Metastatic Tumor Samples Research Articles

Abstract B003: Real-world (RW) characterization and frequency of TSC1 and/or TSC2 alterations collected from tumor tissue and liquid biopsies from the Tempus genomic database in patients with advanced cancer

Abstract Background Tumor suppressor genes tuberous sclerosis complex subunit 1 and 2 (TSC1, TSC2) are critical negative regulators of mTORC1 activity. Patients (pts) harboring inactivating alterations in TSC1 and/or TSC2 (TSC1/2) in the AMPECT PEComa trial (NCT02494570) and in a pan-tumor Expanded Access Program (NCT03817515) demonstrated response to the mTORC1 inhibitor, nab-sirolimus. PRECISION I (NCT05103358), a currently enrolling tumor-agnostic study, will assess the clinical benefit of nab-sirolimus in pts with cancer with inactivating TSC1/2 alterations. To appreciate the potential for TSC1/2 targeted therapy, here we characterize TSC1/2 alterations across a large RW pt population with advanced cancer using data from tissue and liquid biopsies (LBs). Methods De-identified RW data for 154965 cancer pts profiled with the Tempus xT tumor DNA sequencing assay were obtained from the Tempus database (accessed 28 Apr 2023). Records included next-generation sequencing (NGS) data from the Tempus xT tumor DNA and/or xF LB sequencing assays. Somatic TSC1/2 variants were categorized as pathogenic/likely pathogenic mutations, deletions (copy number loss), or variants of uncertain significance (VUS). Inactivating TSC1/2 mutations and copy number variants were further characterized including frequency across primary vs metastatic tumor samples. Results Of the xT patients, pathogenic/likely pathogenic inactivating TSC1/2 alterations were identified in 2621 (1.7%) pts; TSC1/2 VUS (primarily missense) were identified in 9918 (6.4%) pts. Inactivating TSC1/2 variants consisted of frameshifts (35.9%), stop gains (35.8%), multihits (18.4%), deletions (7.0%), and missense mutations (3.0%), and were most frequently identified in urothelial carcinoma (UC) and pancreatic neuroendocrine tumor (each 8.4%), renal clear cell carcinoma (6.3%), hepatocellular carcinoma (HCC; 5.9%), and endometrial carcinoma (3.9%). Pathogenic TSC1 variants were more common (TSC1 56.6%; TSC2 43.4%), as was their detection among primary (~20380) vs metastatic (~70790) tumors (TSC1 1.1% vs 0.91%; TSC2 0.82% vs 0.60%). Of the 53502 xF pts with LB sequencing, inactivating TSC1/2 mutations were identified in 489 (0.9%) pts; TSC2 variants were slightly more common than TSC1 (TSC2 58.5%; TSC1 41.5%). The most common tumor types harboring inactivating TSC1/2 variants in LBs were HCC (5.2%), UC (3.6%), and gastrointestinal stromal tumor (2.1%). Conclusion In a large (N=154965) NGS database of pts with cancer, inactivating TSC1/2 variants occurred in about 1.7% of pts overall and were frequently identified in commonly occurring cancers. The frequency of TSC1/2 alterations and tumor types were generally consistent between tumor tissue samples and LBs. Consistency between primary vs metastatic samples suggests TSC1/2 alterations may not be acquired, although samples were not longitudinal. Additional research is needed to understand the clinical implications of these observations, but these RW observations suggest TSC1/2 alterations in cancer may be a reasonable therapeutic target. Citation Format: David J. Kwiatkowski, Norma A. Palma, Willis H. Navarro, Gopa Iyer. Real-world (RW) characterization and frequency of TSC1 and/or TSC2 alterations collected from tumor tissue and liquid biopsies from the Tempus genomic database in patients with advanced cancer [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr B003.

Synergy of the microRNA Ratio as a Promising Diagnosis Biomarker for Mucinous Borderline and Malignant Ovarian Tumors.

Epithelial ovarian cancers (EOCs) are a heterogeneous collection of malignancies, each with their own developmental origin, clinical behavior and molecular profile. With less than 5% of EOC cases, mucinous ovarian carcinoma is a rare form with a poor prognosis and a 5-year survival of 11% for advanced stages (III/IV). At the early stages, these malignant forms are clinically difficult to distinguish from borderline (15%) and benign (80%) forms with a better prognosis due to the large size and heterogeneity of mucinous tumors. Improving their diagnosis is therefore a challenge with regard to the risk of under-treating a malignant form or of unnecessarily undertaking radical surgical excision. The involvement of microRNAs (miRNAs) in tumor progression and their potential as biomarkers of diagnosis are becoming increasingly recognized. In this study, the comparison of miRNA microarray expression profiles between malignant and borderline tumor FFPE samples identified 10 down-regulated and 5 up-regulated malignant miRNAs, which were validated by individual RT-qPCR. To overcome normalization issues and to improve the accuracy of the results, a ratio analysis combining paired up-regulated and down-regulated miRNAs was performed. Although 21/50 miRNA expression ratios were significantly different between malignant and borderline tumor samples, any ratio could perfectly discriminate the two groups. However, a combination of 14 pairs of miRNA ratios (double ratio) showed high discriminatory potential, with 100% of accuracy in distinguishing malignant and borderline ovarian tumors, which suggests that miRNAs may hold significant clinical potential as a diagnostic tool. In summary, these ratio miRNA-based signatures may help to improve the precision of histological diagnosis, likely to provide a preoperative diagnosis in order to adapt surgical procedures.

