Search

Backgrounds Low-pass whole-genome sequencing (LP-WGS)-based circulating tumor DNA (ctDNA) analysis is a versatile tool for somatic copy number aberration (CNA) detection, and this study aims to explore its clinical implication in breast cancer. Methods We analyzed LP-WGS ctDNA data from 207 metastatic breast cancer (MBC) patients to explore prognostic value of ctDNA CNA burden, and validated it in 465 stage II-III triple-negative breast cancer (TNBC) patients who received neoadjuvant chemotherapy in phase III PEARLY trial (NCT02441933). The clinical implication of locus-level LP-WGS ctDNA profiling was further evaluated. Results We found that a high baseline ctDNA CNA burden predicts poor overall survival and progression-free survival of MBC patients. The post-hoc analysis of PEARLY trial showed that a high baseline ctDNA CNA burden predicted poor disease-free survival independent from pathologic complete response (pCR), validating its robust prognostic significance. The 24-month disease-free survival rate was 96.9% and 55.9% in pCR (+)/Low I-score and non-pCR/High I-score patients, respectively. The locus-level ctDNA CNA profile classified MBC patients into five molecular clusters and revealed targetable oncogenic CNAs. LP-WGS ctDNA and in vitro analysis identified the BCL6 amplification as a resistance factor for CDK4/6 inhibitors. We estimated ctDNA-based homologous recombination deficiency (HRD) status of patients by shallowHRD algorithm, which was highest in the TNBC and correlated with platinum-based chemotherapy response. Conclusions These results demonstrate LP-WGS ctDNA CNA analysis as an essential tool for prognosis prediction and molecular profiling. Particularly, ctDNA CNA burden can serve as a useful determinant for escalating or de-escalating (neo)adjuvant strategy in TNBC patients.

Background Previous studies proposed low-pass whole genome sequencing (LP-WGS)-based circulating tumor DNA (ctDNA) analysis as a versatile tool for genomic profiling and therapeutic monitoring of cancer patients. Here we demonstrate LP-WGS ctDNA genomic profiles and its clinical significance in metastatic breast cancer patients. Patients and methods This prospective exploratory study enrolled 207 treatment-naïve metastatic breast cancer patients from Feb 2017 to September 2020 in Yonsei Cancer Center. The median follow-up duration of patients was 35 months. The baseline (n=207) and post-progression (n=48) plasma samples were prospectively collected on first-line systemic therapy, and LP-WGS was employed for ctDNA somatic copy number alteration (CNA) analysis. The CNA burden of ctDNA was scored by “I-score” method, which was developed to measure genome-wide chromosomal instabilities, to be matched with therapy response. The unsupervised molecular clustering and homologous recombination deficiency (HRD) estimation by shallowHRD algorithm were performed using locus-level CNA profiles with 1 mega base pair resolution. Results The baseline I-score ctDNA CNA burden was highest