Search

Background: Circulating tumor DNA (ctDNA) is being evaluated as a tool to monitor disease and guide therapy escalation and de-escalation in advanced breast cancer. The patient, disease, and treatment characteristics that influence whether tumor DNA is isolated and sequenced from the bloodstream are not well understood. We aimed to describe patient and disease characteristics of cases with undetectable ctDNA levels despite progressive metastatic breast cancer (MBC). Methods: We retrospectively identified patients (pts) with MBC who had ctDNA evaluation by the Guardant 360 assay (Redwood City, CA) as part of their routine care from 2015-2020. We correlated the ctDNA assay with the disease status at collection. We identified the patient cohort with no detectable ctDNA despite evaluation at the time of progressive disease but prior to the initiation of a new therapy (ND ctDNA/PD) and compared the clinical, pathologic, and molecular features of this group to those with detected ctDNA. Differences were tested by two proportion z-tests. Results: Of 1151 ctDNA samples collected among 473 pts with MBC, 87 (7.5%) samples had no detectable (ND) ctDNA. 54 of 87 samples with ND ctDNA were collected at a time of stable or responding disease or after a new line of therapy was started, leaving only 33 ctDNA evaluations (2.8%) among 30 pts with ND ctDNA despite collection at the time of PD prior to new therapy. Among this group there were 14 pts (47%) with HR+ HER-, 11 pts (37%) with HER2+, and 5 pts (17%) with TN MBC. This compared to 254 pts (60%) with HR+ HER2-, 82 pts (19%) with HER2+, and 85 pts (20%) with TN MBC and detectable ctDNA. HER2-positive MBC was more common in the ND ctDNA/PD group than detectable ctDNA group (p-0.02). In the ND ctDNA/PD group, the median time from MBC diagnosis to ctDNA evaluation was 7 months and pts had received a median of 0.5 prior lines of therapy for MBC. Four pts (13%) had lobular breast cancer, 24 pts (80%) had recurrent disease, 14 pts (47%) had visceral metastases, and 12 pts (40%) had oligometastatic disease. Sites of metastases at the time of ND ctDNA/PD were bone n=16 (53%), lung n=7 (23%), liver n=6 (20%), lymph node n=11 (37%), skin and soft tissue n=5 (17%), and CNS n=5 (17%). When compared to those with detectable ctDNA these differences in characteristics were not statistically significant, although numerically pts with ND ctDNA had more CNS disease (17% vs 10%) and less liver disease (20% vs 32%). At the time of ND ctDNA, the site of progression was bone n=14 (47%), CNS n=5 (17%), lymph node n=9 (30%), lung n=7 (23%), and liver n=5 (17%). There was a single site of progression in 20 pts (67%). 16 of 30 pts had repeat ctDNA analysis of which 9 pts had subsequent detectable ctDNA with a median VAF of 0.3%, a median of 1 alteration per sample, and oncogenic or likely oncogenic alterations in TP53 in 3 pts and BRCA1, CCNE1, CDH1, and PIK3CA in 1 pt each. Seven had tissue NGS, all of which showed multiple oncogenic alterations. The remaining 22 pts with ND ctDNA but no PD were responding to therapy based on imaging or had already started a new therapy since the last progression. This group also had a high proportion of HER2+ MBC (n=7, 32%) and low proportion with visceral disease (n=6, 27%) Conclusions: Although ctDNA is a highly sensitive tool to detect active MBC ( Citation Format: Ami N Shah, Lorenzo Gerratana, Shruti Chandra, Dhruvika Mukhija, Neelima Katam, Anthony K Kang, Andrew A Davis, Millen Srivastava, Saya Jacob, Paolo D'Amico, Qiang Zhang, Carolina Reduzzi, Michael Gurley, Firas Wehbe, William J Gradishar, Amir Behdad, Massimo Cristofanilli. Progressive metastatic breast cancer with no detectable circulating tumor DNA: Evaluating limitations of this highly sensitive tool [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 P2-08-04.

Background: The correlation of ERBB2 copy number gain (CNG) from tissue or circulating tumor DNA (ctDNA) by next-generation sequencing (NGS) with standard HER2 tissue evaluation is not well understood. Materials and Methods: We retrospectively identified patients with ERBB2 CNG on commercial NGS. We described their clinical-pathologic features and calculated the positive predictive value (PPV) of ERBB2 CNG by NGS for HER2-positivity by IHC and FISH testing. Results: 176 patients had NGS revealing an ERBB2 CNG (112 by tumor tissue and 91 by ctDNA). The cancer subtypes with the most cases with ERBB2 CNG by NGS were breast (n = 67), non-small cell lung (n = 25), colorectal (n = 18), gastroesophageal (n = 17), pancreatic (n = 11), and uterine (n = 11). The PPV of ERBB2 CNG in determining HER2 positivity by standard IHC/FISH definitions was 88% for tissue NGS (n = 57) and 80% for ctDNA (n = 47). The PPV among breast cancer patients for tissue NGS was 97% (n = 35) and ctDNA was 93% (n = 39). However, for non-breast cancer cases, the PPV of ERBB2 amplification by tissue NGS dropped to 76% (n = 22) and by ctDNA to 44% (n = 7). Conclusions: ERBB2 CNG by NGS is detected in numerous malignancies for which HER2 testing is not standard. Detection of ERBB2 CNG by tissue NGS and ctDNA has a high PPV for true HER2-positivity by standard IHC and/or FISH testing in breast cancer.