THU484 Distinct Genetic Characteristics In Follicular Thyroid Cancer With Different Metastatic Potentials

Abstract Disclosure: S. Kwak: None. E. Lee: None. Y. Park: None. Although most follicular thyroid cancers (FTCs) show excellent prognosis, more than 5% accompany distant metastasis, often diagnosed unexpectedly in the advanced stage. Genetic or genomic markers have been shown to be useful as a prognostic tool in many cancers, but it is unclear whether genetic markers predicting the metastatic potential of FTCs exist. To investigate, 11 pairs of matched formalin fixed paraffin embedded tissue samples of primary (PTM1) and metastatic tumors (MT) from FTC patients with distant metastasis were analyzed by whole exome sequencing. For validation, 78 unmatched FTC samples including 49 well-differentiated tumors, 15 poorly-differentiated tumors and 14 anaplastic thyroid cancer (ATC) samples were analyzed by targeted sequencing. Among the 11 paired PTM1 and MT tumors, 10 showed concordant driver mutations (NRAS, HRAS, DICER1, EIF1AX), and the TERT promoter mutation was concordantly found in 6 pairs. However, the concordance rate of other somatic mutations in paired samples were found to be only 16.2%. Other mutational characteristics such as mutational signatures and CNVs also differed, suggesting the independent subclonal evolution of the paired MT from the PTM1 after metastasis. Interestingly, the frequency of mutations including driver, TERT promoter, and other mutations was lower in primary tumors without metastases (PTM0). Mutational characteristics of PTM0 also differed from those of PTM1 and MT, suggesting that the mutational or molecular characteristics of PTs at the time of diagnosis may distinguish metastatic potential. In a pooled analysis including published data of follicular patterned tumors, we found that metastatic or aggressive tumors were more likely to have the RAS mutation, which tended to be coupled with TERT, DICER1, EIF1AX, and TP53, gradually from minimally invasive FTC, widely invasive FTC to poorly differentiated thyroid cancer (PDTC) and ATC. Transcriptomic profiles from The Cancer Genome Atlas (TCGA) and our data (SNUH) showed RAS(+)/TERT(+) mutated tumors and RAS(+)/TERT(-) tumors had differences in several intracellular signal pathways. Therefore, when follicular neoplasm is diagnosed, NRAS and HRAS mutation tests followed by TERT promoter mutation tests may help diagnose FTCs with metastatic potential. Other genetic mutations could be candidates for a diagnostic or prognostic marker, and large-scale validation studies are warranted.This study was funded by the Basic Science Research Program through the National Research Foundation of Korea, funded by the Ministry of Science, ICT & Future Planning (NRF-2019R1A2C2084332). Presentation: Thursday, June 15, 2023

Expression data of FOS and JUN genes and FTIR spectra provide diagnosis of thyroid carcinoma

We explore the feasibility of using FOS and JUN gene expression and ATR-FTIR for diagnosis of thyroid cancer. For the study, 38 samples (6 non-neoplastic (NN), 10 papillary thyroid carcinoma (PTC), 7 follicular thyroid carcinoma (FTC), and 15 benign tumors (BT) were subjected to RNA extraction followed by quantitative real time PCR (qRT-PCR) and 30 samples (5 NN, 9 PTC, 5 FTC, and 11 BT) were used for Attenuated Total Reflectance – Fourier Transform Infrared (ATR-FTIR) followed by multivariate analysis. Of the above, 20 samples were used for both gene expression and ATR-FTIR studies. We found FOS and JUN expression in malignant tumor samples to be significantly lower than NN and benign. ATR-FIR after multivariate analysis could identify the difficult to diagnose FTC with 93 % efficiency. Overall, results suggest the diagnostic potential of molecular biology techniques combined with ATR-FTIR spectroscopy in differentiated thyroid carcinomas (PTC and FTC) and BT.

Differences in Cell Death and BMP-2 Expression in Core Biopsy Specimens of Malignant Bone Tumors Given Hydrogen Peroxide Compared to Liquid Nitrogen

Introduction: Malignant bone tumors have a global prevalence of 1% of the overall malignancy case, with a very low 5-year survival rate and high local recurrence. The limb-salvage surgery with bone recycling is widely developed in the management of malignant bone tumors. Hydrogen peroxide and liquid nitrogen are bone recycling mediator agents destroying the malignant tumor cells while preserving healthy bones as much as possible. The purpose of this study was to compare the effectiveness of hydrogen peroxide and liquid nitrogen with the number of tumor cell deaths using the Huvos score and BMP-2 expression in malignant bone tumors. Materials and Method: In vitro experimental research was performed on 30 core needle biopsy samples of bone malignant tumors divided into 2 groups of hydrogen peroxide and liquid nitrogen. The parameters measured are Huvos score and BMP-2 level. Results: Based on this study, the characteristics of malignant bone tumors were dominated by primary tumors (83.9%), in female patients (53.3%) with an average age of 30 years. Liquid nitrogen gave better results than hydrogen peroxide, marked by a lower Huvos score (average 12.13 vs. 18.87, p=0.033) and a higher BMP-2 (average 39.53 ± 26.59 vs. 63.87 ± 27.61, p=0.020). Conclusion: Liquid nitrogen is an effective bone recycling agent for the management of malignant bone tumors.

