Abstract Cytokine release syndrome (CRS) is a common acute toxicity in T-cell therapies, including T-cell engaging bispecific antibodies (T-BiSp). Effective CRS management and prevention is crucial in T-BiSp development. Required hospitalization for 7 of the 9 approved T-BiSp and the need for clinical intervention in severe cases highlight the importance of mitigation strategies to reduce healthcare burden and improve patient outcome. In this review, we discuss the emerging evidence on CRS mitigation, management, and prediction. We cover different strategies for dose optimization, current and emerging (pre)treatment strategies, quantitative pharmacology tools employed during drug development, and biomarkers and predictive factors. Insights are gleaned on step-up dosing and formulation effects on CRS and CRS relationships with cytokine dynamics and drug levels gathered through review of T-BiSp licensing applications and emerging data from conferences and publications.

Abstract Pediatric tumors of the central nervous system (CNS) are a leading cause of cancer-related childhood mortality. Along with imaging, it is essential to molecularly characterize tumor tissue for the diagnosis, treatment, and prognosis of these cancers. Biopsies of deeply-seeded lesions may not be possible or may carry significant risks. On the other hand, liquid biopsy, using cerebral spinal fluid (CSF), provides a minimally invasive alternative to identify and monitor tumor DNA at crucial time points, from diagnosis to follow-up. This allows for molecular diagnosis, identification of therapeutic targets, monitoring response to therapy, and early detection of recurrence or progression. The aim of our study was to assess the practicality and usefulness of liquid biopsies in a diverse group of CNS tumor patients. We used a combination of methods, including a targeted hybridization capture panel and low-pass whole genome sequencing (LP-WGS) based on the sample characteristics. We analyzed a total of 243 CSF samples from patients with CNS tumors including: low-grade glioma (LGG; N=57), high-grade glioma (HGG; N=53), medulloblastoma (MB; N=40), germinoma (CNS-GCT; N=29), and no prior biopsy (N=42). We found somatic mutations in 88% of CSF samples from HGG and 67% from LGG with known tumor DNA variants and sufficient sequencing quality. It's important to note that in LGG patients, samples collected from lumbar puncture were positive only in disseminated cases, while a higher positivity rate was found in cases where CSF was collected from ventricular spaces. Moreover, we observed positive copy number alterations (CNA) in CSF samples collected from HGG, embryonal tumors, and CNS-GCT patients at the time of diagnosis and follow-up with a sensitivity of 64%. Within our sample cohort that had an unknown or query tumor due to no prior biopsy, we were able to detect diagnostic, targetable mutations and/or a positive CNA profile in 30% of cases. This supports the utility of liquid biopsy as a diagnostic tool in in CNS lesions. This tool may also assist in monitoring for relapse in high-grade tumors. For example, in a case study, we were able to detect recurrent MB by indication of a positive CNA profile in keeping with the primary tumor CNV, six years post-surgical resection. In CNS- GCT patients, we found that LP-WGS was highly sensitive at diagnosis as we have observed similar CNV in both CSF and tumor DNA. In summary, the utility of our liquid biopsy platform in the context of CNS cancers is an impactful tool that can aid clinicians in diagnosis, treatment and monitoring disease in a minimally invasive manner. Citation Format: Liana Nobre, Mansuba Rana, Yoshiko Nakano, Ian Burns, Robert Siddaway, Richard Yuditskiy, Cyril Li, Andrew Bondoc, Anthony Liu, Uri Taboori, Cynthia Hawkins. Identifying the clinical utility of CSF as a source of liquid biopsy in pediatric brain tumors [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr B071.

Abstract DNA methylation is a fundamental epigenetic mark that governs gene expression and chromatin organization, thus providing a window into cellular identity and developmental processes. Current datasets typically include only a fraction of methylation sites and are often based either on cell lines that underwent massive changes in culture or on tissues containing unspecified mixtures of cells. Here we describe a human methylome atlas, based on deep whole-genome bisulfite sequencing, allowing fragment-level analysis across thousands of unique markers for 39 cell types sorted from 205 healthy tissue samples. Replicates of the same cell type are more than 99.5% identical, demonstrating the robustness of cell identity programmes to environmental perturbation. Unsupervised clustering of the atlas recapitulates key elements of tissue ontogeny and identifies methylation patterns retained since embryonic development. Loci uniquely unmethylated in an individual cell type often reside in transcriptional enhancers and contain DNA binding sites for tissue-specific transcriptional regulators. Uniquely hypermethylated loci are rare and are enriched for CpG islands, Polycomb targets and CTCF binding sites, suggesting a new role in shaping cell-type-specific chromatin looping. The atlas provides an essential resource for study of gene regulation and disease-associated genetic variants, and a wealth of potential tissue-specific biomarkers for use in liquid biopsies. Citation Format: Tommy Kaplan. A DNA methylation atlas of normal human cell types [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr IA025.

