehensive genomic profiling (CGP) examining at least 324 genes for all classes of alterations, including FGFR1-4 was carried out for 32,048 breast cancers during the course of routine clinical care in a Clinical Laboratory Improvement Amendments (CLIA)-certified lab (Foundation Medicine Inc., Cambridge, MA, USA). Tumor mutational burden (TMB) was determined on up to 1.1 Mb, microsatellite instability high (MSI-High) was determined on up to 114 loci and predicted ancestry from &gt;10,000 SNPs. Estrogen receptor (ER) and HER2 status were available for a subset of FM samples. Additionally, 131 patients with metastatic breast cancer from a subset of patients at three Mayo Clinic sites (MC cohort) with clinical characteristics and cancer-panel DNA sequencing data from a CLIA-certified lab (Tempus, Chicago, IL) were included. Results: In the FM cohort, the prevalence of FGFR1-4 high-level amplification (CN≥10) was 10.1%, while mutations (1.5%) and fusions (0.72%) were rare. Most amplifications occurred in FGFR1 (9.2%); most fusions and mutations occurred in FGFR2 (0.46%, 0.77%). FGFR alteration prevalence was highest in ER+/HER2- subtype (14.4%) and lowest in HER2+ disease (7.7%). FGFR alterations were more common in IDC (11.7%) than ILC (7.7%), p&lt;3E-08. FGFR alterations were more prevalent in the metastatic setting relative to breast-biopsied disease (13.6% v 10.1%; OR = 1.4; p=2E-17), especially in the HER2+ (OR =1.9, p=0.004) and ER-/HER2- (OR = 1.9, p = 0.05) disease; no enrichment was seen in the ER+/HER2- metastases (OR =1.0, p = 1). FGFR amplifications were observed at a higher prevalence in patients with predicted East Asian ancestry, relative to patients with European ancestry (12.1% v 10.0%; p = 0.03). Overall, the most common activating mutations in FGFR were FGFR2 N549K (n=85), FGFR1 N546K (n=78), FGFR4 V510M (n=28), FGFR2 K659E (n=28), FGFR4 V510L (n=20), and FGFR2 Y375C (n=15). The most common recurrent fusions were FGFR3:TACC3 (n=36), FGFR2:TACC2 (n=17), FGFR1:TACC1 (n=9), FGFR1:BAG4 (n=6), and FGFR2:ATE1 (n=5). In patients with FGFR amplifications, the most frequently co-occurring alterations were ZNF703 (78.4%), TP53 (51.5%), CCND1 (36.1%), FGF3/4/19 (32.9 - 34.4%), PIK3CA (30.7%), MYC (29.6%), ESR1 (17.2%), EMSY (16.3%), and PTEN (10.6%). Significant co-occurrence was observed for a number of genes including FGF3/4/19, CDK4, and CDK8 (all OR&gt;2, p&lt;1E-07); mutual exclusivity was observed with PIK3R1, BRCA1, and BRCA2 (all OR &lt;0.5, p&lt;4E-13), among other genes. In the 131 metastatic tumors from MC, the prevalence of FGFR1-4 high-level amplifications was 19.8% [FGFR1 (12.4%), FGFR2 (7.4%), and FGFR3 (0.8%)]. The prevalence of high-level FGFR amplifications did not differ by clinical subtypes: HR-/HER2- (7/31), HR+/HER2- (15/79), and HER2+ (2/11), p=0.68. Conclusions: High-level FGFR amplifications are observed in &gt;11% of breast cancers, especially the ER+/HER2- subtype, while mutations/fusions are rare. These data support the ongoing studies evaluating targeted therapies for FGFR amplified ER + breast cancer. Correlations with clinical information (MC cohort) and associations with actionable alterations are ongoing and may inform potential combination strategies. Citation Format: Karthik V Giridhar, Ethan S Sokol, Peter T Vedell, Jason P Sinnwell, Aakash Desai, Tufia C Haddad, Ciara C O’Sullivan, Roberto A Leon-Ferre, Siddhartha Yadav, Kostandinos Sideras, Brenda Ernst, Minetta C Liu, Abe Eyman Casey, Xiaojia Tang, Zoe Fleischmann, Karthikeyan Murugesan, Krishna R Kalari, Matthew P Goetz. The frequency and somatic mutation landscape of Fibroblast growth factor receptor (FGFR) alterations in breast cancer [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-08-02.