The amino-acid stress sensing eIF2α kinase GCN2 is a survival biomarker for malignant mesothelioma

BackgroundMalignant mesothelioma is a tumour that is strongly associated with a history of asbestos exposure, and which derives from mesothelial cells that line the serous cavities of the body. The tumour most commonly arises in the pleural cavity, but can also arise in the pericardium, peritoneum, and tunica vaginalis. At present the lesion has a very poor prognosis and is an incurable form of cancer with median survival times of up to 19 months being quoted for some histological subtypes. A large proportion of mesotheliomas have been shown to be arginine auxotrophic, leading to new research for therapeutics which might exploit this potential vulnerability.MethodsWe measured the levels of General Control Non-derepressible 2 (GCN2) protein in malignant mesothelioma tumour samples and determined whether these levels correlate with clinical outcomes.ResultsWe observed that the expression levels of GCN2 correlated with patient survival and was an independent prognostic variable in pairwise comparisons with all available clinical data.ConclusionThese findings suggest that GCN2 levels provides prognostic information and may allow for stratification of care pathways. It may suggest that targeting GCN2 is a viable strategy for mesothelioma therapy development.

CTNNB1 p.D32A (c.95A > C) somatic mutation in stage I grade 1 endometrioid endometrial carcinoma with lung metastasis: a case report

BackgroundMost endometrial cancers are of low histological grade and uterine-confined, with a high 5-year survival rate. However, a small subset of women with low-grade and early-stage endometrioid endometrial cancer experience recurrence and death; thus, a more precise risk-stratification is needed.Case presentationA 29-year-old woman presented with abnormal vaginal bleeding and was diagnosed with FIGO grade 1 endometrioid endometrial carcinoma by curettage. Comprehensive cancer staging including pelvic and para-aortic lymphadenectomy was then performed. Postoperative pathological findings suggested an FIGO grade 1 endometrioid endometrial carcinoma infiltrating the superficial muscle layer. The patient did not receive adjuvant therapy. After 4 years of follow-up, the patient returned to our institution with lung metastasis. She underwent thoracoscopic resection of the affected lobes, followed by six cycles of combined chemotherapy of paclitaxel and carboplatin. Next-generation sequencing showed that the primary and lung metastatic tumors shared 4 mutations: PTEN (p.P248Lfs*8), CTNNB1 (p.D32A), BCOR (p.N1425S) and CBL (p.S439N). Immunohistochemistry revealed nuclear location of β–catenin in the primary and lung metastatic tumor samples, indicating abnormal activation of β–catenin.ConclusionCTNNB1p.D32A (c.95A > C) mutation may be related to lung metastasis in this patient with low-grade early-stage endometrioid endometrial carcinoma.

Single cell RNA sequencing to reveal immune microenvironment in primary and brain metastatic lesion of non-small cell lung cancer.

e21009 Background: Despite immune checkpoint inhibitors (ICIs) have been proven effective in patients with advanced NSCLC, suboptimal therapeutic responses are seen in brain metastatic lesions. The heterogeneous cellular components in the tumor immune microenvironment (TIME) pose a challenge in understanding the low immune response. A comprehensive understanding of the heterogeneous TIME in primary and brain metastatic NSCLC is currently lacking, but crucial for improving ICI therapeutic responses in NSCLC patients with brain metastasis. Therefore, we use single cell RNA sequencing (scRNA-seq) to reveal distinct immune microenvironment in primary and brain metastatic lesion of NSCLC. Methods: Eight primary and nine brain metastatic NSCLC tumor samples were obtained through surgical resection. Immune cells (CD45+) were enriched through fluorescence-activated cell sorting (FACS) and subjected to scRNA-seq using the 10X Genomics platform. Sequencing reads were normalized and analyzed using R/Seurat package. Cellular components of each sample were determined based on known marker genes. Results: In our study, we observed a significant difference in the immune cell composition between primary and brain metastatic tumor samples. We found brain metastatic tumors are marked by an enrichment of TIMP1+ monocytes and PLTP+ tumor-associated macrophages (TAMs), both of which display heightened anti-inflammatory properties. In contrast, primary tumors exhibit an enrichment of FCN1+ monocytes and MARCO+ TAMs. Further analysis has revealed that the enrichment of TIMP1+ monocytes and PLTP+ TAMs is associated with a worse prognosis in patients receiving immunotherapy. Additionally, our examination of the transcriptome profile of CD8+ T cells showed an upregulation of inhibitory and exhausted signatures in brain metastatic tumor samples. Conclusions: Our study provided a comprehensive understanding of the TIME of primary and brain metastatic NSCLC and to gain deeper insights into the immunological mechanisms involved in the response to ICIs. The results of our study could prove valuable for the development of future immunotherapy strategies for NSCLC patients with brain metastasis.

Association of smoking history extracted from electronic health records (EHR) using machine-learning methods and tumor characteristics in patients with lung cancer.

