Low-pass Whole-genome Sequencing Research Articles

High-dose Furmonertinib in patients with EGFR-mutated non-small cell lung cancer and leptomeningeal metastases: A prospective real-world study

IntroductionLeptomeningeal metastasis (LM) is one of the most severe complications of non-small cell lung cancer (NSCLC). Furmonertinib is a pan-epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) with a high rate of brain penetration and a wide therapeutic window. Here, we evaluated the efficacy and safety of high-dose furmonertinib in patients with EGFR-mutated NSCLC and LM. MethodsThis prospective real-world study included patients with EGFR-mutated NSCLC and LM treated with a high-dose furmonertinib (240 mg once daily) as a monotherapy or in combination with other treatments. The primary endpoint was overall survival (OS), and the secondary endpoints included time to treatment discontinuation (TTD) and clinical response rate. Additional efficacy evaluations included changes in brain magnetic resonance imaging (MRI) by the RANO-LM radiologic criteria. We also introduced next generation sequencing (NGS)-based assays to evaluate genomic and epigenomic features of cell-free DNA (cfDNA) in patients’ cerebrospinal fluid (CSF) samples and to analyze their associations with patient outcomes. ResultsWe enrolled 48 patients, of whom 35 (72.9%) had received third-generation EGFR-TKIs. The median OS was 8.43 months (95%CI, 5.48 to 11.39 months), while the median TTD was 8.27 months (95%CI, 5.40 to 11.14 months), and the clinical response rate was 75%. The LM objective response rate (ORR) and disease control rate (DCR) assessed with RANO-LM radiologic criteria were 50.0% and 92.1%, respectively. The adverse event profiles were consistent with previous reports of furmonertinib. Briefly, 22 (45.8%) had adverse events (AEs) possibly related to furmonertinib and three (6.3%) had a grade 3-elevated aminotransaminase or nausea or leucopenia, leading to dose reduction to 160 mg daily. Furthermore, methylation analysis of cfDNA in CSF showed that there was a significant correlation between the changes of aberrant methylated fragments (AMFs) from lung cancer cells and the response of the patients. Meanwhile, the copy number burden (CNB) scores derived from the low-pass whole genome sequencing (LP-WGS) assay may offer another objective and effective method for the diagnosis and evaluation of treatment efficacy in LM. ConclusionIn the real world, the high-dose furmonertinib-based treatment may potentially have clinical efficacy and tolerable safety in patients of EGFR-mutated NSCLC with LM, even in patients previously treated with other third-generation EGFR-TKIs. Methylation and CNB analysis of cfDNA in CSF may be considered objective indicators for the diagnosis of LM and evaluation of treatment response.

Low-pass whole genome sequencing as a cost-effective alternative to chromosomal microarray analysis for low- and middle-income countries.

Low-pass whole genome sequencing (LP-WGS) has been applied as alternative method to detect copy number variants (CNVs) in the clinical setting. Compared with chromosomal microarray analysis (CMA), the sequencing-based approach provides a similar resolution of CNV detection at a lower cost. In this study, we assessed the efficiency and reliability of LP-WGS as a more affordable alternative to CMA. A total of 1363 patients with unexplained neurodevelopmental delay/intellectual disability, autism spectrum disorders, and/or multiple congenital anomalies were enrolled. Those patients were referred from 15 nonprofit organizations and university centers located in different states in Brazil. The analysis of LP-WGS at 1x coverage (>50kb) revealed a positive testing result in 22% of the cases (304/1363), in which 219 and 85 correspond to pathogenic/likely pathogenic (P/LP) CNVs and variants of uncertain significance (VUS), respectively. The 16% (219/1363) diagnostic yield observed in our cohort is comparable to the 15%-20% reported for CMA in the literature. The use of commercial software, as demonstrated in this study, simplifies the implementation of the test in clinical settings. Particularly for countries like Brazil, where the cost of CMA presents a substantial barrier to most of the population, LP-WGS emerges as a cost-effective alternative for investigating copy number changes in cytogenetics.

METB-05. CSF LIQUID BIOPSY: PAVING THE WAY FOR MINIMALLY INVASIVE BRAIN TUMOR SURVEILLANCE USING LOW-PASS WHOLE GENOME SEQUENCING

Abstract OBJECTIVE We describe a clinical liquid biopsy platform for disease detection and surveillance in pediatric patients with primary central nervous system (CNS) tumors using low-pass whole genome sequencing (LP-WGS) on cell-free DNA (cfDNA) isolated from cerebrospinal fluid (CSF). METHODS CSF samples were collected during routine clinical procedures. cfDNA was extracted and LP-WGS performed using a validated Illumina NextSeq sequencing with a 37 bp paired-end read configuration. Sequencing results were compared to tumor molecular profiling performed by UW-Oncoplex™, a targeted DNA-based next-generation sequencing panel. RESULTS Sixty-one CSF specimens were collected from 29 individual patients with primary brain tumors across 7 institutions, with samples obtained at initial diagnosis or staging (n=17), during treatment (n=29), at therapy completion (n=4), or at disease recurrence (n=11). Patients with medulloblastoma (n=13), ATRT (n=3), high-grade glioma (n=3), low-grade glioma (n=6), ependymoma (n=1), and high-grade neuroepithelial tumor/ embryonal tumors (n=3) were included. Copy number variants consistent with tumor were detected in 24 samples from a median CSF volume of 2.5 mL (range: 1-8 mL) and 54.6 pg/ul of cfDNA (median, range: 1.3-5,220). 31/61 samples yielded enough DNA for LP-WGS while 30 were insufficient due to absence of tumor DNA (true negatives) or low DNA quantity (technical insufficiency). Among the 10 diagnostic samples collected from patients with MRI evidence of disease at initial staging, LP-WGS was positive in 8/10 (80%), compared to 2/10 (20%) positive by cytologic analysis for malignant cells. All 4 staging samples from patients with leptomeningeal metastasis were positive by LP-WGS. At recurrence, LP-WGS was positive for tumor in 6/8 (75%) of MRI-positive patients. CONCLUSIONS LP-WGS of CSF-derived cfDNA can identify clinically relevant somatic copy number variants, offering a minimally invasive liquid biopsy option for ongoing surveillance in pediatric CNS tumors. Further prospective analysis is needed to mitigate interpretation error and refine clinical applications.