Supplemental Table 1: Treatments by subgroup. Supplemental Figure 1A-D: Association of baseline MAF and NOA with survival

Purpose:Circulating tumor DNA (ctDNA) is a promising tool for noninvasive longitudinal monitoring of genomic alterations. We analyzed serial ctDNA to characterize genomic evolution in progressive metastatic breast cancer.Experimental Design:This was a retrospective cohort between 2015 and 2019 obtained under an Institutional Review Board–approved protocol at Northwestern University (Chicago, IL). ctDNA samples were analyzed with Guardant360 next-generation sequencing (NGS) assay. A total of 86 patients had at least two serial ctDNA collections with the second drawn at first post-NGS progression (PN1) by imaging and clinical assessment. A total of 27 participants had ctDNA drawn at second post-NGS clinical progression (PN2). We analyzed alterations, mutant allele frequency (MAF), number of alterations (NOA), and sites of disease on imaging in close proximity to ctDNA evaluation. Matched pairs' variations in MAF, NOA, and alterations at progression were tested through Wilcoxon test. We identified an independent control cohort at Massachusetts General Hospital (Boston, MA) of 63 patients with serial ctDNA sampling and no evidence of progression.Results:We identified 44 hormone receptor–positive, 20 HER2+, and 22 triple-negative breast cancer cases. The significant alterations observed between baseline and PN1 were TP53 (P < 0.0075), PIK3CA (P < 0.0126), AR (P < 0.0126), FGFR1 (P < 0.0455), and ESR1 (P < 0.0143). Paired analyses revealed increased MAF and NOA from baseline to PN1 (P = 0.0026, and P < 0.0001, respectively). When compared with controls without progression, patients with ctDNA collection at times of progression were associated with increased MAF and NOA (P = 0.0042 and P < 0.0001, respectively).Conclusions:Serial ctDNA testing identified resistance alterations and increased NOA and MAF were associated with disease progression. Prospective longitudinal ctDNA evaluation could potentially monitor tumor genomic evolution.

The challenges of reliably collecting, storing, organizing, and analyzing research data are critical in low- and middle-income countries (LMICs), particularly in Sub-Saharan Africa where several healthcare and biomedical research organizations have limited data infrastructure. The Research Electronic Data Capture (REDCap) System has been widely used by many institutions and hospitals in the USA for data collection, entry, and management and could help solve this problem. This study reports on the experiences, challenges, and lessons learned from establishing and applying REDCap for a large US-Nigeria research partnership that includes two sites in Nigeria, (the College of Medicine of the University of Lagos (CMUL) and Jos University Teaching Hospital (JUTH)) and Northwestern University (NU) in Chicago, Illinois in the United States. The largest challenges to this implementation were significant technical obstacles: the lack of REDCap-trained personnel, transient electrical power supply, and slow/ intermittent internet connectivity. However, asynchronous communication and on-site hands-on collaboration between the Nigerian sites and NU led to the successful installation and configuration of REDCap to meet the needs of the Nigerian sites. An example of one lesson learned is the use of Virtual Private Network (VPN) as a solution to poor internet connectivity at one of the sites, and its adoption is underway at the other. Virtual Private Servers (VPS) or shared online hosting were also evaluated and offer alternative solutions. Installing and using REDCap in LMIC institutions for research data management is feasible; however, planning for trained personnel and addressing electrical and internet infrastructural requirements are essential to optimize its use. Building this fundamental research capacity within LMICs across Africa could substantially enhance the potential for more cross-institutional and cross-country collaboration in future research endeavors.

Purpose: Circulating tumor DNA (ctDNA) is a promising tool for noninvasive longitudinal monitoring of genomic alterations. We analyzed serial ctDNA to characterize genomic evolution in progressive metastatic breast cancer. Experimental Design: This was a retrospective cohort between 2015 and 2019 obtained under an Institutional Review Board–approved protocol at Northwestern University (Chicago, IL). ctDNA samples were analyzed with Guardant360 next-generation sequencing (NGS) assay. A total of 86 patients had at least two serial ctDNA collections with the second drawn at first post-NGS progression (PN1) by imaging and clinical assessment. A total of 27 participants had ctDNA drawn at second post-NGS clinical progression (PN2). We analyzed alterations, mutant allele frequency (MAF), number of alterations (NOA), and sites of disease on imaging in close proximity to ctDNA evaluation. Matched pairs' variations in MAF, NOA, and alterations at progression were tested through Wilcoxon test. We identified an independent control cohort at Massachusetts General Hospital (Boston, MA) of 63 patients with serial ctDNA sampling and no evidence of progression. Results: We identified 44 hormone receptor–positive, 20 HER2+, and 22 triple-negative breast cancer cases. The significant alterations observed between baseline and PN1 were TP53 (P < 0.0075), PIK3CA (P < 0.0126), AR (P < 0.0126), FGFR1 (P < 0.0455), and ESR1 (P < 0.0143). Paired analyses revealed increased MAF and NOA from baseline to PN1 (P = 0.0026, and P < 0.0001, respectively). When compared with controls without progression, patients with ctDNA collection at times of progression were associated with increased MAF and NOA (P = 0.0042 and P < 0.0001, respectively). Conclusions: Serial ctDNA testing identified resistance alterations and increased NOA and MAF were associated with disease progression. Prospective longitudinal ctDNA evaluation could potentially monitor tumor genomic evolution.