1559 Background: Though smoking is a major risk factor for lung cancer, it has been a challenge to collect patients’ smoking history information accurately from the EH due to data inconsistency and incompleteness. To address these challenges, we utilized a weak supervision methodology to automatically annotate smoking status of patients with lung cancer and correlated it with tumor characteristics. Methods: We analyzed 6,355 patients with lung cancer who underwent tumor profiling with MSK-IMPACT. In total, 14,555 unstructured clinical notes were extracted from EHR at the Memorial Sloan Kettering Cancer Center. The weak supervision methodology used a generative model for intermediate labels that were subsequently tuned by machine-learning classifier to generate the final labels. Clinical notes from a randomly sampled set of 564 patients were manually curated and used for performance assessment. The rest of the patients were split into training and validation datasets used for model training and hyperparameter tuning. Pack years were also extracted from clinical notes using Natural Language Processing. We next conducted multivariate analyses for primary and metastatic tumor samples separately to correlate smoking metrics with tumor characteristics including tumor mutation burden (TMB) and chromosomal instability, as inferred by the fraction of genome altered (FGA) after controlling for age at sequencing, gender, histological subtypes, ancestry, coverage and tumor purity. Results: The weak supervision classifier had almost perfect performance for 2-label classification model (ever smokers and never smokers) with macro F1-score: 97.7%, balanced accuracy: 97.1%, 97.1%, precision:98.4%, 98.4% and recall: 99.5%,94.6% respectively. For 3-label classification model (never smoker, former smoker, and current smoker), the macro F1-score was 79.8%; balanced accuracy: 97.1%, 86.7%, 71.2%, precision: 93.9%, 90.1%, 61.7%, recall: 96.1%, 93.3%, 46.0% respectively. Analyzing genomic data, we observed that smoking status (smoker vs. never smoker) and pack-years were associated with TMB in both primary and metastatic tumor samples (p<2e-16). FGA was marginally associated with smokers compared to never smokers in primary tumor samples (p=0.06). Among smokers diagnosed with lung adenocarcinoma, significantly high FGA in primary tumor samples was observed in males compared to females after adjusting for pack-years and other variables (p= 3.3e-3). Conclusions: We demonstrated high performance of our approach for automated curation of smoking history from EHR. The genomic results confirmed distinct mutational patterns associated with smoking behavior in patients with lung cancer. We are currently exploring multimodal approaches by including chest CT images and “time of quitting” to improve performance of the 3-class model.

Trained dogs can accurately discriminate between scents of saliva samples from dogs with cancer versus healthy controls.

To determine whether dogs can be trained to utilize olfaction to differentiate between saliva samples from dogs with cancer and those from healthy control dogs. Canine patient saliva samples were collected (October 2020 to July 2022) from 139 dogs diagnosed with malignant tumors and from 161 healthy dogs (control samples) for use during training and testing of the dog detection team. Samples from canine patients were collected prior to treatment with radiation therapy or chemotherapy. Six pet dogs (mean ± SD age, 5.4 ± 1.9 years) were trained for odor discrimination between healthy control and malignant tumor samples. Training of the dogs, using a reward-based positive reinforcement method, took place 1 to 3 times per week for a period of 6 months (January to June 2022). After training was complete, a subset of samples not utilized during the training sessions were selected for use during odor discrimination testing of the dog team. The trained dogs could accurately distinguish between samples from cancer patients versus control dogs with a mean sensitivity of 90% and mean specificity of 98%, and with mean positive and negative predictive values of 95%. This study serves as preliminary evidence that dogs can be trained to detect differences in scent between saliva samples from cancer and normal patients. Further studies should expand upon these results with a larger sample, varied tumor types, use of non-cancer diseases as controls, and exploration of this technique in feline patients.

Single-cell RNA sequencing and influences of necroptosis-associated myeloid lineages in immune microenvironment of PDAC.

e16261 Background: Tumor-infiltrating myeloid cells (TIMs) are abundant in tumor stroma and emerging evidence indicate that the presence of these cells influences clinical outcomes in many cancer types. TIMs play important roles in tumor progression and specific subsets may exhibit divergent anti-tumor or tumor-promoting functions. Currently, the molecular characteristics of TIMs in pancreatic ductal adenocarcinoma (PDAC) largely remain unknown. Like other programmed cell death procedures such as ferroptosis and pyroptosis, necroptosis has recently emerged as an important cellular event that modulates tumorigenesis and tumor progression. Combination of necroptosis and TIMs is of great interest to study the molecular mechanism of clinical outcomes for PDAC patients. Methods: This study enrolled 16 PDAC patients. The scRNA-seq data were obtained from a published dataset and each patient included primary tumor, metastatic tumor, primary tumor-adjacent and PBMC samples. All myeloid cells comprising granulocytes, monocytes, macrophages, and dendritic cells from enrolled patients were extracted for further analysis. We performed comprehensive analysis about the characteristics of myeloid lineages, differentially expressed genes, pathway enrichment analysis and novel cell subtype identification. The necroptosis-associated novel TIMs cell sub-clusters were elucidated in detail. Results: To comprehensively analyze the heterogeneous TIMs microenvironment in PDAC patients using scRNA-seq dataset, we firstly obtained consistently upregulated genes among granulocytes, monocytes, macrophages, and dendritic cell types between tumor samples and non-tumor samples. Those uniquely expressed genes represent different functions in necroptosis-associated TIMs cells. In total, we identified 10 necroptosis-associated upregulated genes in PDAC tumors and 5 upregulated genes in tumor-adjacent and peripheral blood samples, respectively. Based on the cells expressed those necroptosis-related genes, we identified three novel cell populations, including GLUL-SQSTM1- RTM, HSP90AA1+HSP90AB1+ mast cells, and JAK3+TLR4+ CD16 monocytes, which were regulators of immunity either tumor-promoters or tumor-suppressors. Clinical outcomes further supported their function roles in the PDAC tumor microenvironment. Conclusions: This study comprehensively studied the heterogeneity of myeloid microenvironment in PDAC by scRNA-Seq. Several necroptosis-associated genes and myeloid cell sub-clusters were identified and they have prognostic clinical values in PDAC. These preliminary results were helpful for further understanding the molecular mechanism of PDAC treatment strategies.