BIOM-12. LIQUID BIOPSIES IN CNS TUMORS: A COMBINED APPROACH TO PROFILING CSF

Abstract In precision medicine, molecular profiling of pediatric tumors is crucial for diagnosis, prognosis, and therapeutic decision-making. Traditional biopsy methods, often invasive, have inherent risks and may not capture a tumor’s heterogeneity. Liquid biopsy, however, offers a minimally invasive and comprehensive alternative. We demonstrate its utility in diagnosing and monitoring brain tumors using cell-free DNA (cfDNA) from cerebrospinal fluid (CSF), where conventional cfDNA plasma analysis has shown low sensitivity. Through digital droplet PCR, we analyzed 23 plasma samples from patients with BRAFV600E or H3K27M gliomas, detecting H3K27M in a single case. Conversely, CSF samples revealed 8 positives out of 9 for these mutations, underscoring CSF’s diagnostic value. We advanced our methodology by developing a 21-gene hybrid-capture panel paired with low-pass whole-genome sequencing to pinpoint copy number variations. In our reference sample set, the panel achieved a detection limit with variant allele frequencies as low as 0.5%. With 10ng of cfDNA, we reached 83% sensitivity and 100% specificity, which increased to 100% sensitivity with a 30ng input. Using this assay, we profiled 141 CSF samples from 119 patients, achieving 70% positivity rate in active disease cases in a validation cohort—ranging from 50% in low-grade gliomas to 84% in high-grade gliomas and 75% in medulloblastomas. Intriguingly, of 14 queried tumor samples lacking biopsy confirmation, half showed positive ctDNA results using the panel and/or low-pass whole-genome sequencing. This approach allowed for differential diagnoses in suspected medulloblastoma relapses, tailored therapeutic interventions based on CSF profiles, and effective treatment response monitoring. In summary, our integrated liquid biopsy strategy showcases a powerful, clinically applicable tool that significantly enhances patient care in pediatric neuro-oncology.

Longitudinal cell-free DNA characterization by low-coverage whole-genome sequencing in patients undergoing high-dose radiotherapy

Background and PurposeCurrent radiotherapy guidelines rely heavily on imaging-based monitoring. Liquid biopsy monitoring promises to complement imaging by providing frequent systemic information about the tumor. In particular, cell-free DNA (cfDNA) sequencing offers a tumor-agnostic approach, which lends itself to monitoring heterogeneous cohorts of cancer patients. MethodsWe collected plasma cfDNA from oligometastatic patients (OMD) and head-and-neck cancer patients (SCCHN) at six time points before, during, and after radiotherapy, and compared them to the plasma samples of healthy and polymetastatic volunteers. We performed low-pass (on average 7x) whole-genome sequencing on 93 plasma cfDNA samples and correlated copy number alterations and fragment length distributions to clinical and imaging findings. ResultsWe observed copy number alterations in 4/7 polymetastatic cancer patients, 1/7 OMD and 1/7 SCCHN patients, these patients’ imaging showed progression following radiotherapy. Using unsupervised learning, we identified cancer-specific fragment length features that showed a strong correlation with copy number-based tumor fraction estimates. In 4/4 HPV-positive SCCHN patient samples, we detected viral DNA that enabled the monitoring of very low tumor fraction samples. ConclusionsOur results indicate that an elevated tumor fraction is associated with tumor aggressiveness and systemic tumor spread. This information may be used to adapt treatment strategies. Further, we show that by detecting specific sequences such as viral DNA, the sensitivity of detecting cancer from cell-free DNA sequencing data can be greatly increased.

Glioma monitoring via longitudinal intracranial cerebrospinal fluid cell-free DNA.

2053 Background: Current methods for glioma treatment response assessment are limited. This study aimed to assess the technical and clinical feasibility of molecular profiling using intracranial CSF from patients with gliomas during their disease course. Methods: Adults with gliomas were recruited for longitudinal intracranial CSF collection using 1) Ommaya reservoirs, from which CSF would be sampled on at least two separate occasions, or 2) CSF collection from other clinically indicated CSF access devices, such as ventriculoperitoneal (VP) shunts. Patients were followed until their last visit or present time for ongoing participants. For this initial cfDNA feasibility study, CSF was collected from Ommaya reservoirs in four patients and from an existing VP shunt in one patient. cfDNA was then extracted and analyzed using two sequencing platforms, PredicineCARE for cancer variant profiling and PredicineSCORE for low-pass whole genome sequencing (LP-WGS). This analysis generated genomic aberration profiles and genome-wide copy number burden (CNB) score. Results: Five patients (2 females, 3 males; median age: 40 years, range 32-64 years) underwent longitudinal intracranial CSF collection via Ommaya reservoirs (n=4) or VP shunt (n=1). In total, thirty-five CSF samples were obtained (median volume: 3.80 mL; 0.5-5 mL), with 30 samples (85.7%) yielding sufficient cfDNA for variant profiling and LP-WGS. Our initial findings suggest that tumor fraction increased by 6.28x and 2.87x with radiographic progression in two patients. Tumor fraction decreased by 0.08x and 0.04x in two patients with paired immediate pre-versus-post chemoradiation samples. Despite ongoing pseudoprogression in one patient, tumor fraction decreased to unmeasurable levels from a post-resection baseline of 0.26. Patient-specific tumor-associated variant allelic frequencies (VAFs), including TP53, PTEN, TERT, and CDKN2A/B, decreased within individual patients after resection and chemoradiation. In two patients with isocitrate dehydrogenase (IDH) mutant gliomas, decreasing IDH1 VAF after resection correlated with decreased CSF D-2-hydroxyglutarate (D-2-HG) levels (0.64x and 0.62x, respectively, for the first patient, and 0.01x and 0.07x for the other patient). Both CSF IDH1 VAF and CSF D-2-HG increased in the patient who had radiographic progression (2.56x and 9.21x, respectively). Moreover, CNB decreased below the limit of quantification during treatment and increased above the limit at progression. Conclusions: Longitudinal CSF cfDNA can feasibly be obtained via CSF access devices in patients with gliomas during their disease course. Ongoing studies will evaluate hypotheses generated in this case series regarding how longitudinal CSF cfDNA could be utilized to sensitively detect changes in disease burden. Clinical trial information: NCT04692337 ; NCT04692324 .