in triple-negative breast cancer (TNBC) patients among subtypes, and the patients were dichotomized by median I-score level 5.54 (range 2.55 to 12.98). The high baseline ctDNA I-score was independently associated with poor overall survival (hazard ratio [HR] = 3.98, p < 0.001) with adjustment of tumor subtype, visceral metastasis, and disease status (de novo stage IV versus recurrent). The progression-free survival (PFS) on endocrine plus CDK4/6 inhibitors (HR = 2.75, p = 0.005), anti-HER2 therapy (HR = 2.52, p = 0.032), and cytotoxic chemotherapy (HR = 2.33, p = 0.012) was also shorter in high baseline I-score patients than in low I-score patients. The locus-level CNA profile was analyzed in high I-score patients (n=103), and the patients were classified into five molecular clusters with distinct overall survival by unsupervised k-means clustering of CNA profile: basal-like, EGFR-high basal-like, CCND1-high, luminal, and HER2-enriched clusters. Patients with BCL6 (p = 0.009) and PIK3CA amplification (p < 0.001) on baseline ctDNA showed significantly shorter PFS on CDK4/6 inhibitor treatment. The matched baseline and post-progression ctDNA analysis found emergence of FGFR1 amplification and MYC amplification after CDK4/6 inhibitor treatment (n=1, each). The ctDNA shallowHRD score was highest in TNBC patients among subtypes, and TNBC patients with high shallowHRD score (≥10) showed high response rate on (58.3% versus 28.6%) on platinum-based chemotherapy. Conclusion LP WGS-based ctDNA analysis provides a robust tool for non-invasive genomic clustering, therapy response prediction, and HRD estimation in metastatic breast cancer patients. All patients (n=207)Low I-score (n=104)High I-score (n=103)N (%)N (%)N (%)Age, Median (Interquartile range)54 (46-62)53 (47-60)54(44-62)GenderFemale205 (99)102 (98.1)103Male2 (1)2 (1.9)0SubtypeHR+ HER2-106 (51.2)61 (58.7)45 (43.7)HR- HER2+33 (15.9)14 (13.5)19 (18.4)HR+ HER2+22 (10.6)11 (10.6)11 (10.7)HR- HER2- (TNBC)46 (22.2)18 (17.3)28 (27.2)Disease statusDe novo stage IV74 (35.7)31 (29.8)43 (41.7)Recurrent133 (64.3)73 (70.2)60 (58.3)Primary therapyEndocrine + CDK 4/6 inhibitor97 (46.9)55 (52.9)42 (40.8)Anti-HER2 based therapy54 (26.1)24 (23.1)30 (29.1)Chemotherapy45 (21.7)16 (15.4)29 (28.2)Others11 (5.3)9 (8.7)2 (1.9)Visceral metastasisYes142 (68.6)60 (57.7)82 (79.6)No65 (31.4)44 (42.3)21 (20.4)Metastasis SitesLung89 (43)43 (41.3)46 (44.7)Brain19 (9.2)4 (3.8)15 (14.6)Liver59 (28.5)13 (12.5)46 (44.7)Bone120 (58)47 (45.2)73 (70.9)Lymph node90 (43.7)32 (30.8)58 (56.9)Pleura33 (15.9)17 (16.3)16 (15.5) Citation Format: Joohyuk Sohn, Min Hwan Kim, Jin Mo Ahn, Won-Ji Ryu, Seul-Gi Kim, Jee Hung Kim, Tae Yeong Kim, Hyun Ju Han, Jee Ye Kim, Hyung Seok Park, Seho Park, Byeong Woo Park, Seung Il Kim, Eun Hae Cho, Gun Min Kim. Whole genome sequencing-based circulating tumor DNA profiling of metastatic breast cancer patients for molecular characterization and therapy response prediction [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD6-07.