Gene expression comparison between primary aggressive luminal breast cancer and paired axillary and sentinel lymph node metastasis.

e13060 Background: More aggressive and therapy-resistant hormone receptor (HR)-positive breast cancer (hereafter luminal breast cancer, LBC) remains a great challenge. Few studies examine the genomics of axillary lymph node (ALN) metastasis in LBC. The aim of this study was to characterize and compare the mRNA expression patterns of primary breast cancers and paired ALN metastases. Methods: Aggressive pT1-2N2-3M0 LBC patients aged over 18 years who were treated at Shandong Provincial Hospital Affiliated to Shandong First Medical University and had available matched primary and metastatic tumor samples for gene expression analyses were retrospectively recruited. Bulk mRNA expression levels were determined by RNA sequencing (Illumina NovaSeq 6000 system) and calculated using featureCounts. Differential gene expression analysis was performed using DESeq2. All analyses were carried out with differentially expressed genes (DEGs) with an FDR <0.05. Changes of intrinsic subtypes from primary to metastatic disease were assessed. Results: Ten LBC patients with paired tumor samples were included. Compared with the primary tumor, ALN metastatic tumor had 1509 statistically significant DEGs, including 743 upregulated genes and 766 downregulated genes. Notably, there was an upregulation of genes related to immune response-activating signal transduction, T cell receptor signaling pathway, B cell receptor signaling pathway, and cytokine-cytokine receptor interaction in the ALN metastasis. The downregulation of genes were mainly enriched in the PI3K-Akt signaling pathway, ECM-receptor interaction, Wnt signaling pathway, and TGF-beta signaling pathway. Intrinsic subtype distribution in primary tumors was different from those in ALN metastatic tumors, such as Luminal A (30% vs. 50%), Luminal B (10% vs. 20%), HER2-enriched (20% vs. 10%), Basal-like (10% vs. 10%), and Normal-like (30% vs. 10%). These results suggest that LBC is a highly heterogeneous disease. Conclusions: In aggressive pT1-2N2-3M0 LBC, ALN metastases have a distinct gene expression profile and show upregulation of immune-related genes ( CR2, BLK, PAX5, CD19, FCRL1, TNFRSF13B). Genes associated with signaling pathways are involved in the development of breast cancer ( FST, TNN, LAMC3, SFRP2, FGF18), and are downregulated when compared to autologous primary cancer. These pathways are potential targeted pathways in the treatment of breast cancer. LBC cells metastatic to lymph nodes undergo a change to metastasize and survive in the new microenvironment, which may provide insights into the metastatic process. Besides, a switch towards more aggressive intrinsic subtype from primary to advanced disease is also observed.

Acquired DNA damage repairs deficiency-driven immune evolution and involved immune factors of local versus distant metastases in non-small cell lung cancer

ABSTRACT The evolution of immune profile from primary tumors to distant and local metastases in non-small cell lung cancer (NSCLC), as well as the impact of the immune background of primary tumors on metastatic potential, remains unclear. To address this, we performed whole-exome sequencing and immunohistochemistry for 73 paired primary and metastatic tumor samples from 41 NSCLC patients, and analyzed the change of immune profile from primary tumors to metastases and involved genetic factors. We found that distant metastases tended to have a decreased CD8+ T cell level along with an increased chromosomal instability (CIN) compared with primary tumors, which was partially ascribed to acquired DNA damage repair (DDR) deficiency. Distant metastases were characterized by immunosuppression (low CD8+ T cell level) and immune evasion (high PD-L1 level) whereas local metastases (pleura) were immune-competent with high CD8+ T cell, low CD4+ T cell and low PD-L1 level. Primary tumors with high levels of CD4+ T cells were associated with distant metastases rather than local metastases. Analysis of TCGA data and a single-cell RNA-sequencing dataset revealed a decreasing trend of major immune cells, such as CD8+ T cells, and an increasing trend of CD4 T helper cells (Th2 and Th1) in primary tumors with metastases from local to distant sites. Our study indicates that there are differences in the immune evolution between distant and local metastases, and that acquired DDR deficiency contributes to the immunosuppression in distant metastases of NSCLC. Moreover, the immune background of primary tumors may affect their metastatic potential.

Cellular landscape of adrenocortical carcinomas determined by molecular profiling of primary, recurrent and metastatic tumour samples at single-nuclei resolution.

Searchable abstracts of presentations at key conferences in endocrinology ISSN 1470-3947 (print) | ISSN 1479-6848 (online)

HOPE (SOLTI-1903) breast cancer study: real-world, patient-centric, clinical practice study to assess the impact of genomic data on next treatment decision-choice in patients with locally advanced or metastatic breast cancer.