Prognostic significance of ctDNA mutation analysis in patients with HER2-positive breast cancer undergoing neoadjuvant chemotherapy.

e12654 Background: Neoadjuvant chemotherapy (NACT) combined with dual anti-HER2 agents followed by surgical resection is the standard treatment for stage II-III HER2- positive breast cancer (BC). Despite the availability of effective therapies such as ado-trastuzumab emtansine (T-DM1) and neratinib, identifying patients at high risk of disease recurrence remains challenging. To evaluate whether circulating tumor DNA (ctDNA) can be used as a biomarker to assess treatment response in patients with HER2-positive EBC treated with neoadjuvant anti-HER2 therapy. ctDNA burden can serve as a useful determinant for escalating or de-escalating (neo)adjuvant strategy in EBC patients. This study aimed to assess the role of ctDNA mutation analysis using low-pass whole-genome sequencing (lpWGS) and whole exome sequencing (WES) in evaluating blood copy-number burden (bCNB), genomic alterations, mutational signatures, blood tumor mutational burden (bTMB), and minimal residual disease (MRD) monitoring. Methods: A total of 46 HER2-positive BC patients who underwent NACT combined with dual anti-HER2 agents and surgical resection were enrolled, including 28 clinical stage II and 17 III patients. Plasma samples were collected at four time points: before NACT, after 3 cycles, at the completion of NACT, and after surgery. lpWGS was utilized to assess bCNB, while WES was performed to evaluate genomic alterations, mutational signatures, and bTMB. Mutations selected from the baseline sample were monitored for MRD. In cases where lpWGS yielded negative results, targeted NGS sequencing was conducted. Somatic mutations are identified by whole exome sequencing across 20,000 genes (with boosted sequencing coverage for 600 cancer-related genes). Between 4-50 personalized somatic mutations or fusions are selected for each patient. This personalized panel together with a fixed MRD core panel is utilized for MRD detection and monitoring. Results: Among the enrolled patients, nodal metastasis was observed in 29 patients and 29 were hormone receptor-positive, and 36 achieved pCR (78.2%). Positive ctDNA was detected 58.8% of patients. Higher nodal stage was associated with ctDNA positivity at baseline. MRD positivity after chemotherapy was observed in 4 patients and 2 patients achieved pCR TP53 was the most frequently altered gene, followed by PIK3CA, MAP3K1, REV3L, MYC, MAP2K4, ERBB3, ATRX, ATM and ALK. This pattern is comparable to that observed in previous studies on BC. Non-PCR patients commonly exhibit MAP3K1 and MAP2K4 alteration is their ctDNA. Conclusions: The identification of MRD, regardless of pCR status, can be valuable for risk stratification and personalized treatment decision-making. Further studies are warranted to validate these findings and explore the clinical implications of ctDNA mutation analysis in HER2-positive BC management.

Evaluation of cost-effective multiple cancer early detection with extremely low coverage whole genome sequencing from plasma.

10538 Background: Cancer continues to be a significant global health concern, with approximately 19.3 million new cases diagnosed and 10.0 million cancer-related deaths reported in 2020. Multi-cancer early detection (MCED) test using the peripheral blood offers a great opportunity to improve the current cancer screening tests with better performance and benefit-to-harm balance. Methods: Here we introduce a novel MCED test called HIFI-PROF, which utilizes extremely low pass whole-genome sequencing (median coverage of 0.6X) to create a multi-dimensional fragmentation signatures model through machine learning approach. HIFI-PROF was developed to detect and localize six types of cancer (colorectal, pancreas, liver, lung, esophagus and gastric cancer). Results: The study included a total of 1,487 participants, comprising 235 colorectal cancers, 227 pancreatic cancers, 150 liver cancers, 125 lung cancers, 107 esophageal cancers, 106 gastric cancers and 537 healthy volunteers. Most of the cancer cases were I/II stage patients (stage I: 38%, stage II: 35%, stage III: 23%, stage IV: 4%). The participants, stratified by age and clinical status, were divided into two sets: a training set (n=1,079) and a validation set (n=408).HIFI-PROF exhibited an excellent performance in the early detection of multiple cancer types, achieving a sensitivity of 87.58% at a specificity of 99.09% and identifying the tissue of origin (TOO) with an accuracy of 81.99%. The sensitivity rates for detecting cancer at different stages were also promising, with a sensitivity of 84.62% for stage I, 85.59% for stage II, 89.61% for stage III and 92.86% for stage IV. Conclusions: These results highlight the high sensitivity and specificity of HIFI-PROF in identifying cancer cases while maintaining a low rate of false-positive results. The use of HIFI-PROF offers a cost-effective MCED test with excellent performance. Such advancements will contribute to improve strategies for multi-cancer early detection, leading to more effective clinical interventions and better outcomes.