Abstract Liquid biopsy is an emerging tool in oncology, providing minimally invasive detection of tumor- specific DNA in body fluids, known as circulating tumor DNA (ctDNA). Colorectal cancer (CRC) has shown a high rate of shedding of ctDNA, making it a promising prognostic and possibly predictive biomarker in the personalized management of patients with CRC. Therefore, the half-life of ctDNA in circulation is estimated to be between 16 minutes and several hours, indicating a need for rapid processing. This study aimed to explore the concordance between the mutational profiles in tumor tissue and ctDNA from plasma obtained using EDTA tubes as well as the correlation between ctDNA and disease stage, to evaluate the potential of implementing liquid biopsy testing in CRC cases, following biobank standards. In this retrospective cohort, 58 patients with CRC stage T1 to T4 were enrolled at the Research Institute McGill University Health Centre (RI-MUHC). Tumor tissue was obtained through surgical resection or biopsies of formalin-fixed paraffin-embedded tissue, and blood samples were drawn pre-surgery using EDTA tubes. Cell-free DNA (cfDNA) was extracted from 1-2mL plasma samples using QIAamp Circulating Nucleic Acids. Next-Generation Sequencing using a 52-gene (AmpliSeq for Illumina Focus Panel) was performed to identify the tumor mutations and subsequently droplet digital PCR (ddPCR) was performed to identify the plasma mutations. Among the 58 patients, the mean age was 62 years (range 29-86), and 29 were women. Forty- one patients had left-sided tumor. At least 1 tumor-specific mutation was detected in all tissue tumor samples. Overall, the most frequently identified mutations in this cohort were: BRAF V600E (22.4%), KRAS G12D (12.1%), KRAS G12V and KRAS G13D (8.6%). Cell-free DNA was obtained from 32 patients and showed a mean concentration of 975 ng/mL (range 400- 1,995) for stage T1, 584 ng/mL (range 228-1,300) for stage T2, 1,127 ng/mL (range 249-4,160) for stage T3 and 2,083 ng/mL (range 860-3,680) for stage T4. ddPCR revealed mutation concordance between ctDNA and matched tumor tissue of 28.1%. When tumors were stratified by stage, the concordance was 0% (n=0/3) in patients with T1 tumors, 16.7% (n=1/6) in patients with T2 tumors, 25% (n=5/20) in patients with T3 tumors and 100% (3/3) in patients with T4 tumors. This study showed that ctDNA detection of mutations has excellent concordance in stage T4 tumors, highlighting the potential of ctDNA in advanced stage disease. However, in T1-3 tumor stages the concordance was low, showing the importance of preanalytical steps such as the use of cell-free DNA preservative tubes for sensitive and accurate detection of ctDNA in patients with earlier stage disease. Citation Format: Monyse de Nobrega, Kyle Dickinson, Saba Alsaddah, Alexandra Bartolomucci, Tadhg Ferrier, Anne Mahalia Olivier Mahalia Olivier, Mina Farag, Gertuda Evaristo, Sophie Camilleri-Broet, Andrea Gomez, Thupten Tsering, Lawrence Lee, Pierre Olivier Fiset, Julia V. Burnier. Circulating tumor DNA profiling in colorectal cancer to detect guideline-based targeted mutations at a Quebec health care center [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr B011.

Abstract Pancreatic ductal adenocarcinoma (PDAC) is an aggressive cancer with a poor prognosis, largely due to late diagnosis and limited treatment options. The immunosuppressive tumor microenvironment (TME) further exacerbates the situation, hindering the effectiveness of traditional therapies. This study investigates the potential of IL-10R2/IL-22R1 expressing immune cells as novel biomarkers for early PDAC detection and recurrence monitoring, offering a potential avenue for improved patient outcomes. The research revealed a significant increase in the number of IL-10R2/IL-22R1 expressing immune cells in the peripheral blood of PDAC patients compared to healthy controls. These cells were found to infiltrate PDAC tissue and exhibit immunosuppressive functions, contributing to the TME's suppression of the immune response. Further investigation using a mouse model of PDAC demonstrated that early tumor progression was accompanied by a rise in IL-10R2/IL-22R1 expressing immune cells in the blood, highlighting their potential role in tumor development. In post-pancreatectomy patients, the study observed that elevated levels of IL-10R2+ immune cells were associated with earlier recurrence compared to the currently employed biomarker, CA19-9. This finding suggests that IL-10R2/IL-22R1 expressing immune cells could serve as valuable tools for predicting PDAC recurrence and potentially improving treatment outcomes. This research provides compelling evidence for the potential of IL-10R2/IL-22R1 expressing immune cells as promising biomarkers for early PDAC diagnosis and recurrence monitoring. However, further research is crucial to comprehensively understand their generation, function, and clinical utility. Larger-scale studies are needed to validate these findings and translate them into practical clinical applications for improved patient care. Citation Format: Hyung Keun Lee, So Young Kim, Soo-Hyun Chung, Sung Ill Jang, Bongkun Choi, Ji-Eun Kim, Dohee Yoon, Areum Yeo, Hyun Goo Kang, Jusung Lee, Yoon Ha Choi, Joon Seong Park, Yoolim Sung, Jong Kyoung Kim, Eun-Ju Chang, Dong Ki Lee. Tumour-associated myeloid cells expressing IL-10R2/IL-22R1 as a potential biomarker for diagnosis and recurrence of pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr A058.