Metastatic breast cancer (mBC) causes nearly all BC-related deaths. Next-generation sequencing (NGS) technologies allow for the application of personalized medicine using targeted therapies that could improve patients' outcomes. However, NGS is not routinely used in the clinical practice and its cost induces access-inequity among patients. We hypothesized that promoting active patient participation in the management of their disease offering access to NGS testing and to the subsequent medical interpretation and recommendations provided by a multidisciplinary molecular advisory board (MAB) could contribute to progressively overcome this challenge. We designed HOPE (SOLTI-1903) breast cancer trial, a study where patients voluntarily lead their inclusion through a digital tool (DT). The main objectives of HOPE study are to empower mBC patients, gather real-world data on the use of molecular information in the management of mBC and to generate evidence to assess the clinical utility for healthcare systems. After self-registration through the DT, the study team validates eligibility criteria and assists patients with mBC in the subsequent steps. Patients get access to the information sheet and sign the informed consent form through an advanced digital signature. Afterwards, they provide the most recent (preferably) metastatic archival tumor sample for DNA-sequencing and a blood sample obtained at the time of disease progression for ctDNA analysis. Paired results are reviewed by the MAB, considering patient's medical history. The MAB provides a further interpretation of molecular results and potential treatment recommendations, including ongoing clinical trials and further (germline) genetic testing. Participants self-document their treatment and disease evolution for the next 2 years. Patients are encouraged to involve their physicians in the study. HOPE also includes a patient empowerment program with educational workshops and videos about mBC and precision medicine in oncology. The primary endpoint of the study was to describe the feasibility of a patient-centric precision oncology program in mBC patients when a comprehensive genomic profile is available to decide on a subsequent line of treatment. www.soltihope.com, identifier NCT04497285.

Abstract 2037: Tracking the emergence of immunotherapy resistance in lung cancer metastases

Abstract Metastatic non-small cell lung cancer (NSCLC) results from a complex evolutionary process in which cancer cells migrate from a primary tumour to a new anatomical site. Immunotherapy is frequently used to treat NSCLC, but drug resistance is frequent (>50%) and metastasis remains the most common cause of death. Recent studies of primary and metastatic tumours suggest that anti-cancer treatments can impact cancer evolutionary dynamics by applying selective pressures that promote the development of treatment resistance in genetically distinct subpopulations of cancer cells, or tumour clones. However, the relationship between metastatic progression and the development of resistance to immunotherapy has not been fully explored. In particular, whether metastases become resistant to immunotherapy as a result of the migration of resistant cells or whether distinct metastatic sites converge in parallel toward an immunotherapy resistance phenotype irrespective of cell migration, is not known. In this work, we investigate the role of metastatic cell migration in the development of immunotherapy resistance using DNA sequencing data from primary and metastatic tumour samples from twelve immunotherapy-treated patients with NSCLC co-recruited to the national TRACERx and PEACE studies. Using computational approaches, we identified distinct tumour clones in primary and metastatic tumours and used these to reconstruct tumour phylogenies for each patient. By evaluating tumour clonal dynamics in relation to treatment, we classified tumour clones as either sensitive or resistant to immunotherapy. We reconstructed the metastatic migration histories of these clones using existing computational methods based on somatic single nucleotide variants. Our results suggest that heterogeneous mechanisms of immune evasion develop in parallel at distinct metastatic sites. The ongoing analysis of the genetic changes that distinguish resistant from sensitive clones, and migrating from non-migrating clones, will provide further detail regarding the impact of metastatic migrations on immunotherapy resistance. Using this approach to map the evolutionary history of metastatic NSCLC, we provide insight into the mechanism by which immunotherapy resistance develops at distinct metastatic sites. Citation Format: Sonya Hessey, Ariana Huebner, Oriol Pich, Abi Bunkum, Wing Liu, David A. Moore, Cristina Naceur-Lombardelli, Selvaraju Veeriah, Sophie Ward, Roberto Salgado, Nicholas McGranahan, Simone Zaccaria, Charles Swanton, Mariam Jamal-Hanjani. Tracking the emergence of immunotherapy resistance in lung cancer metastases [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2037.

Abstract 4277: Measuring proliferation rates of distinct tumour clones using single-cell DNA sequencing

Abstract Background: Tumour proliferation rate is a key phenotypic feature of cancer, with higher rates linked to poorer clinical outcomes. Thus far, proliferation rates have been measured using pathological or experimental techniques on bulk tumour samples. However, while tumours are heterogeneous compositions of distinct clones with varying levels of fitness, measuring the proliferation of individual clones has not been possible to date. We hypothesise that enabling the identification of the most proliferative clones would reveal genomic hallmarks of aggressive clones, or the prediction of their potential phenotype, e.g., metastatic potential. Methods: We have developed SPRINTER (Single-cell Proliferation Rate Inference in Neoplasms Through Evolutionary Routes), a novel computational method to measure proliferation rates in individual tumour clones using single-cell whole-genome DNA sequencing. To assess the accuracy and power of SPRINTER, we have also developed an experimental approach to DNA sequence >18,000 single cells, accurately separated in different DNA-replication phases. We have sequenced and applied SPRINTER to >10,000 non-small cell lung cancer cells from longitudinal and metastatic tumour samples within the TRACERx study and PEACE autopsy programme. We have further analysed published data from >10,000 breast cancer cells. Results: We demonstrate that SPRINTER can accurately identify subpopulations of cells with different proliferation rates using relatively small numbers of cells, in contrast to previous preliminary approaches. While our estimates are concordant with previous bulk experimental studies (5-40%), we importantly have identified clonal heterogeneity in proliferation rates. Using bulk analysis, we have identified patterns of dissemination of tumour clones in non-small cell lung cancer. Integrating this with single-cell data, our results indicate that more widely disseminating tumour clones have higher proliferation rates, suggesting a link between proliferation and dissemination potential. We additionally find that clones are more proliferative in the metastatic versus the primary setting. Furthermore, we have identified high proliferation clones that may have a selective advantage in a breast tumour, and have inferred that they likely arose recently in cancer evolution. Conclusions: We have developed a novel method that enables accurate identification of proliferation rates of individual tumour clones using single-cell DNA sequencing data, allowing the investigation of genomic hallmarks in highly proliferative clones that might lead to higher fitness. Citation Format: Olivia Lucas, Sophie Ward, Rija Zaidi, Mark Hill, Emilia Lim, Haoran Zhai, Abigail Bunkum, Sonya Hessey, Michelle Dietzen, Andrew Rowan, Cristina Naceur-Lombardelli, Nnenna Kanu, Mariam Jamal-Hanjani, Charles Swanton, Simone Zaccaria. Measuring proliferation rates of distinct tumour clones using single-cell DNA sequencing. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4277.