Multi-dimensional cell-free DNA-based liquid biopsy for sensitive detection of head and neck squamous cell carcinomas.

e18054 Background: Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer type worldwide. The mortality from HNSCC is closely linked to stage. However, the majority of HNSCC patients were diagnosed at an advanced stage. Therefore, a non-invasive assay for the early detection of HNSCCs is highly desirable for reducing the associated mortality. Methods: We collected a study cohort of 281 participants including 140 pathologically confirmed with head and neck squamous cell carcinomas (HNSCCs) and 141 with non-cancerous conditions. A plasma sample was collected from each participant along with general health screening examinations before any invasive examinations. We performed low-pass whole-genome sequencing with plasma samples and profiled three types of cell-free DNA (cfDNA) characteristics, fragment size pattern (FSP), copy number variation (CNV), and single nucleotide substitution (SNS). We developed an ensemble model with these differential profiles to identify HNSCC patients from non-cancer individuals. We further collected a validation cohort consisting of 90 HNSCC patients and 95 non-cancer individuals to validate the cfDNA-based assay. Results: The liquid biopsy assay that incorporated three types of cfDNA characteristics was able to identify HNSCC patients from non-cancer individuals at an AUROC of 0.989 in the study cohort and 0.984 in the validation cohort. At a specificity of 95.7% (135/141), it achieved a sensitivity of 95.7% (134/140) in the study cohort. With this threshold (0.580), in the validation cohort, 93.7% (89/95) of non-cancer individuals and 93.3% (84/90) of HNSCC patients were correctly identified. Specifically, the sensitivities in early stage (I&II) HNSCCs were 94.5 % (52/55) and 87.5% (28/32) in the study and validation cohort, respectively. Of note, the inferred probabilities of cancer showed consistent trends to increase with pathological stages in both cohorts (P=0.032 and 0.047, respectively). On the contrary, the differentiation grades and sub-locations did not significantly affect the performance of the assay. Conclusions: We introduced a liquid biopsy assay using multiple dimensions of cfDNA characteristics that could accurately identify HNSCCs from non-cancerous conditions. As a cost-effective non-invasive approach, it may be conveniently incorporated into annual general physical examinations to provide supplementary information about the risk of HNSCCs.

A multi-center study for colorectal cancer early detection among patients with high-risk disease using a cell-free fragmentomics assay.

3613 Background: Colorectal cancer (CRC) is the third most prevalent and the second deadliest cancer globally, accounting for an estimated 1,926,000 new cases and 903,800 deaths in the year 2022 alone. It's also established that chronic inflammatory conditions, such as Crohn's disease and ulcerative colitis, elevate the risk of developing CRC. Therefore, a timely CRC diagnosis for patients with colorectal disease can significantly improve their prognosis and therefore reduce cancer-related mortalities. Methods: In this study, we enrolled a total of 167 patients diagnosed with CRC (stages I: 43, II: 55, III: 56, IV: 13) and 217 patients with benign colorectal disease (BCD), from five participating hospitals. All participants underwent colonoscopy procedures, and their CRC or BCD status was pathologically confirmed. The development of the cancer detection model was based on a training cohort of 98 CRC patients and 131 BCD patients from three hospitals, and the evaluation of the model's performance was carried out on an independent test cohort comprising 69 CRC patients and 96 BCD patients from the remaining two hospitals. The predictive model leveraged cell-free fragmentomics profiling, employing low-pass whole-genome sequencing (WGS) on pre-operative plasma samples to identify CRC. Results: The early detection model exhibited remarkable efficacy in differentiating CRC patients from those with BCD, achieving an Area Under the Curve (AUC) of 0.930 in the training cohort and a nearly equivalent AUC of 0.926 in the independent test cohort. By applying the same cutoff determined in the training cohort, our model demonstrated a high sensitivity (91.8% in the training cohort and 91.3% in the test cohort) and a satisfactory specificity (78.6% in the training cohort and 82.3% in the test cohort). Notably, the model showed consistent and robust performance across various CRC characteristics, including tumor location, histology subtypes, grades, and mismatch repair deficiency status in both cohorts. Moreover, our model outperformed traditional CRC screening methods such as the fecal immunochemical test, carcinoembryonic antigen (CEA), and CA19-9, which achieved AUCs of 0.742, 0.720, and 0.543, respectively. Further validation was conducted using an additional cohort of 31 patients with advanced colorectal adenoma (advCRA), where the model achieved an AUC of 0.846 and a 66.7% sensitivity for distinguishing advCRA from BCD. Conclusions: The study introduces a non-invasive, cell-free fragmentomics assay that incorporates low-pass WGS and machine learning algorithms, demonstrating significant clinical potential in accurately distinguishing CRC and BCD patients. Our innovative approach offers a promising alternative to traditional screening methods that can improve early CRC detection, enhance CRC patient outcomes, and reduce mortality rates.

Recurrent neural network for predicting absence of heterozygosity from low pass WGS with ultra-low depth

BackgroundThe absence of heterozygosity (AOH) is a kind of genomic change characterized by a long contiguous region of homozygous alleles in a chromosome, which may cause human genetic disorders. However, no method of low-pass whole genome sequencing (LP-WGS) has been reported for the detection of AOH in a low-pass setting of less than onefold. We developed a method, termed CNVseq-AOH, for predicting the absence of heterozygosity using LP-WGS with ultra-low sequencing data, which overcomes the sparse nature of typical LP-WGS data by combing population-based haplotype information, adjustable sliding windows, and recurrent neural network (RNN). We tested the feasibility of CNVseq-AOH for the detection of AOH in 409 cases (11 AOH regions for model training and 863 AOH regions for validation) from the 1000 Genomes Project (1KGP). AOH detection using CNVseq-AOH was also performed on 6 clinical cases with previously ascertained AOHs by whole exome sequencing (WES).ResultsUsing SNP-based microarray results as reference (AOHs detected by CNVseq-AOH with at least a 50% overlap with the AOHs detected by chromosomal microarray analysis), 409 samples (863 AOH regions) in the 1KGP were used for concordant analysis. For 784 AOHs on autosomes and 79 AOHs on the X chromosome, CNVseq-AOH can predict AOHs with a concordant rate of 96.23% and 59.49% respectively based on the analysis of 0.1-fold LP-WGS data, which is far lower than the current standard in the field. Using 0.1-fold LP-WGS data, CNVseq-AOH revealed 5 additional AOHs (larger than 10 Mb in size) in the 409 samples. We further analyzed AOHs larger than 10 Mb, which is recommended for reporting the possibility of UPD. For the 291 AOH regions larger than 10 Mb, CNVseq-AOH can predict AOHs with a concordant rate of 99.66% with only 0.1-fold LP-WGS data. In the 6 clinical cases, CNVseq-AOH revealed all 15 known AOH regions.ConclusionsHere we reported a method for analyzing LP-WGS data to accurately identify regions of AOH, which possesses great potential to improve genetic testing of AOH.