Introduction: Previously reported data from our group and others have demonstrated an increase in genomic complexity and number of alterations over the course of treatment in MBC. Our study aims to categorize the pathogenicity of genomic aberrations found on serial ctDNA evaluations in MBC to explore the relevance of emerging mutations. Methods: Patients with MBC and ctDNA next-generation sequencing (NGS) analysis by Guardant360 (Guardant Health, Redwood City, CA) completed from 2015 to 2019 were retrospectively identified with an IRB-approved protocol. Clinical-pathologic features and time point for progression of disease were abstracted from review of the electronic medical record. Among the 255 patients with ctDNA analysis, 85 patients had serial ctDNA analysis from baseline and at subsequent progression of disease (PD1) and 27 had repeat ctDNA analysis at next progression (PD2). Genomic alterations were classified as oncogenic, likely oncogenic, predictive oncogenic, likely benign, inconclusive, unknown significance, and synonymous based on categorization by OncoKB.org (Chakravarty et al, JCO PO 2017). Oncogenic, likely oncogenic, and predicted oncogenic were considered ‘pathogenic’ while the remainder were ‘not known pathogenic’. Two-sample t-tests for difference of means (α = 0.05) was used for analysis of changes in pathogenic proportions and mutant allele frequencies (MAF). Results: The median age was 53.8 years, 41 patients had hormone receptor positive (HR+) HER2-negative, 22 had HER2+, and 22 had triple negative (TN) MBC. The median and interquartile range (IQR) of the time of MBC diagnosis to first ctDNA collection was 10 mo (0.3-23.1 mo), from baseline ctDNA to analysis at the time of disease progression (PD1) was 6.7 mo (3.6-12.5 mo), and from PD1 to second progression (PD2) was 4.6 mo (3.1-8.5 mo). At baseline 66% of alterations were pathogenic (38.4% oncogenic, 27.6% likely oncogenic, 26.5% unknown significance). The proportion of pathogenic variants did not significantly change over time: 62% at PD1 (p=0.41) and 56% at PD2 (p=0.19). The percent of pathogenic alterations by disease subtype at baseline and PD1 was 66% and 60% in HR+ HER2-, 73% and 57% in HER2+, and 62% and 71% in TN. 57% and 70% of patients had a new pathogenic alteration at PD1 and PD2, respectively. The most frequently detected new pathogenic alterations at the time of disease progression were TP53, PIK3CA, ESR1, FGFR1, and MYC, which occurred in 28%, 18%, 15%, 12%, and 11% of patients, respectively. Among genes that were altered in at least 10% of patients, the aberrations were very likely to be pathogenic (>75% of alterations pathogenic) for TP53, PIK3CA, ESR1, MYC, BRAF, FGFR1, often pathogenic (50-75% of alterations pathogenic) for NF1, KIT, GATA3, ERBB2, EGFR, CCNE1, PDGFRA, and often not known pathogenic ( Citation Format: Ami N Shah, Andrew A Davis, Saya Jacob, Lorenzo Gerratana, Brian Finkelman, Shruti Chandra, Neelima Katam, Firas Wehbe, Jeeven Srivastava, Ashwin Sunderraj, Qiang Zhang, Leonidas Platanias, Amir Behdad, William Gradishar, Massimo Cristofanilli. Evaluating the pathogenicity of emerging genomic aberrations detected by circulating tumor DNA over the course of metastatic breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-26.