Abstract Osteosarcoma (OS) is the most prevalent bone cancer in children, characterized by aggressive primary tumors and high metastatic potential. Current treatment protocols involve a combination of neoadjuvant chemotherapy, surgery, and adjuvant chemotherapy, tailored based on histological grade and metastatic risk. Identifying metastatic lesions through routine imaging significantly impacts treatment outcomes, highlighting the need for innovative diagnostic tools for timely monitoring. Noninvasive in vitro diagnostics, analyzing circulating tumor components like tumor-derived extracellular vesicles (EVs), would offer promise for monitoring treatment responses in pediatric sarcomas. Matrix metalloproteinases (MMPs) play a crucial role in tumor progression by degradation of the extracellular matrix; thus, recent research has focused on understanding their functional roles by profiling their proteolytic activities. Tumor EVs are crucial in transporting MMPs to facilitate invasion and metastasis, making them suitable surrogate for noninvasive profiling of MMP activities. Herein, we introduce an innovative OS EV MMP Activity Assay to assess MMP activities in OS patients noninvasively. This assay evaluates six different combinations of surface markers on OS EVs and OS-associated MMPs, generating a unique profile for each patient. Using a logistic regression model, an OS EV MMP Activity Score is generated from the most effective combinations. This Score has shown excellent performance in distinguishing between metastatic and localized OS. In a longitudinal study of six OS patients undergoing treatment, the Score strongly correlated with treatment response as measured by routine imaging. The assay represents a significant advancement in harnessing tumor EVs for diagnostics, providing a noninvasive, quantitative, and reliable approach to assess MMP activities in solid tumors. Citation Format: Junseok Lee, You-Ren Ji, Hyoyong Kim, Zhao Chen, Ryan Zhang, Diego Banuet, Yusef Mathkour, Kenny Vo, Steven Jonas, Yazhen Zhu, Hsian-Rong Tseng. Precise monitoring of osteosarcoma patients by noninvasive assessment of matrix metalloproteinases activities in plasma extracellular vesicles [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr A028.

Abstract Background. Extracellular vesicles (EV) are emerging as new biomarkers for cancer diagnostics and therapeutic monitoring1. For example, in patients that receive PD1/PD-L1 immune checkpoint inhibitors (a key pillar of cancer immunotherapy), circulating EV-derived PD-L1 is gaining attention as a non-invasive biomarker for therapeutic efficacy. Standard methodologies to measure EV-derived PD-L1 requires ultracentrifugation to first separate EVs and then quantify those that express PD-L1 using nano flow cytometry. These are tedious processes that require trained technologists in centralized facilities, which is a key barrier to frequently monitoring EV-based biomarkers. Yet, frequent monitoring is crucial to quickly identify and stop an ineffective therapy while more fruitful options are still available. Method. To bridge the gap, we develop a proof-of-concept digital microfluidic (DMF) device that can separate EVs directly from biofluids and quantify PD-L1+ EVs automatedly. Briefly, DMF device supports automatedly droplet handling, including droplet transport, splitting and mixing, based on electrowetting principles2. Immunomagnetic beads were used to capture EVs from biofluids, and EVs were then detected by electrochemical sensors inserted into the device. These tips of the electrochemical sensors were modified with gold nanoparticle self-assembled layers to enhance target binding, and CD9 antibodies were immobilized on the tips to bind to EVs. We then used anti-PD-L1 as secondary antibodies for target detection. Differential Pulse Voltammetry (DPV)3 were performed to obtain the results. Results. We used the DMF device to extract 109- 1011 #/mL EVs secreted by a human breast cancer cell line (MDA-MB-231). Briefly, 10 mL of cell culture media was loaded into the DMF device, and the sample was mixed and incubated with a 10 mL droplet of immunomagnetic beads to capture EVs. After 5 min, EVs were eluted in 20 mL of elution buffer, and the beads were discarded. We estimated the number of EVs using Nanoparticle Tracking Analysis; the number of EVs extracted on-chip was comparable with those that were extracted using standard in-tube method. We also verified the purity of the EVs extracted on-chip using Western Blot. Our results showed that the electrochemical sensors can detect as low as 104 #/mL PD-L1+ EVs, with a linear range of 104 – 107 #/mL (R2=0.9701). The entire process can be completed within 1 hour. Conclusion. We integrated automated EV extraction and their surface marker detection in a single DMF device. Our next step is to extract and detect EV-based biomarkers from plasma samples. This platform holds promise for the regular monitoring of EV-based biomarkers that are indicators of therapeutic responses in patients that receive cancer immunotherapies.