Abstract PO4-01-04: Clinical significance of Circulating tumor DNA (ctDNA) analysis in predicting pathologic complete remission in HER2-positive Breast Cancer Patients Undergoing Neoadjuvant Chemotherapy

Abstract Background: Neoadjuvant chemotherapy (NACT) combined with dual anti-HER2 agents followed by surgical resection is the standard treatment for stage II-III HER2-positive breast cancer. Despite the availability of effective therapies as an adjuvant treatment such as ado-trastuzumab emtansine (T-DM1) and neratinib, pathologic complete remission (pCR) strongly predicts survival. This study aimed to investigate the role of ctDNA analysis in predicting pCR and clinical outcome using comprehensive ctDNA analysis. Methods: A total of 34 HER2-positive breast cancer patients who underwent NACT combined with dual anti-HER2 agents and surgical resection were enrolled, including 22 clinical stage II and 12 III patients. Plasma samples were collected at four-time points: before neoadjuvant chemotherapy, after 3 cycles, at the completion of neoadjuvant chemotherapy, and after surgery. low-pass whole-genome sequencing (lpWGS) and whole exome sequencing (WES) in evaluating blood copy-number burden (bCNB), genomic alterations, mutational signatures, blood tumor mutational burden (bTMB). Mutations selected from the baseline sample were monitored for Molecular residual disease (MRD) monitoring. lpWGS yielded negative results, targeted NGS sequencing was conducted. Results: Among the enrolled patients (N=34), 21 patients (61.8%) had nodal metastasis and 22 patients (64.7%) were hormone receptor-positive. A total 26 patients (76.5%) achieved pCR. Positive ctDNA was detected in 20 patients (58.8%) at baseline. Higher nodal stage was significantly associated with ctDNA positivity at baseline. After surgery, MRD positivity was observed in 3 patients (8.8%). Among the 26 patents who acheived pCR, only 1 patient (3.8%) showed MRD positivity, while 2 non-pCR patients (25%) demonstrated MRD positivity. Conclusion: The presence of MRD was significantly associated with non-pCR status. The presence of ctDNA at baseline was associated with a higher nodal stage, suggesting its potential as a prognostic marker. Further studies with larger cohorts are warranted to validate these findings and explore the utility of ctDNA analysis for personalized treatment decisions in HER2-positive breast cancer. Citation Format: Byeongju Kang, Soo Jung Lee, Jeeyeon Lee, In Hee Lee, Jee-Young Park, Ho Yong Park, Tiantian Zheng, Lu Tan, Liam Il-Young Chung, Shidong Jia, Pan Du, Young Kwang Chae, Yee Soo Chae. Clinical significance of Circulating tumor DNA (ctDNA) analysis in predicting pathologic complete remission in HER2-positive Breast Cancer Patients Undergoing Neoadjuvant Chemotherapy [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-01-04.

Abstract PO1-07-02: Sensitive tumor detection, accurate quantification, and ER subtyping using low-pass methylome of liquid biopsy samples from patients (pts) with metastatic breast cancer (mBC)

Abstract Background The presence and amount of blood circulating tumor DNA (Tumor Content, TC) is emerging as a clinically relevant factor in many cancers, including BC. In the metastatic setting, a commonly used tumor tissue agnostic approach involves analyzing Copy Number Alterations (CNA) from low-pass whole-genome sequencing (lpWGS) by ichorCNA. However, with a lower limit of detection of 3%, it is unsuitable for low TC applications, such as monitoring disease in response to treatment. Here we report on the use of low-pass whole genome bisulfite sequencing (lpWGBS) coupled with a novel tool, METER (METhylation analyzER), to overcome current limitations of lpWGS, enabling sensitive TC detection, accurate quantification, and ER subtyping from mBC liquid biopsies. Methods METER is a computational tool to analyse TC exploiting BC-specific Differentially Methylated Sites (DMS) and Regions (DMR) in lpWGBS data. To define BC DMSs and DMRs, Rocker-meth was applied to the WGBS profiles of 30 BC tissue (BASIS dataset) and 23 healthy donor cell free DNA (cfDNA) (Fox-Fisher 2021). METER consists of three modules: 1) METER-quant, a DMS-based quantifier of TC; 2) METER-detect, a DMR-based z-score method to classify samples as TC+ (METER+) or TC- (METER-); 3) METER-subtype, to infer ER status via Robust Partial Correlation applied to ER+ vs. ER- BC DMR. We generated lpWGBS data (coverage 0.5-1X) of 135 cfDNA samples from 58 pts (45 ER+, 13 ER-, 55 (95%) up to 3 previous lines of treatment; before starting treatment (T0, n=58), after the first cycle (T1, n=25), and at progression (T2, n=22)) and 30 from healthy individuals as controls. ichorCNA using sensitive parameters (allowing normal fraction up to 0.99) was used to estimate reference TC. A leave-one-out strategy applied to the control samples was used to estimate the false discovery rate (FDR) of METER-detect classification. METER+/- pts were tested for association with OS and PFS using the log-rank test and compared with ichorCNA detection (TC by ichorCNA above or below 3%, ichorCNA+/-). Results TC of 105 cfDNA BC samples by METER-quant was concordant with estimates from the state-of-the-art tool ichorCNA (R >0.90, p< 1e-10). In 22 pts with complete longitudinal data, METER-quant median TC of 0.05 at T0 (IQR 0.02-0.12), 0.02 at T1 (0.01-0.07), and 0.06 at T2 (0.03-0.11) were obtained, with significant difference observed for T0 vs T1 and T2 vs T1 (paired Wilcoxon p< 0.05). Across all time points, METER-detect classified 43% of the n=42 ichorCNA- samples as METER+. The reliability and prognostic performance of METER and ichorCNA detection classification were then compared using clinical outcome data. Based on T0 data, METER+ pts had significantly worse OS than METER- (HR=4.2 CI=2.0-8.9, p< 0.001). This effect was stronger than using ichorCNA (HR=2.3 CI=1.3-4.0, p=0.006). Worse PFS at T0 was observed for METER+ compared with METER- pts (HR=3.7 CI=1.8-7.9, p< 0.001), while no significant association was observed for ichorCNA+ pts (HR=1.4 CI=0.8-2.5, p=0.2). Of note, ichorCNA-/METER+ pts (N=11) at T0 had worse PFS compared with ichorCNA-/METER- (N=11) (HR=3.7 CI 1.4-9.7, p=0.005). Considering the 73 METER+ and 53 samples with METER-quant >5%, METER-subtype showed an accuracy of 0.84 (CI=0.73-0.92) and 0.94 (CI=0.83-0.99) respectively in classifying ER status based on IHC. Discussion lpWGBS enables tumor tissue agnostic analysis and concurrent investigation of complementary molecular features, including CNA and DNA methylation patterns. In a small and heterogeneous cohort, METER showed comparable performance to state-of-the-art tools in terms of TC quantification (ichorCNA) and ER subtyping (IHC), while offering enhanced sensitivity and ensuring a low FDR. METER showed promising prognostic stratification capabilities, particularly for patients with low TC. Validation in an extended prospective cohort is currently ongoing. Citation Format: Marta Paoli, Agostina Nardone, Francesca Galardi, Chiara Biagioni, Dario Romagnoli, Francesca De Luca, Gian Marco Franceschini, Ilenia Migliaccio, Marta Pestrin, Giuseppina Sanna, Emanuela Risi, Luca Livraghi, Erica Moretti, Laura Biganzoli, Luca Malorni, Francesca Demichelis, Matteo Benelli. Sensitive tumor detection, accurate quantification, and ER subtyping using low-pass methylome of liquid biopsy samples from patients (pts) with metastatic breast cancer (mBC) [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO1-07-02.