Background: Inflammatory breast cancer (IBC) has a distinctive and aggressive clinical behavior but its underlying biological characteristics have not been fully elucidated. The extended analysis of somatic alterations in The Cancer Genome Atlas (TCGA) highlighted canonical oncogenic pathways that were consistently represented across different tumor subtypes. The aim of this study was to translate such pathway-based characterization to the clinical setting through ctDNA NGS to dissect IBC’s biology and prognosis. Methods: The study retrospectively analyzed 255 metastatic breast cancer (MBC) patients (pts) treated and characterized for ctDNA at Northwestern University (Chicago, IL). ctDNA was analyzed using the Guardant360 NGS assay (Guardant Health). Only non-synonymous alterations were analyzed. Pathway classification was defined based on prior work (Sanchez-Vega F et al, Cell. 2018). Associations among clinical characteristics, pathway classification, and IBC were explored through uni- and multivariate logistic regression; survival was tested though uni- and multivariate Cox regression both for progression-free survival (PFS) and overall survival (OS). Results: Of 255 enrolled pts, 124 (48%) were diagnosed with hormone receptor positive (HR pos) MBC, 75 (30%) with HER2-positive (HER2_pos) MBC and 56 (22%) with triple negative (TNBC) MBC. IBC was diagnosed in 74 pts (30%). Receptor-tyrosine kinase, RTK (130 pts, 51%), p53 (130 pts, 51%), PI3K/Akt (116 pts, 46%), and cell cycle (91 pts, 36%) were the most often altered pathways. The multivariate model highlighted the association of IBC with HER2_pos (OR: 2.19; 95%CI: 1.09 - 4.38; P=0.0276), an increased number of alterations in the p53 pathway (OR: 2.05; 95%CI: 1.12 - 3.75; P=0.0197) and a decreased number of alterations in the RAS pathway (OR: 0.34; 95%CI: 0.14 - 0.80; P=0.0137). Decreased alterations in the ER pathway were borderline significant (OR: 0.48; 95%CI: 0.22 - 1.03; P=0.0584). Only cell cycle alterations had an impact on PFS for IBC (HR: 2.20; 95%CI: 1.18 - 4.08; P=0.0127), while p53 and Wnt had an impact on nonIBC (respectively HR: 2.00; 95%CI: 1.23 - 3.25; P=0.0052 and HR: 3.40; 95%CI: 1.20 - 9.64; P=0.0212). The univariate model showed a significant impact on OS RAF, ER, and cell cycle pathways alterations for IBC, the role of ER and cell cycle pathways alterations was confirmed in the multivariate model (respectively HR: 6.19; 95%CI: 1.63 - 23.48; P=0.0073 and HR: 3.79; 95%CI: 1.04 - 13.75; P=0.0431). The multivariate model showed a prognostic impact only for p53 in the nonIBC subgroup (HR: 2.20; 95%CI: 1.11 - 4.36; P=0.0237) Conclusion: The ctDNA-based oncogenic signaling pathway characterization showed different biological and prognostic features across IBC and nonIBC MBC patients. Alterations of the p53 pathway were more likely to be present in IBC pts, while alterations in the RAS pathway were less represented in this cohort. ER and cell cycle pathways’ alterations impacted the OS of IBC MBC patients. Although preliminary, these results suggest a more comprehensive biological characterization based on ctDNA for treatment selection and clinical decision-making. Main alterations and pathwayGeneNumber of alterations%PathwayTP5318815.58p53PIK3CA14111.68PI3K/AktERBB2725.97RTKESR1705.8ERMYC574.72MycFGFR1463.81RTKEGFR453.73RTKCCNE1383.15Cell cycleMET322.65RTKNF1322.65RAS Citation Format: Lorenzo Gerratana, Andrew A Davis, Marco Mina, Saya L Jacob, Qiang Zhang, Ami N Shah, Paolo D’Amico, Neelima Katam, Firas Wehbe, Elena Vagia, Lisa Flaum, Leonidas Platanias, Amir Behdad, Fabio Puglisi, William J Gradishar, Giovanni Ciriello, Massimo Cristofanilli. Investigating oncogenic signaling pathways in inflammatory metastatic breast cancer (MBC) though circulating tumor DNA (ctDNA) next-generation sequencing (NGS) [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-33.

Background: Circulating tumor DNA (ctDNA) has emerged as a potential tool for detecting disease recurrence, monitoring response to therapy, and identifying resistance mutations in the peripheral blood. With increased frequency of testing, there is an unmet need to recognize putative germline variants in ctDNA, and the probability that these variants are associated with inherited conditions. Here, we evaluated a large cohort of breast cancer patients who underwent ctDNA evaluation to determine the type and frequency of ctDNA mutations identified with confirmed germline testing. Methods: We reviewed ctDNA testing from a single institution (Northwestern University). All breast cancer patients who had next-generation sequencing (NGS) performed by Guardant Health (Redwood City, CA) from 2015-2020 were included in this retrospective study. An allele frequency cutoff of 30% was pre-established as a threshold to review patient charts to determine whether genetic counseling and germline testing were performed, along with the timeframe of this testing (e.g. before or after ctDNA evaluation). Clinical information including demographics, pathology, tissue NGS testing, and germline testing were collected. Descriptive analyses and statistical associations were performed using STATA. Results: The initial cohort consisted of 520 patients with breast cancer who underwent ctDNA testing. From this, we identified 84 patients (16.2%) who had at least one variant with allele frequency ≥30%. The most common variants identified were the following: TP53 (34%), PIK3CA (27%), BRCA1 (9%), BRCA2 (8%), and AKT1 (4%). Guardant360 classified 99% of these variants as pathogenic and 1% as a variant of unknown significance. Germline positivity using a separate CLIA-approved test for this indication was confirmed at the following frequencies: BRCA1 (2 of 8 positive, 25%), BRCA2 (2 of 5 positive, 40%), PIK3CA (0 of 5 positive), and TP53 (0 of 26 positive). In total, 14% of patients with ctDNA allele frequency ≥30% had a confirmed germline mutation. Lower age at breast cancer diagnosis was significantly associated with the probability of germline testing prior to ctDNA evaluation (P=0.0001). For patients who had a variant with allele frequency ≥30%, 24.3% never received genetic counseling or germline testing. Conclusion: High allele frequency ctDNA variants (≥30%) were present in 16% of patients who underwent ctDNA evaluation with 14% of these variants confirmed as true germline variants. Consenting patients for ctDNA testing should include the possibility of identifying putative germline variants, and criteria should be established to refer patients for subsequent genetic counseling and germline testing, given the potential implications for patients and their family members. Citation Format: Andrew A Davis, Michael C Burns, Lorenzo Gerratana, Paolo D'Amico, Saya Jacob, Ami Shah, Neelima Katam, Firas Wehbe, Qiang Zhang, Elena Vagia, Lisa Flaum, Kalliopi P. Siziopikou, Leonidas C Platanias, Amir Behdad, William J Gradishar, Massimo Cristofanilli. Identification of incidental putative germline variants in circulating tumor DNA [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-08.