Abstract Introduction: Head and neck cancer (HNC) is one of the most common cancer types globally without an accessible rapid screening method. Conventional screening occurs at the clinician’s office, where grading and staging of the tumors are physically performed, and the clinician’s preliminary diagnosis is usually confirmed through laborious procedures of histology and CT/PET/MR scans. Patient’s quality of life and expected outcomes can be improved through early-stage screening, yet there is no existing technology on the market that tackles this dire need. As such, we propose a rapid liquid biopsy diagnostic platform empowered by deep learning algorithms that makes early-stage non-invasive screening possible. The platform utilizes the complementary vibrational spectral biomarkers information from Fourier Transform Infrared Spectroscopy (FTIR) and Raman spectroscopy. We coupled the techniques with two biofluids: saliva and plasma, which offered information on circulating and localized metabolites. Materials and Methods: The patient’s biofluid is first drop-casted on top of a substrate and air-dried at room temperature. Using an FTIR microscope or a Raman microscope, 25 spot measurements are made in a 5 by 5 grid pattern to produce a comprehensive scan of the sample. The measured spectra are subsequently fed into the deep learning framework to perform classification. The deep machine learning framework is composed of spectrum and subject-level models. Specifically, the spectrum-level model employs cascading convolutional layers to identify the prominent features within the spectrum. This is followed by a sequence of linear layers that further extract features from the convolutional layers and carry out binary classification. For each patient, preliminary classification results of the spectral level model are then evaluated on a per- modality basis across our four modalities (FTIR/Raman measurements made on saliva/plasma) where the patient is classified as positive if any of the modality models deem it so. This framework is applied to a cohort of 96 patients, with 36 non-cancerous patients, 28 early-stage patients, and 32 late-stage patients. Results and Discussion: Our model optimized for general cancer detection has a sensitivity of 91.7%, specificity of 77.8%, and accuracy of 86.5%. Late- stage cancer has proven to be less challenging to detect since our model optimized for late-stage cancer diagnosis has a sensitivity of 93.8%, a specificity of 83.3%, and an accuracy of 88.2%. Our best early-stage model has a sensitivity of 89.3%, a specificity of 69.4%, and an accuracy of 78.1%. Conclusions: Our proposed framework has demonstrated promising capability as the preliminary screening method for high-risk patients (e.g. tobacco users) while providing further interpretability of the deep learning model that is often characterized as a “black box”. The modular nature of our model structure offers substantial flexibility, enabling effective performance even when specific biofluids or spectroscopy modalities are unavailable. Citation Format: Kwan Lun Chiu, Antonio Guillen-Perez, Rebecca Mayer, Wesley Viner, Hanna Koster, Matthew Benson, Jeremy Rowe, Maria Navas-Moreno, Andrew Birkeland, Maria-Dolores Cano, J. Sebastián Gomez-Diaz, Randy Carney. Biofluid-Ensemble Analysis through Multi-modal Spectroscopy (BEAMS): A deep learning architecture for rapid early-stage liquid biopsy cancer diagnostics [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr A010.