Abstract 3448: A novel liquid biopsy lung cancer detection method using a coverage and fragment end motif machine learning analysis

Abstract Lung cancer remains the leading cause of cancer death in the United States. One reason for this high mortality is that >40% of cases are diagnosed at a late stage, where the 5-year survival rate is low. Liquid biopsy molecular testing has transformed lung cancer care through the identification of driver mutations when tissue is inaccessible. It also has promise as a tool in early detection and diagnosis, thereby reducing mortality. The existing paradigm for lung cancer screening and diagnosis includes low dose CT (LDCT) scans and tissue biopsy. LDCT scans yield a visualization of lung abnormalities. However, many of these abnormalities are difficult to characterize and next steps are often unclear to physicians. One option is to continue to monitor the patient, which may risk a delay in cancer treatment. Up to 15% of all individuals who are put on a monitoring regime will ultimately be diagnosed with lung cancer. A second option is to biopsy the nodule, which risks unnecessary invasive procedures and expenses to patients. 42-62% of indeterminate nodules detected by LDCT are found to be benign after biopsy. Furthermore, 15% of transthoracic needle biopsies result in complications. Genece Health aims to decrease uncertainty in cancer diagnosis and improve patient outcomes through a minimally invasive liquid biopsy assay that detects the presence of lung cancer. The technology utilizes a proprietary algorithm that leverages machine learning models to detect lung cancer signals in low-pass whole genome sequencing (LP-WGS) of patient cfDNA isolated from plasma. To establish assay performance, 50 clinical lung cancer samples from a commercial biobank, and 100 presumed normal samples were collected in Streck cfDNA BCT Devices. Following cfDNA extraction from plasma, LP-WGS was performed. Sequencing data was then processed to generate fragment end-motif and size (FEMS) and fragment coverage data that was input into Genece Health’s proprietary machine learning classifier to generate a final cancer prediction score. In addition, a limit of detection was determined using three late-stage lung cancer samples, with tumor fractions estimated using whole exome sequencing and ichorCNA scores. Our results show >80% specificity, >90% sensitivity, and a limit of detection of <1% tumor fraction. Through this testing, we demonstrate the assay’s ability to detect lung cancer across various lung cancer stages, including early stage disease, and histologies. Genece has developed a highly sensitive and specific lung cancer detection liquid biopsy assay. We present this assay as a cost-effective approach to discriminate between benign and malignant nodules identified by LDCT. In addition, this assay has the potential to be applied across the cancer care continuum. For example, future work will involve characterizing the assay’s ability to be applied in early cancer screening. Citation Format: Bryan Leatham, Michael Salmans, Molly Smith, Mengchi Wang, Andrew Carson, Kristin Fathe, Byung In Lee. A novel liquid biopsy lung cancer detection method using a coverage and fragment end motif machine learning analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3448.

Abstract 323: Optimization of a liquid biopsy assay to accommodate low cfDNA samples and increase throughput

Abstract Molecular testing that interrogates cell free DNA (cfDNA) from plasma has significant potential value across the cancer care continuum, including early cancer detection, driver mutation identification, and disease recurrence monitoring. To achieve this potential, liquid biopsy assays must be optimized to address multiple challenges. For one, cfDNA quantity in the blood is highly variable, with different amounts present depending on malignancy and stage of progression. This requires optimization to detect low concentrations of circulating tumor DNA (ctDNA) and to accommodate variable tumor fractions. In addition, minimizing assay costs will improve test accessibility and adoption. To address these issues, Genece Health is optimizing their lung cancer liquid biopsy assay to accommodate minimal cfDNA input into library preparation. Furthermore, Genece is testing algorithms that tolerate reduced read depths from low-pass whole genome sequencing (LP-WGS) while maintaining high performance. The Genece lung cancer liquid biopsy assay profiles fragment end-motif and size (FEMS) and fragment coverage signals using LP-WGS (Illumina) and proprietary machine learning algorithms. This study first tested the frequency of end motifs in pooled ctDNA after modifying library preparation (IDT) to reduce reagent volumes from 1X down to 0.5X, 0.25X, and 0.1X of manufacturer specification. A range of cfDNA input from 0.25 ng to 6 ng was tested with each library preparation volume. In addition, LP-WGS read coverage from 5X to 1X was evaluated to assess impact on end motif frequency and assay sensitivity. Performance characterization was further evaluated with a cohort of 50 clinical lung cancer and 100 healthy donor plasma samples tested at 1X and 0.25X library reagent volume. We observed an expected increase in library complexity at higher cfDNA input and library reagent volume. Assay performance was minimally impacted by variable cfDNA input and library reagent volume. However, reducing sequencing read depth is tolerated down to 1.25X depth before increasing the variability of end motif frequencies. These observations are corroborated in the 150 clinical sample cohort. This data shows that by altering standard library preparation conditions and reducing unique read depth, liquid biopsy assays can accommodate a wider range of cfDNA, while concurrently decreasing overall costs. Citation Format: Michael Salmans, Dasom Kim, Andrew Carson, Mengchi Wang, Byung In Lee, Bryan Leatham. Optimization of a liquid biopsy assay to accommodate low cfDNA samples and increase throughput [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 323.