Background: As the role of precision medicine in metastatic breast cancer (MBC) expands, there is an increasing emphasis on diagnostic tools to characterize both somatic and germline alterations that drive carcinogenesis. Circulating tumor DNA (ctDNA) has provided an important means for real-time, non-invasive detection and monitoring of tumor somatic alterations without standard paired white blood cell testing. However, its application for the detection of germline mutations remains relatively understudied. Here, we characterize the role of BRCA1&2 ctDNA mutant allele frequency (MAF) to identify underlying putative germline BRCA1&2 mutations. Methods: Patient data were retrospectively obtained under an IRB-approved protocol to review ctDNA data at Northwestern University between 2015 and 2020. All ctDNA samples were analyzed using the Guardant360 next-generation sequencing (NGS) assay (Guardant Health). Patients with ctDNA alterations in the BRCA genes were identified at all mutant allele frequencies (MAF). Clinical data were collected including breast cancer subtype, prior lines of therapy, tissue-based NGS and germline testing information. Separate CLIA-approved testing performed per standard of care was used as the gold standard for germline testing. Statistical analysis was used to test the association of MAF cut-offs with the presence or absence of a germline alteration. Receiver operating characteristic (ROC) analysis was performed. Results: We identified 127 patients with breast cancer who underwent ctDNA collection. There were 69 HR+ HER2-, 24 HER2+, and 34 triple negative breast cancer (TN) patients. Of these, 8 patients had known BRCA1 germline mutations, as confirmed with separate germline testing, and 9 patients had known BRCA2 germline mutations. BRCA1 germline mutations were significantly association with age < 40 (p=0.016) and triple negative subtype (p=0.042). Mean ctDNA BRCA1 MAF was 2.27% (Standard Deviation {SD} 10.19%) and mean BRCA2 MAF was 2.04% (SD 9.51%). BRCA1/2 MAF was analyzed with respect to germline mutations through ROC analysis. For BRCA1 ctDNA MAF cut-off 32.4% was associated with confirmed putative germline mutation. This cutoff demonstrated sensitivity of 1 and specificity of 0.99. Area under the curve (AUC) was 1.00 at this cut-off. For BRCA2, ctDNA MAF cutoff of 28.5% was associated with confirmed putative germline mutations. This cutoff demonstrated sensitivity of 0.86, specificity of 0.99 and AUC of 0.93. Discussion: ctDNA alterations of BRCA1 with MAF>32.4% and BRCA2 with MAF>28.5% in clinical ctDNA testing were useful criterion to identify germline BRCA mutations. The data suggest the potential to utilize ctDNA as a diagnostic tool to predict germline mutations based on MAF thresholds. Future prospective studies are needed to determine how ctDNA testing may be incorporated into existing clinical algorithms to refer patients for germline testing. Citation Format: Saya Jacob, Andrew Davis, Paolo D'Amico, Lorenzo Gerratana, Michael Burns, Ami Shah, Neelima Katam, Firas Wehbe, Qiang Zhang, Elena Vagia, Lisa Flaum, Kalliopi Siziopikou, Leonidas Platanias, Amir Behdad, William Gradishar, Massimo Cristofanilli. Circulating tumor DNA (ctDNA) as a diagnostic tool to identify putative germline BRCA mutations [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-10.

PURPOSE Circulating tumor DNA (ctDNA) has emerged as a promising noninvasive biomarker for baseline characterization and longitudinal monitoring of a tumor throughout disease management. The aim of this study was to evaluate the utility of ctDNA across a wide spectrum of tumor types. METHODS We retrospectively identified 1,763 patients with advanced cancers who had next-generation sequencing of ctDNA or tumor tissue completed by a designated commercial assay at Northwestern University. RESULTS ctDNA identified at least one gene alteration in 90% of patients. The number of detected alterations (NDA) and mutant allele frequency (MAF) of the most frequently mutated genes varied significantly across tumor types, with the highest MAF observed in gastric, colorectal, and breast cancers and the highest NDA observed in colorectal, lung squamous, and ovarian/endometrial cancers. TP53 was the most mutated gene in all tumor types. PIK3CA, ERBB2, BRCA1, and FGFR1 alterations were associated with breast cancer, and ESR1 mutations were exclusively detected in this tumor type. Colorectal cancer was characterized by alterations in KRAS and APC mutations, whereas KRAS, EGFR, PIK3CA, and BRAF mutations were common in lung adenocarcinoma. Concordance between blood and tissue sequencing was notably observed for truncal gene alterations (eg, APC and KRAS), whereas low concordance was often observed in genes associated with treatment resistance mechanisms (eg, RB1 and NF1). Tumor mutational burden (TMB) varied significantly across tumor types, and patients with high MAF or NDA had a significantly higher TMB score with one of the investigated platforms. CONCLUSION The study provided new insights into the ctDNA mutational landscape across solid tumors, suggesting new hypotheses-generating data and caveats for future histotype-agnostic workflows integrated with tissue-based biomarkers such as TMB.