Purpose: Several cell-free DNA (cf-DNA) features, such as genome-wide coverage, fragment size, and fragment end motif frequency, have shown their potentials for cancer detection. In this study, we developed two independent models, GC (gross chromatin), and FEMS (fragment end motif frequency and size). Each model uses images generated from genome-wide normalized sequencing coverage and cf-DNA fragment end motif frequencies according to the different cf-DNA size profiles. Then we integrated them into a single ensemble model to improve cancer detection and multi-cancer type classification accuracy. Methods: Low depth cf-WGS data was generated from 1,396 patients (stage I: 14.9%, stage II: 35.6%, stage III: 24.9%, stage IV: 24.2%, unknown: 0.4%) with breast (n=702), liver (n=213), esophageal (n=155), ovarian (n=151), pancreatic (n=85), lung (n=53), head and neck (n=16), biliary tract (n=15), and colon cancer (n=6) and 417 healthy individuals. Samples were randomly split into training, validation, and test set stratifying cancer type and stages. Cancer types with a small number of samples ( Results: For cancer detection, the ensemble model achieved sensitivities of 85.2% [95% confidence interval (CI): 71.8% to 94.5%], 74.9% (CI: 68.0% to 88.0%), 73.2% (CI: 66.7% to 85.9%) at a specificity of 95%, 98% and 99% and the AUC value of 0.97(CI: 0.95-0.99) in the test dataset. By the cancer stages, sensitivity was 62.8% (CI: 48.8% to 83.7%) in stage I, 66.3% (CI: 57.7% to 82.7%) in stage II, 85.9% (CI: 77.5% to 94.4%) in stage III, and 76.1% (CI: 63.4% to 87.3%) in stage IV at 99% specificity. For multi-cancer classification, the overall accuracy of 85.1% (CI: 80.4% to 89.3%) was achieved including 6 cancer types. Conclusions: Highly sensitive and accurate deep learning model for cancer detection and multi-cancer classification was generated by combining different types of cf-DNA features. This result provides the opportunity for general population multi-cancer screening using various cf-DNA features. Citation Format: Tae-Rim Lee, Jin Mo Ahn, Joo Hyuk Sohn, Sook Ryun Park, Min Hwan Kim, Gun Min Kim, Ki-Byung Song, Eunsung Jun, Dongryul Oh, Jeong-Won Lee, Joseph J Noh, Young Sik Park, Sun-Young Kong, Sang Myung Woo, Bo Hyun Kim, Eui Kyu Chie, Hyun-Cheol Kang, Youn Jin Choi, Ki-Won Song, Jeong-Sik Byeon, Junnam Lee, Dasom Kim, Chang-Seok Ki, Eunhae Cho. Deep learning algorithm for multi-cancer detection and classification using cf-WGS [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6371.

603 Background: Our previous study reported prognostic significance of copy number aberration (CNA) burden on low-pass whole genome sequencing (LP-WGS) based circulating tumor DNA (ctDNA) analysis in metastatic breast cancer patients. Here, we report the prognostic value of ctDNA CNA burden measured before neoadjuvant chemotherapy in stage II-III triple-negative breast cancer (TNBC) patients enrolled in phase III PEARLY trial (NCT02441933, BIG Supporter Study BIG 19-01, KCSG BR15-01). Methods: The PEARLY trial was performed as a randomized, open-label, multicenter, phase III study to test the efficacy and safety of adding carboplatin to (neo)adjuvant chemotherapy in patients with stage II-III TNBC. Patients were randomized in a 1:1 ratio to receive 4 cycles of AC followed by the taxane or taxane plus carboplatin (AUC 5, tri-weekly 4 cycles) as neoadjuvant or adjuvant therapy. This post-hoc baseline ctDNA analysis (before neoadjuvant chemotherapy) included only the neoadjuvant patient cohort with available baseline ctDNA results (n = 465, median follow-up 16.8 months), while it was blinded for randomization information (carboplatin or not). We used "I-score" method to estimate CNA burden of ctDNA by LP-WGS to be matched with disease-free survival (DFS) after primary surgery. Results: The baseline ctDNA I-score level was positively associated with clinical T and N stage, while baseline I-score was not different between patients with pathologic complete response (pCR) and non-pCR. We listed 465 patients in the order in which they underwent primary surgery, and then alternated patients to be assigned to exploratory cohort (n =232) and validation cohort (n = 233). The DFS was significantly shorter in high I-score (I-score ≥ 7.81) patients compared with low I-score (I-score < 7.81) patients in exploratory cohort. The high I-score independently predicted poor DFS adjusted for clinical T stage, clinical N stage, and pCR status (hazard ratio [HR] 3.88, p = 0.003). In the validation cohort, high I-score was validated to be associated poorer DFS, and multivariate Cox analysis validated the independent prognostic impact of I-score on DFS (HR 2.04 , p = 0.050). The high baseline I-score patients showed shorter DFS both in pCR-positive and pCR-negative patients. The 12-month DFS rate for pCR (+)/Low I-score patients was 98%, whereas that of pCR(-)/High I-score patients was 61.3 % in the validation cohort. Conclusions: The baseline ctDNA CNA burden on LP-WGS before neoadjuvant chemotherapy robustly predicts recurrence risk in stage II-III TNBC patients. The ctDNA I-score showed prognostic value independently from pCR status, suggesting ctDNA I-score can serve as a useful clinical determinant for escalating or de-escalating (neo)adjuvant strategy in TNBC patients. Clinical trial information: NCT02441933.