Abstract 6166: TBL1, a multifunctional transcriptional regulator, is highly expressed in osteosarcoma and correlates with activated beta catenin

Abstract Background: Osteosarcoma (OS) is the most common primary bone cancer in children and adolescents. Two-thirds of OS patients are cured through chemotherapy and tumor resection. However, survival rates have plateaued over the past 30 years, especially for patients with relapsed and/or metastatic disease, with overall long-term outcomes of less than 30%. Transducin beta-like protein 1 (TBL1) protects beta-catenin from nuclear degradation and mediates Wnt targeted transcription by forming a TBL1-beta-catenin complex. Overexpression of TBL1 and nuclear beta-catenin are positively correlated to adverse clinicopathological features and poor prognosis for osteosarcoma patients. Functional experiments reveal that down-regulation of TBL1 and nuclear beta-catenin activity inhibit proliferation, migration, and invasion of osteosarcoma cells. The objective of this study was to determine TBL1 levels within a panel of OS patient-derived xenograft (PDX) tumors. Methods: TBL1 and beta-catenin protein levels were measured via western blot and immunohistochemistry in OS samples. We examined protein levels in OS PDX cell lines collected at distinct clinical stages of metastatic OS and OS primary tumor samples. TBL1, total beta-catenin, and activated beta-catenin levels were measured by immunohistochemical (IHC) staining in tissue microarray (TMA) derived from patient derived xenograft (PDX) tumor and clinical samples. Each tumor sample were formalin fixed, paraffin embedded, and spotted in triplicate on the TMA. Results: Beta-catenin and TBL1 protein were detected in OS samples, and we found higher TBL1 protein levels in cells derived from advanced disease stages. Assessment of TBL1 protein was expanded to OS tissue microarray where TBL1, total beta-catenin, and active beta-catenin were independently stained in a TMA. TBL1, total beta-catenin, and activated beta-catenin were present in 100% of OS PDX cores. Finally, a majority of OS cores were shown to have high expression for both TBL1 and active beta-catenin. Conclusion: Elevated TBL1 levels have been associated with metastatic disease and poor survival. We have measured TBL1 and activated beta-catenin protein levels in a variety of osteosarcoma samples and identified a large percentage of osteosarcoma that co-express both proteins. These results indicate that metastatic OS represents a TBL1 enriched patient population and is a candidate for therapeutics that target TBL1. Citation Format: Kimberly R. Holloway, Taku Yamamichi, Bikesh Nirala, Nino Rainusso, Jason T. Yustein, Stephen Horrigan. TBL1, a multifunctional transcriptional regulator, is highly expressed in osteosarcoma and correlates with activated beta catenin [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6166.

Abstract 961: HER2-low expression in patients with advanced or metastatic solid tumors