Abstract Non-invasive monitoring of ctDNA following curative intent IO therapy has been used to build an understanding of the dynamics of clinical response. Defining clinical response for patients identified with stable disease has always been difficult with imaging. To improve the resolution of ctDNA dynamics, we examined the plasma from a set of exceptional responders to combination IO therapy. These patients did not progress by imaging, however, the measurement of ctDNA showed that there was a persistent level of disease burden throughout the time on study. By evaluating these patients, we gained a comprehensive understanding of the response dynamics including the depth of Molecular Response (MR), the level of molecular persistence and the timing of molecular progression. Current NGS based methods for evaluating ctDNA have been shown to have considerable variability particularly at lower Tumor Fractions (TF). MR dynamics of ctDNA utilize a decrease of 50% TF, which may be related to the accuracy and resolution of assay rather than being a biologically relevant limit. We describe a comprehensive monitoring ctDNA platform using a novel digital PCR assay strategy that achieves a dynamic range not achievable by any current assay. METHODS: Plasma samples from the exceptional responders were evaluated using a novel tumor informed ctDNA platform. Personalized panels, targeting 3-5 variants, were selected from patient tumor data. In total, 225 unique timepoints were extracted (avg 25 timepoints per patient). A novel PCR primer algorithm was designed which showed improved amplification for the individual reactions and optimized multiplexing of the assays. Utilizing machine learning methods, the aggregated data analysis calculates quantitative TF and an accurate percent change between each timepoint. RESULTS: By examining the entire treatment cycle of these patients, we found that the patient timeline separated into molecular response, molecular persistence, and molecular progression. The maximum decrease in TF (molecular response) was found by C4D1 with the largest decrease occurring at either C1D1 or C2D1. Molecular persistence was found in one patient to continue for 55 cycles of treatment (1381 days on trial). Molecular progression occurred in 2 of the 3 exceptional responders. One patient had progression at cycle 12, while the second occurred at cycle 43. Progression occurred very rapidly with the TF doubling from baseline value in 28 days. While patient 1 did see a decrease in the TF for 3 cycles, progression was observed, prompting the end of treatment. To our knowledge this is the first study to evaluate the complete treatment cycle of exceptional responders to IO therapy. The ddPCR assay provided a unique and expansive window into the ctDNA dynamics of treatment suggesting that longitudinal monitoring provides valuable insights. Future studies utilizing this assay will allow for better understanding of the complex tumor dynamics during treatment. Citation Format: Daniel Stetson. Novel digital PCR ctDNA monitoring assay evaluating exceptional responders to combination immune-oncology (IO) therapy [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr PR015.

Abstract Background: Gastroesophageal cancers commonly spread to the peritoneum. Peritoneal disease significantly alters patient prognosis and treatment-intent. Currently, peritoneal lavage cytology (PLC) is the method of choice for detecting peritoneal micro-metastasis within the peritoneum. Despite its pervasive use, PLC has a variable sensitivity of 10% - 80% in detecting peritoneal micro-metastasis in gastroesophageal cancers. Peritoneal tumor DNA (ptDNA) is tumor-derived DNA detectable in peritoneal lavage fluid. ptDNA-positivity may indicate peritoneal micro- metastasis and may be more sensitive than PLC in staging the peritoneum and predicting disease- free survival (DFS) and overall survival (OS). We aimed to develop and test a tumor-informed platform for ptDNA detection and validate its sensitivity and specificity against PLC. Methods: For this pilot study, peritoneal lavage fluid were collected at staging laparoscopy from 15 patients with gastroesophageal cancer. Cytology and peritoneal metastases (where clinically detectable) were confirmed by histopathology. A tumor-informed ptDNA testing was performed via whole-genome sequencing of tumor tissue and buffy coat to identify somatic variants for each patients, followed by tracking of up to 96 variants in each patient’s peritoneal fluid. Results: 5 out of 15 patients were either cytology positive or had macroscopic peritoneal disease and ptDNA was detectable in all 5 cases. Additionally, out of the 10 cytology-negative patients without macroscopic peritoneal disease, ptDNA positivity was detected in 6 patients prior to commencing neoadjuvant chemotherapy or surgery. Two of these 6 patients demonstrated radiological evidence of recurrence at 15- and 18-months follow-up respectively. Interestingly, one patient with macroscopic peritoneal disease tested negative for cytology. In this case, peritoneal lavage was carried out to the upper two abdominal quadrants only, and the pelvic deposit of disease was detected after the lavage had been completed. Despite a negative cytology result, ptDNA was detected using the same fluid sample, suggesting a higher sensitivity compared to PLC. Conclusions: This pilot study demonstrates that a tumor-informed ptDNA detection platform is feasible and potentially more sensitive than PLC in gastroesophageal cancer. To further validate the clinical utility of ptDNA we are proceeding to a multi-center prospective cohort study. Citation Format: Zexi Allan, Yuxuan Wang, Jeanne Tie, Niall Tebbutt, Nicholas Clemons, Nickolas Papadopoulos, Kenneth Kinzler, Bert Vogelstein, David SH Liu. Peritoneal tumor DNA as a prognostic biomarker in locally advanced gastroesophageal cancer: A pilot prospective cohort study [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr A049.