Patients with a history of malignancy have been shown to be at an increased risk of COVID-19-related morbidity and mortality. Poorer clinical outcomes in that patient population are likely due to the underlying systemic illness, comorbidities, and the cytotoxic and immunosuppressive anti-tumor treatments they are subjected to. We identified 416 cancer patients with SARS-CoV-2 infection being managed for their malignancy at Northwestern Medicine in Chicago, Illinois, between March and July of 2020. Seventy-five (18.0%) patients died due to COVID-related complications. Older age (>60), male gender, and current treatment with immunotherapy were associated with shorter overall survival. Laboratory findings showed that higher platelet counts, ALC, and hemoglobin were protective against critical illness and death from COVID-19. Conversely, elevated inflammatory markers such as ferritin, d-dimer, procalcitonin, CRP, and LDH led to worse clinical outcomes. Our findings suggest that a thorough clinical and laboratory assessment of infected patients with cancer might help identify a more vulnerable population and implement more aggressive proactive strategies.

2634 Background: High tumor mutational burden (TMB-H) has shown promise as a predictive biomarker in certain tumors, but broad applicability across tumors is unclear. In 2020, pembrolizumab was approved for treatment refractory metastatic or unresectable solid tumors with TMB ≥10 mutations/megabase based on exploratory analysis of the KEYNOTE-158 study. Growing evidence suggest limitations of TMB-H across tumors. Our aim is to evaluate outcomes of patients (pts) with advanced microsatellite-stable (MSS) solid tumors with TMB ≥10 mutations/megabase treated with immune checkpoint inhibitors (ICIs) and identify genetic alterations that predict poor response to therapy. Methods: Pts treated between 01/2015 to 12/2020 at Robert H Lurie Cancer Center, Northwestern University, with ICIs and TMB ≥10 mutations/megabase on next generation sequencing by FoundationOne CDx, Tempus xT or Guardant360 CDx platforms were identified. Tumors other than carcinomas and microsatellite instability were excluded. Radiologic response was assessed using iRECIST criteria. Responders (CR/PR) were compared with non-responders (SD/PD). Subgroup analyses were conducted based on tumor type, genetic alterations within signaling pathways. Survival curves were calculated using Kaplan-Meier method. Multivariate analyses were performed to determine impact of clinical variables on response. Results: Of 119 pts, median age was 68 yrs, and 61% of pts were male. 15 tumor types were represented, most commonly melanoma (33%), NSCLC (28%), small cell lung carcinoma (5.9%), colorectal carcinoma (5%), and pancreatobiliary carcinoma (5%). When evaluating for efficacy (n = 106), ORR was 36% with CR 14%. Median PFS and OS overall were 10.9 mo and 29.9 mo. Lack of exposure to previous therapy was associated with response (p = 0.039). PD-L1 status, specific ICI regimen, age, and sex did not have prognostic significance. When comparing melanoma/NSCLC (n = 66) and other tumor types (n = 40) there was no significant difference in response (p = 0.3). When comparing response by specific gene mutations, TP53 (p = 0.026) and PIK3CA (p = 0.025) were associated with worse response. Similarly, ROS1 alteration trended to poorer response (p = 0.057). When evaluating genetic pathways, mutations within TP53 pathway were associated inferior response (p = 0.018). There was no significant difference in RTK/RAS, NRF2, PI3K, TGFβ, WNT, MYC, Cell Cycle, and Notch pathways. Conclusions: Growing evidence shows limitations of TMB-H as a predictive biomarker for ICI therapy in MSS tumors. Our data suggests alterations within TP53 and its signaling pathway, PIK3CA mutations, and likely ROS1 alterations are associated with non-responders in MSS TMB-H solid tumors. Further analysis of tumor-specific cohorts and genetic alterations will be presented.

Background: We describe the genomic landscape of circulating tumour DNA (ctDNA) across pathological subtypes of metastatic breast cancer. Methods: 255 clinically annotated patients with ctDNA testing by Guardant360 were stratified into HR+, HER2+, and TNBC cohorts. Frequency and heterogeneity of alterations were reported. Paired ctDNA and tissue sequencing were compared for a subset of patients. The association of ctDNA and metastatic sites of disease on imaging was also assessed. Findings: 89% of patients had at least one ctDNA alteration detected. The most common single nucleotide variants (SNVs) for HR+ patients were PIK3CA, ESR1, and TP53. For HER2+, these were TP53, PIK3CA, and ERBB2 with ERBB2 as the most frequent copy number variant (CNV). For TNBC, the most common SNVs were TP53 and PIK3CA, and the most frequent CNVs were MYC, CCNE1, and PIK3CA. TNBC patients had a significantly higher mutant allele frequency (MAF) of the highest variant compared to HR+ or HER2+ patients (P