e18789 Background: Methylation analysis of cfDNA has been used to diagnose cancer in its early stages. Previous research has concentrated on local methylation signals using cancer type specific methylation markers. We used not just methylation markers, but also global methylation patterns for sensitive cancer detection. Methods: We generated methylation data from cancer patients (N = 717) and normal controls (N = 190) using cell-free methylated DNA immunoprecipitation and high-throughput sequencing (cfMeDIP-seq, N = 907) and cell free whole genome enzymatic methyl seq (cfWGEM-seq, N = 162) from cancer patients (N = 137) and normal controls (N = 25). We analyzed at the Illumina 450K methylation microarray (N = 3,479) from The Cancer Genome Atlas (TCGA) to find differentially methylated regions (DMR) in 6 cancer types (breast, lung, liver, ovarian, esophageal, and pancreatic cancer). After determining the overlapping DMRs of each dataset, the best 1661 regions that differed the most between the cancer patient group and the normal group were left. The selected marker-based model was cross-validated using cfMeDIP samples separated into training, validation, and test sets. Additionally, global methylation count values of cfMeDIP-seq data were used to train convolutional neural network. Finally, the global methylation pattern deep learning algorithm and the marker-based algorithm were combined to detect cancer. Results: Deep learning models based on selected markers and global methylation patterns achieved test data accuracy of 0.88-0.92 and 0.90-0.91, respectively, with AUC 0.94-0.96 and 0.95-0.96. The ensemble model of two models showed test data accuracy 0.91-0.92 and AUC 0.96-0.97 with the detection of early stage of cancers (stage 1:detection rate of 88-100%, stage 2:detection rate of 75-100%, stage 3:detection rate of 90-97%, stage 4:detection rate of 92-100%). Conclusions: In this study, we selected best markers by using tissue methylation dataset (TCGA) and cfDNA methylation datasets (cfMeDIP-seq, cfWGEM-seq). To train cancer detection models, we used not only the DMR pattern but also the global methylation pattern. And the ensemble model that included these features outperformed a single model. In the field of early cancer detection, our models show potential.

Background: This study is to evaluate the accuracy of machine learning for differentiation between optic neuropathies and pseudopapilledema (PPE). Methods: Two hundred and ninety-five images of optic neuropathies, 295 images of PPE, and 779 control images were used. Pseudopapilledema was defined as follows: cases with elevated optic nerve head and blurred disc margin, with normal visual acuity (>0.8 Snellen visual acuity), visual field, color vision, and pupillary reflex. The optic neuropathy group included cases of ischemic optic neuropathy (177), optic neuritis (48), diabetic optic neuropathy (17), papilledema (22), and retinal disorders (31). We compared four machine learning classifiers (our model, GoogleNet Inception v3, 19-layer Very Deep Convolution Network from Visual Geometry group (VGG), and 50-layer Deep Residual Learning (ResNet)). Accuracy and area under receiver operating characteristic curve (AUROC) were analyzed Results: The accuracy of machine learning classifiers ranged from 95.89% to 98.63% (our model: 95.89%, Inception V3: 96.45%, ResNet: 98.63%, and VGG: 96.80%). A high AUROC score was noted in both ResNet and VGG (0.999). Conclusions: Machine learning techniques can be combined with fundus photography as an effective approach to distinguish between PPE and elevated optic disc associated with optic neuropathies. Keywords: Machine Learning; Pseudopapilledema; Optic neuropathy; Optic disc swelling.