Abstract 967: Clinical applicability of targeted panel sequencing in liquid biopsies for disease diagnosis and monitoring of patients with pediatric solid tumors

Abstract Liquid biopsy assays based on sequence analysis of cell-free DNA (cfDNA) have the potential to transform patient care. We recently validated a liquid biopsy platform, LBSeq4Kids, for patients with retinoblastoma, brain tumors and solid tumors using low pass whole genome sequencing (LP-WGS) of cfDNA from aqueous humor (AH), cerebrospinal fluid (CSF), and plasma (PL) to detect Copy Number Alterations (CNAs) in circulating tumor DNA (ctDNA). Herein, we present phase 2, a high-depth comprehensive Next-Generation Sequencing (NGS) panel using cfDNA that allows us to detect SNV’s, indels and gene fusions in clinically relevant genes for pediatric solid and brain tumors. Seventy-four cfDNA specimens have been analyzed to date, using plasma from 43 patients with solid tumors, CSF from 25 patients with brain tumors and AH from six patients with ocular tumors including retinoblastoma. cfDNA was extracted at diagnosis, during therapy, at remission and/or relapse. Sequencing libraries were constructed with the xGen Prism DNA Library Prep Kit using 5ng of cfDNA. For the detection of recurrent mutations and fusions we designed a custom panel using a hybridization-based capture method (Twist Bioscience) that included full coverage of the exons of 136 genes harboring recurrent mutations and selected introns for five genes with recurrent rearrangement breakpoints including BRAF, EWSR1, SYT, FOXO1 and NTRK1. Hybridized libraries were paired-end sequenced (Illumina NextSeq 500) at an average 1000x collapsed coverage. Mutations identified in cfDNA were compared with those in the primary or metastatic tumors profiled with our OncoKids® NGS panel. A total of 35/43 (81%) plasma specimens were positive for either an SNV, deletion or fusion with the most common events being a somatic missense variant in TP53 (13/35, 37%) or translocation involving EWSR1 (9/35, 26%). A total of 16/25 (64%) CSF specimens were positive for SNVs and fusions, most often involving KIAA1549::BRAF (7/25, 28%) in low grade gliomas. A germline variant in RB1 was detectable in two AH specimens from the left and right eye of a patient with bilateral retinoblastoma while plasma from a patient with metastatic retinoblastoma demonstrated clinically relevant variants in RB1 and DICER1. Serial cfDNA profiling revealed a clinically reported TP53 c.746G>C missense variant in the plasma of a patient with osteosarcoma at multiple timepoints through the disease course, with evidence for relapse prior to standard imaging. These data suggest that clinical implementation of a second version of LBSeq4Kids that combines CNA and mutation analysis allows for diagnosis, residual disease detection and risk stratification for pediatric patients with diverse tumor types, regardless of histologic grade, with a variety of pediatric solid and brain tumors. Citation Format: Eirini Christodoulou, Venkata Yellapantula, Katrina O'Halloran, Liya Xu, Jesse L. Berry, Fariba Navid, Dejerianne Ostrow, Jaclyn A. Biegel. Clinical applicability of targeted panel sequencing in liquid biopsies for disease diagnosis and monitoring of patients with pediatric solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 967.

Abstract 2294: Monitoring glioma treatment response through longitudinal analysis of cell-free DNA in cerebrospinal fluid: Insights from a comprehensive study using next-generation sequencing

Abstract Background: Assessing treatment response in gliomas with high fidelity remains challenging. Cell-free DNA (cfDNA) is increasingly recognized as a liquid biomarker in plasma and cerebrospinal fluid (CSF), showing promise in optimizing therapeutic regimens for various cancers. While CSF has been explored as a source of cfDNA for gliomas, understanding the dynamics of cfDNA abundance and sequencing changes during treatment and recurrence is essential. Leveraging our biobank of longitudinal CSF specimens obtained through CSF access devices, we aimed to determine the potential utility of CSF cfDNA as a monitoring tool for treatment response in glioma patients. Materials and methods: Longitudinal CSF specimens were collected through Ommaya reservoirs (NCT04692337) or ventriculoperitoneal shunts (NCT04692324) from patients with brain tumors, including gliomas. cfDNA was extracted from 1-5 mL of CSF based on available sample volumes and quantified using Qubit. Next-Generation sequencing using PredicineCARE or PredicineSCORE low-pass whole genome sequencing (LPWGS) was performed based on the available cfDNA. Results: CSF collected before versus after resection showed an increase in quantified cfDNA (2.97x, range: 1.58-5.26x), indicating the impact of increased parenchymal disruption and closer contact with CSF. Subsequently, CSF cfDNA decreased during chemoradiation and adjuvant treatment and increased at recurrence, even in patients co-enrolled in immunotherapy trials for IL-7 agonists and pembrolizumab. Notably, copy number burden (CNB) increased at recurrence in three patients with known disease progression and decreased throughout immunotherapy treatment in one patient despite concerns of pseudoprogression. In a patient with a hypermutated glioblastoma, CSF cfDNA revealed over 59 genomic alterations, including MAP2K1, KIT, and PDGFRA. Another patient with a progressing EGFR-amplified GBM showed over 200 new variants in the final sample, with EGFR copy number increasing from 2 to over 30. Conclusion: Analysis of cfDNA throughout a patient’s disease course, encompassing resection and treatment with standard-of-care and experimental therapies, may aid in disease monitoring and treatment response. Further evaluation in a larger patient cohort is needed to determine the sensitivity of changes in mutations, CNB, and cfDNA quantity for glioma disease burden. Citation Format: Riviere-Cazaux Cecile, Xiaoxi Dong, Chao Dai, Wei Mo, Pan Du, Shidong Jia, Terence Burns. Monitoring glioma treatment response through longitudinal analysis of cell-free DNA in cerebrospinal fluid: Insights from a comprehensive study using next-generation sequencing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2294.