10602 Background: As the use of circulating tumor DNA (ctDNA) is more widely implemented, incidental identification of pathogenic variants reflecting germline alterations in cancer predisposition genes will occur more frequently. Such mutations are expected to have a high mutant allele frequency (MAF) or occur in genes typically associated with inherited syndromes. When a similar analysis was conducted by our group, we found that MAF of about 30% or greater in BRCA1/ 2 was associated with confirmed putative germline mutations in patients with breast cancer (Jacob et al & Davis et al, SABCS 2020). In this study, we extended this analysis to non-breast malignancies. Methods: Patients with non-breast solid malignancies and ctDNA testing between 2015-2020 were retrospectively identified from Northwestern Medicine. All ctDNA was analyzed using Guardant 360 (Guardant Health, Inc. Redwood City, Ca). Patients with ctDNA samples with mutations at high MAF (>30%) and those with BRCA1/2 mutations at any MAF were identified. We reviewed these charts for referral to genetic counselors and/or CLIA-approved germline testing. Descriptive analysis was reported for these findings. Genetic alterations were classified as pathogenic or of unknown significance based on OncoKB (Chakravarty et al, JCO PO 2017). Results: We identified ctDNA samples of 548 patients with non-breast solid malignancies, of whom 56 had gene mutations occurring at high MAF (>30%). Predominant cancer subtypes were lung (48%), colorectal (21%), pancreatic (7%), ovarian (3.5%), prostate (3.5%), and gastroesophageal (3.5%). The most common gene mutations identified were TP53 (46%), BRCA1/2 (18%), EGFR (18%), APC (13%), and KRAS (9%). 87.5% were pathogenic and 12.5% were of unknown significance. 11 patients (19.6%) had germline testing of whom 6 tested positive. These germline mutations were in BRCA2 (n = 3), EGFR, APC, and TP53. In addition to the 10 patients with BRCA1/2 mutations at high MAF (>30%), we identified 70 patients with BRCA1/2 mutations at low MAF ( < 30%). 54% were pathogenic and 46% were of unknown significance. 11 patients (14%) had germline testing of whom 3 tested positive for BRCA2, all at high MAF. 1 patient with a BRCA2 mutation at low MAF of 1.4% tested positive for a different germline BRCA2 variant. Conclusions: In patients with advanced cancers, ctDNA analysis can reveal variants with MAF >30% that are reflective of a germline mutation. Unfortunately the rate of genetic testing in these patients was low (20%). Future studies with germline testing in patients with high MAF variants would help understand the prevalence of germline variants. This can facilitate developing a more standardized approach for genetic counselor referral to identify families that may benefit from interventions for early detection or prevention of future cancers.

1084 Background: Metastatic breast cancer (MBC) is associated with genomic evolution, representing a challenge at clinical progression. While tissue and blood next-generation sequencing (NGS) allows for the baseline detection of alterations, non-invasive longitudinal assessment of ctDNA can provide a tool for monitoring tumor evolution. We characterized genomic changes using serial ctDNA testing in patients with clinical progression. Methods: Patient data was obtained under an IRB-approved protocol and ctDNA was collected at Northwestern University between 2015-2019. All ctDNA samples were analyzed using the Guardant360 NGS assay. Of 255 patients with MBC, 86 had at least two serial ctDNA collections with the second collection drawn at first progression (P1) by imaging and clinical assessment. Participants were followed until second clinical progression (P2). We analyzed type of alterations, mutant allele frequency (MAF), number of alterations (NOA), and sites of disease on imaging in close proximity to ctDNA evaluation. Matched pairs variations in MAF and NOA at P1 and P2 were tested through Wilcoxon test. Results: We identified 44 HR+, 20 HER2+ and 22 TNBC cases. Median lines of therapy were 3 (interquartile range (IQR): 1-6) for HR+, 3 (IQR: 1-5) for HER2+, and 2 (IQR: 1-4) for TNBC. The most likely alterations between baseline to P1 were TP53 (p < 0.0075), PIK3CA (p < 0.0126), AR (p < 0.0126), FGFR1 (p < 0.0455) and ESR1 (p < 0.0143). In the HR+ subset , ESR1 was statistically more likely at P1. ESR1 at P1 was also associated with development of new liver lesions (p < 0.0320). ERBB2 mutation at P1 was associated with new lung (p < 0.0050) or bone lesions (p < 0.0030). Increase in NOA was observed between baseline and P1 (p < 0.0001), P1 and P2 (p < 0.0001), and baseline to P2 (p < 0.0004). MAF was increased between baseline and P2 (p < 0.0480). Conclusions: Serial ctDNA testing identified resistance alterations ( TP53, PIK3CA, AR, ESR1, FGFR1), with some mutations indicating new sites of disease ( ESR1, ERBB2). Heterogeneity of ctDNA was significantly associated with progressive disease. Prospective evaluation of the impact of serial ctDNA testing on treatment decisions is needed to expand the role of precision medicine in MBC. [Table: see text]

3565 Background: In advanced breast cancer (ABC) HER2 status is based on ASCO/CAP immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) criteria. Next generation sequencing (NGS) of tissue and blood can detect aberrations in ERBB2 such as copy number gain/amplifications (cng/amp) and mutations. Methods: We retrospectively identified patients (pts) seen at Northwestern University between 2015 and 2019 with ABC and an alteration in ERBB2 identified by tissue and/or circulating tumor DNA (ctDNA) NGS. We included pts with testing by Guardant360, TempusX, and/or FoundationOne platforms. NGS reports were evaluated for non-synonymous mutations and cng/amp. HR and HER2 status were determined based on the most recent pathologic assessment. Mutations were categorized as pathologic if they were consider oncogenic (level 1-2 evidence with direct functional data), likely oncogenic, or predicted oncogenic, based on OncoKB (Chakravarty et al., JCO PO 2017). Results: 109 cases of ABC (6 locally advanced, 103 metastatic) with ERBB2 alterations were identified. Tissue NGS was available from 43%, ctDNA from 72%, and both from 19%. The positive predictive value (PPV) of ERBB2 amp/cng by tissue NGS to predict HER2+ using the gold standard as IHC/FISH was 94% (33/35). The PPV of ERBB2 amp by ctDNA was 93% (40/43). ERBB2 mutations were detected in 52 pts. Of these, 23 pts were considered to harbor pathologic ERBB2 mutations, (19 oncogenic, 2 likely oncogenic, 1 predicted oncogenic) detected by ctDNA and tissue in 4, ctDNA in 16, and tissue in 3 pts. The most frequently detected mutations were V777L and S310. Four pts had co-mutations of ERBB2 V777L and S310F. Disease subtype among those with ERBB2 pathologic mutations was HR+ HER2- in 57%, HER2+ in 26%, and triple negative in 17%. In all patients with serial ctDNA analysis and pathologic ERBB2 mutations, the mutation was detected on the first analysis. Pathologic ERBB2 mutation represented the mutant with the highest mutant allele frequency (MAF) in 30% and top 3 highest MAF in an additional 35%. PIK3CA was co-mutated in 48%. Conclusions: The PPV of ERBB2 amp/cng by tissue and ctDNA NGS was high, and has potential utility for cancers where HER2 IHC/FISH is not standardly assessed or cases where biopsy is challenging. ERBB2 pathologic mutations were found in all breast cancer subtypes. When present, they were identified on the initial ctDNA analysis and often represented a significant clone, supporting its role as a ‘driver mutation’.