Abstract Extensive biological and clinical evidence supports the notion many CRC tumors are addicted to EGFR/KRAS/MAPK signaling. For example, mutations in KRAS, BRAF and other MAPK pathway components confer intrinsic or acquired resistance to anti-EGFR therapy. However, a significant proportion of acquired resistance cases have remained unexplained. ARID1A mutations are found in approximately 11% of CRC tumors, but a connection between ARID1A mutations and EGFR/MAPK signaling activation in CRC has not been previously suspected. By leveraging the genomic biomarker data collected from 333 patients at baseline and at progression who participated in the CALGB/SWOG 80405 trial, we found that mutations in ARID1A, a key component of SWI/SNF complex, were enriched in the cell free DNA of 6/16 of the patients whose tumors had become resistant to anti-EGFR therapy with ARID1A mutations detected at either timepoint (adj. p = 0.03, OR = 0.09, two-tailed fisher's exact test). In contrast, there was no evidence of enriched ARID1A mutations in patients treated with bevacizumab. To investigate the potential role of a broader ARID1A deficiency phenotype in resistance to cetuximab therapy, we developed an ARID1A mutant-like signature from TCGA that captures the transcriptional profile characteristic of ARID1A mutants. Using this signature to stratify CALGB patients based on gene expression data collected on tissue at diagnosis, we also found patients with an ARID1A mutant-like gene expression profile had worst outcome in patients treated with cetuximab than bevacizumab with adjustment for established clinical variables for both OS [p = 0.0002, hazard ratio (HR) 6.2, 95% confidence interval (CI) 2.4-16], and PFS [p = 0.0008, HR 4.7 (1.9 -12)], suggesting that ARID1A mutations may be also a mechanism of intrinsic resistance to anti-EGFR therapies. Gene expression analysis demonstrated that ARID1A deficiency leads to re-activation of an EGFR-like signature, suggesting reactivation of this pathway as a possible mechanism of resistance. Furthermore, CRISPR knockout of ARID1A in the cetuximab-sensitive CRC cell line NCI-H508 conferred elevated MAPK signaling relative to parental line and resistance to cetuximab in culture. Consistent with these findings, we also observed strong mutual exclusivity between ARID1A mutations and mutations in the EGFR/MAPK signaling pathway in more than 40,000 lung and colorectal cancer patients profiled in the FoundationCore® database. Strikingly, in lung cancers where EGFR mutations structural variants (SV) are more prevalent, EGFR SV was the top mutually exclusive alteration with ARID1A mutations in lung cancers. Of 5980 lung cancer patients with ARID1A and/or EGFR SVs, only 100 patients (1.7%) had both SVs. Taken together, our data suggest that ARID1A loss-of-function mutations may promote resistance to cetuximab by driving an EGFR-like transcription program in the absence of ligand-dependent activation of EGFR. Our results suggest that ARID1A defects could be potentially used as an exclusion biomarker for cetuximab treatment decisions, and they provide a rationale for exploring therapeutic MAPK inhibition in ARID1A mutant CRC patients. Citation Format: Radia Marie Johnson, Xueping Qu, Joshua Thomas, Yvonne Kschonsak, Ling-Yuh Huw, Fang-Shu Ou, Ethan Sokol, Nnamdi Ihuegbu, Oliver Zill, Omar Kabbarah, Anneleen Daemen, Richard Bourgon, Alan Venook, Federico Innocenti, Heinz-Josef Lenz, Felipe de Sousa e Melo, Carlos Bais. ARID1A mutations induce an EGFR-like gene expression signature and confer intrinsic and acquired resistance to cetuximab treatment in first line metastatic CRC [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr LB-063.

Increasing data support the importance of preexisting host immune response and neoantigen burden for determining response to immune checkpoint inhibitors (ICIs). In lung cancer and melanoma, tumor mutational burden (TMB) has emerged as an independent biomarker for ICI response. However, the significance of TMB in breast cancer, particularly in the context of PD-L1 negativity, remains unclear. This report describes a patient with HER2-negative breast cancer with high TMB and an apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like (APOBEC) trinucleotide signature; her disease was refractory to multiple lines of treatments but achieved durable complete response using ICIs and capecitabine. Additional analysis of the tumor revealed a low amount of stromal tumor-infiltrating lymphocytes (sTILs) and PD-L1 negativity, reflecting a poor preexisting host immune response. In collaboration with Foundation Medicine, comprehensive genomic profiling from 14,867 patients with breast cancer with the FoundationOne test was evaluated. Using the cutoff of ≥10 mutations/megabase (mut/Mb) for high TMB, PD-L1 positivity and TMB-high populations were not significantly overlapping (odds ratio, 1.02;P=.87). Up to 79% of TMB-high tumors with >20 mut/Mb were PD-L1–negative. Our study highlights that despite having low TILs and PD-L1 negativity, some patients may still experience response to ICIs.