Abstract Introduction: HER2 protein expression by immunohistochemistry (IHC) has been the main testing method for breast and gastric cancers. Much remains unknown about the frequency of HER2-low (IHC 1+ or 2+) expression and the change of HER2 status from primary to metastatic tumor samples. There is limited data available about the HER2-low frequency in solid tumors beyond breast and gastric/gastroesophageal (GEJ) cancer. Methods: HER2 IHC was performed in 923 patients (pts) with metastatic breast and 51 pts with gastric/GEJ cancers in CLIA-certified labs. We evaluated the HER2 IHC in paired primary and metastatic samples in breast cancer. We also assessed the HER2 status in non-breast, non-gastric/GEJ histologies. In available cases, concurrent HER2 genomic alterations were evaluated. Results: A total of 867 pts with breast cancer had a definitive HER2 IHC score for their metastatic sample: 101 pts (11.6%) were 3+, 171 pts (19.7%) were 2+, 252 pts (29%) were 1+ and 343 (39.6%) had no HER2 expression. The HER2 IHC scores in metastatic samples as compared to primary was in concordance in 54.9% of cases, while HER2 3+ remained unchanged in 87.2%, kappa=0.85 with a 95% CI = (0.79, 0.91). In 48 cases with metastatic gastric/GEJ cancer, HER2 was: 3+ in 2 pts (4.2%), 2+ in 5 pts (10.4%), 1+ in 9 pts (18.8%) and 32 (66.7%) had no HER2 expression. In non-breast and non-gastric/GEJ cancers, HER2-low was found in 631 pts (33.5%) as shown in the table below. A total of 1,552 pts who had HER2 IHC also had genomic testing: 36 had HER2 activating mutations, 48 had HER2 amplification (amp) and 3 had HER2 amp and mutations. 22 of 33 pts with HER2 mutations (without amp) had HER2 IHC expression: 10 pts (30%) were 1+, 10 (30%) were 2+ and 2 (6%) were 3+. Conclusion: Our results show that HER2-low expression is frequently found across tumor types and HER2 activating mutations are enriched in cancers where HER2 is expressed. These findings suggest that a high number of pts with solid tumors with HER2-low expression could potentially benefit from HER2 directed therapies. HER2 expression in solid tumors Cancer Types HER2 expression levels (IHC) Total* 3+(N pts) 3+(%) 2+(N pts) 2+(%) 1+(N pts) 1+(%) 0 (N pts) 0 (%) Appendiceal 0 0 3 6 8 16 39 78 50 Cancer of Unknown Primary 2 4 8 15 15 28 29 54 54 Cholangiocarcinoma 5 4 15 12 35 29 67 55 122 Colorectal 19 3 46 8 118 20 394 68 577 Endometrial 16 13 49 39 10 8 52 41 127 Gallbladder 3 9 8 24 7 21 15 45 33 HNSCC 2 9 1 4 4 17 16 70 23 Salivary 6 13 10 21 14 29 18 38 48 Hepatocellular Carcinoma 0 0 8 25 6 19 18 56 32 Non-small cell lung cancer 6 3 31 15 68 33 100 49 205 Ovarian/Fallopian Tube 4 5 17 20 11 13 53 62 85 Pancreas 1 1 5 7 16 24 45 67 67 Prostate 0 0 0 0 4 13 27 87 31 Renal Cell Carcinoma 0 0 0 0 1 5 18 95 19 Sarcoma 0 0 1 3 1 3 35 95 37 Small Bowel Cancer 2 9 2 9 1 4 18 78 23 Thyroid 0 0 0 0 2 4 54 96 56 Urothelial Carcinoma 38 22 48 27 32 18 57 33 175 *more than 15 samples available; HNSCC=head and neck squamous cell carcinoma Citation Format: Burak Uzunparmak, Cara Haymaker, Gabriela Raso, Chen Zhu, Serena Masciari, Lei Wang, Bryce Kirby, Heather Lin, Aysegul Gorur, Ann-Marie Cimo, Amber Kennon, Qingqing Ding, Andrea Palmieri, Gabriele Elizabeth Urschel, Ying Yuan, Guoxin Feng, Yasmeen Rizvi, Aisha Hussain, Vivek Subbiah, Timothy Anthony Yap, Jordi Rodon-Ahnert, Sarina Anne Piha-Paul, David S. Hong, Funda Meric-Bernstam, Ecaterina E. Dumbrava. HER2-low expression in patients with advanced or metastatic solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 961.

Abstract 213: Phylogenetic reconstruction across 303 metastatic tumor samples using Ultima whole-genome sequencing dramatically increases subclonal resolution

Abstract Comprehensive characterization of tumor evolution is essential for understanding the drivers of metastasis and treatment resistance, which are still largely unknown. Studying evolution and resistance requires large cohorts of patients; for each patient, a comprehensive phylogenetic tree is used to follow clones over time and space. Deep whole-genome sequencing (WGS) enables the creation of more robust phylogenies, since (i) each clone is characterized by far more clone-specific mutations (compared to whole-exome or panel sequencing); and (ii) WGS yields more accurate clonal and sub-clonal allelic copy-number estimates, enabling more precise estimation of the fraction of cancer cells that harbor each mutation. Due to recent development of more affordable WGS by Ultima Genomics, we set out to test our ability to construct phylogenetic trees based on Ultima WGS. Ultima sequencing has demonstrated high sensitivity and specificity for detecting germline polymorphisms, but its performance on somatic mutation calling has not yet been rigorously studied. First, we adjusted our somatic mutation detection pipeline to work with Ultima-generated WGS data and tested its performance using well-studied cell line tumor/normal pairs. We assembled a ground truth set of somatic mutations based on deep sequencing data from multiple platforms for the breast cancer cell line pairs (HCC1954/HCC1954-BL and HCC1143/HCC1143-BL) and the melanoma tumor/normal cell line pair (COLO829/COLO829-BL). We found that mutation detection performance was on-par with other platforms for single nucleotide variations. Next, we identified 32 patients from various cancer types (breast, lung, cholangiocarcinoma, and melanoma) treated with a variety of therapies. We collected between 4 to 14 samples per patient at autopsy (total of 303), and sequenced them with Ultima to an average coverage of 63x (range 48x-91x). We identified mutations using our Ultima-adjusted WGS pipeline and constructed phylogenetic trees using our PhylogicNDT suite of tools. We then compared the results to trees previously constructed from mostly whole-exomes (and a few whole-genomes) sequenced on Illumina. We demonstrate that the trees generated from Ultima WGS data are consistent with the previously generated trees, and proved to be much more detailed. At commercialization, we anticipate that WGS for projects like this will be performed at ~1/6th the cost of current NGS offerings (on par with current WES costs). Therefore, it will enable larger studies and more accurate phylogenetic reconstruction, which will advance the study of tumor evolution and resistance. Citation Format: Julian Hess, Elizabeth Martin, Ilya Soifer, Hila Benjamin, Mendy Miller, Carrie Cibulskis, Brian P. Danysh, Matthew Coole, Stacey Gabriel, Dejan Juric, Doron Lipson, Gad Getz. Phylogenetic reconstruction across 303 metastatic tumor samples using Ultima whole-genome sequencing dramatically increases subclonal resolution [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 213.