Background: The clinical utility of routine clinical assessment of ctDNA in metastatic breast cancer (MBC) has not been clearly established. Recently, the Food and Drug Administration approved alpelisib (Piqray®) with a ctDNA-based companion diagnostic. Therefore, there is significant interest in describing the landscape of ctDNA alterations and establishing further clinical utility of this assessment. Methods: We retrospectively identified ctDNA data from a single institution (Northwestern University). All patients had next-generation sequencing performed by Guardant Health (Redwood City, CA) from 2015-2019 for clinical evaluation, and patients were consented for participation in a prospective registry study. Reports were evaluated for type of genomic alterations, number of alterations, and mutant allele frequency (MAF) of the dominant clone. Clinical information including demographics, pathology, imaging, treatment, and response data were collected via patient chart review. Associations were tested through Fisher’s exact test or Kruskal-Wallis test as statistically appropriate. Matched paired samples were investigated through McNemar test. Results: The cohort consisted of 255 patients with MBC. In total, 371 ctDNA reports were reviewed, including 116 patients with serial assessments. Baseline ctDNA evaluation occurred at a median of 3 lines of therapy (interquartile range (IQR): 1 - 5). Median number of alterations was 4 (IQR: 2-7), while median MAF was 4% (IQR: 0.6% - 17.9%). Across the entire cohort, PIK3CA and TP53 were the most commonly detected mutations, while MYC, FGFR1, PIK3CA, and ERBB2 were the most commonly detected amplifications. 40% of patients had PIK3CA aberrations observed in the analyzed plasma samples (28% amplification, 22% H1047R, 11% E542K), which was independent of breast cancer subtype. In the subset of patients with hormone-receptor positive (HR+) breast cancer, 30% of patients had ESR1 mutations (27% D538G, 19% Y537S, 12% Y537N, 11% E380Q). Patients with triple negative breast cancer (TNBC) had statistically higher MAF compared to HR+ or HER2+ breast cancer (P Conclusion: Liquid biopsy evaluation using ctDNA identified clinically meaningful genomic alterations at high frequency. In addition, our data indicate the potential clinical utility of ctDNA as a surrogate for tumor burden on imaging based on the observed correlation with lines of therapy received and sites of disease. Serial assessment of ctDNA demonstrated clonal evolution, particularly in the HR+ cohort, and identified resistant clones. Validation with an independent cohort is warranted. Citation Format: Andrew A Davis, Saya Jacob, Lorenzo Gerratana, Ami N Shah, Firas Wehbe, Neelima Katam, Qiang Zhang, Lisa Flaum, Kalliopi Siziopikou, Leonidas C Platanias, William J Gradishar, Amir Behdad, Massimo Cristofanilli. Landscape of circulating tumor DNA (ctDNA) in metastatic breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P5-01-08.

Purpose We determined Semaphorin 7a (Sema7a) localization and abundance in naive corneas and in corneas after nerve-transecting lamellar flap surgery, and determined the effect of Sema7a supplementation on corneal nerve regeneration and inflammation. Methods Immunolocalization and Western blot analyses were performed to evaluate the abundance of Sema7a in naive corneas and corneas undergoing nerve regeneration after lamellar corneal surgery in thy1-YFP+ neurofluorescent mice. We used compartmental cultures of dissociated trigeminal ganglion cells to determine the effect of Sema7a exposure on neurite outgrowth in vitro. Finally, a Sema7a pellet was implanted under the corneal flap after lamellar transection surgery to determine the neuronal and inflammatory effects of Sema7a supplementation in vivo. Results Sema7a was expressed in the corneal epithelium and stromal keratocytes, but was more abundant in the epithelium (74.3%) compared to the stroma (25.7%, P = 0.02). Sema7a expression was increased significantly in the cornea after lamellar corneal surgery and was localized to stromal cells near the regenerating nerve fronds. Exposure of trigeminal neurites to Sema7a (20 nM) in the side compartment increased neurite length significantly. The implanted Sema7a pellet increased significantly YFP+ inflammatory cell influx into the cornea as well as increased corneal nerve length. Conclusions Sema7a is expressed constitutively in the cornea, and potently stimulates nerve regeneration and inflammatory cell influx. Therefore, this immune semaphorin links nerve regeneration and inflammatory processes in the cornea.