Abstract 4796: Improving cancer screening performance for women with dense breast tissue

Abstract Breast cancer is the most common cancer in women in the United States. To reduce mortality, screening by mammography in females is recommended by the United States Preventive Services Task Force, successfully reducing mortality by 15 - 32% in women aged 40 - 69 years. This is partially due to the high compliance rate (60 - 80%) among the candidate population. Mammography increases early-stage disease detection, which is correlated with better outcomes. While mammography has been largely successful, this screening method has limitations. Overall, mammography has false-negative and false-positive rates estimated at 12 - 20% and 7 - 12%, respectively. Consequently, annual mammograms result in 50 - 60% of women having a false-positive result within a 10-year span. Additionally, high breast density, found in approximately 50% of females over the age of 40, increases susceptibility to false positive and negative results. For this reason, in 2023 the FDA required that dense breast tissue status, an independent risk factor for cancer, be disclosed to mammogram recipients. The subsequent diagnostic pathway for women with dense breast tissue is not well defined for the care provider. Alternative imaging options have performance issues or are less accessible. Here we describe a liquid biopsy test to complement mammography in women with dense breast tissue to help resolve ambiguity in results. Blood from early- and late-stage breast cancer patients was prospectively collected in Streck cfDNA BCT Devices (n=100) or retrospectively acquired (n=100); and blood from presumed normal samples was collected (n=200). cfDNA was extracted and converted into libraries for low-pass whole genome sequencing (LP-WGS). Sequenced reads were analyzed to generate fragment end-motif and size (FEMS) and fragment coverage dataframes for classification using Genece Health’s proprietary analysis pipeline and machine-learning (ML) algorithm. 150 breast cancer samples distributed across stages and 150 presumed normal samples were selected to create a training cohort. Performance was tested using a 5-fold cross validation strategy. The best performance was observed using a convolution neural network (CNN) ML model with >80% sensitivity when specificity was set at 85%. The remaining 50 breast cancer samples and 50 presumed normal samples were used as an external test set. The performance in this cohort was comparable to the training cohort, with both specificity and sensitivity >80%. Genece presents an early proof of concept breast cancer screening liquid biopsy assay that uses LP-WGS fragmentomics and ML to detect cancer from a single blood sample. Continuing development with additional cohorts and optimization will yield even greater performance. The Genece liquid biopsy test has the potential to be paired with mammography, especially in women with dense breast tissue, to improve screening sensitivity, specificity, and overall outcomes. Citation Format: Molly Smith, Mengchi Wang, Andrew Carson, Byung In Lee, Michael Salmans, Bryan Leatham. Improving cancer screening performance for women with dense breast tissue [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4796.

Detection of PSMA expression on circulating tumor cells by blood-based liquid biopsy in prostate cancer.

For patients with mCRPC, PSMA-targeted radioligand treatment has significantly improved the clinical outcome. A blood-based liquid biopsy assay for recognizing PSMA protein expression on circulating tumor cells may be beneficial for better informing therapeutic decision-making and identifying the patients most likely to benefit from PSMA-targeted radioligand therapy. Using high-throughput imaging and digital AI pathology algorithms, a four-color immunofluorescence assay has been developed to find PSMA protein expression on CTCs on a glass slide. Cell line cells (LNCaP/PC3s/22Rv1) spiked into healthy donor blood were used to study the precision, specificity, sensitivity, limit of detection, and overall accuracy of the assay. Clinical validation and low-pass whole-genome sequencing were performed in PSMA-PET-positive patients with high-risk mCRPC (N = 24) utilizing 3 mL of blood. The PSMA CTC IF assay achieved analytical specificity, sensitivity, and overall accuracy above 99% with high precision. In the clinical validation, 76% (16/21) of the cases were PSMA positive with CTC heterogeneity, and 88% (21/24) of the patients contained at least one conventional CTC per milliliter of blood. Thirty-six low-pass-sequenced CTCs from 11 individuals with mCRPC frequently exhibited copy number increases in AR and MYC and losses in RB1, PTEN, TP53, and BRCA2 locus. The analytical validation utilizing Epic Sciences' liquid biopsy CTC platform demonstrated the potential to detect PSMA protein expression in CTCs from patients with mCRPC. This assay is positioned as an effective research tool to evaluate PSMA expression, heterogeneity, and therapeutic response in many ongoing clinical studies to target tumors that express PSMA.

Genotype imputation in F2 crosses of inbred lines.

Crosses among inbred lines are a fundamental tool for the discovery of genetic loci associated with phenotypes of interest. In organisms for which large reference panels or SNP chips are not available, imputation from low-pass whole-genome sequencing is an effective method for obtaining genotype data from a large number of individuals. To date, a structured analysis of the conditions required for optimal genotype imputation has not been performed. We report a systematic exploration of the effect of several design variables on imputation performance in F2 crosses of inbred medaka lines using the imputation software STITCH. We determined that, depending on the number of samples, imputation performance reaches a plateau when increasing the per-sample sequencing coverage. We also systematically explored the trade-offs between cost, imputation accuracy, and sample numbers. We developed a computational pipeline to streamline the process, enabling other researchers to perform a similar cost-benefit analysis on their population of interest. The source code for the pipeline is available at https://github.com/birneylab/stitchimpute. While our pipeline has been developed and tested for an F2 population, the software can also be used to analyse populations with a different structure.