26 Background: Haystack MRD is a highly accurate, tumor-informed ctDNA assay designed for minimal residual disease (MRD) detection. To evaluate clinical performance and inter-laboratory reproducibility, retrospective testing of samples from early-stage colorectal cancer (CRC) patients enrolled in the randomized DYNAMIC and DYNAMIC-III trials was performed. Methods: Residual specimens from 59 patients (51 stage II, 8 stage III CRC) with annotated clinical outcomes were analyzed using the CLIA-validated Haystack MRD assay, which targets up to 50 tumor-informed, patient-specific variants derived from whole-exome sequencing of tumor and matched normal DNA. Testing was performed on 60 cell-free DNA (cfDNA) samples from 59 unique patients, containing a median of 3,800 genomic equivalents from plasma collected post-operatively (n=40) and post-adjuvant chemotherapy (n=20). Concordance of MRD results with clinical recurrence was assessed. Orthogonal validation was also conducted to evaluate reproducibility across two independent CLIA-certified laboratories: Quest Diagnostics Oncology Center of Excellence (Lewisville, TX) and Haystack Oncology (Baltimore, MD). Results: All 60 samples tested with Haystack MRD in parallel at the two sites yielded concordant MRD results (100% agreement). MRD positivity was observed in 10 of 12 samples from patients who experienced clinical recurrence, while 48 samples from patients who remained recurrence-free tested MRD-negative (sensitivity: 83%; specificity: 100%; positive predictive value [PPV]: 100%; negative predictive value [NPV]: 96%). Quantitative ctDNA measurements were between 0.32 (min) and 316.21 (max) mean ctDNA molecules/ml and were highly correlated across sites (Pearson correlation coefficient = 1.000; p<0.0001), supporting strong inter-laboratory reproducibility. Additional testing is ongoing. Conclusions: This study provides both clinical and orthogonal validation of the Haystack MRD assay using well-characterized specimens from the DYNAMIC trials. Results demonstrate high concordance with clinical recurrence outcomes and reproducibility across independent laboratories. These findings support the use of Haystack MRD as a robust approach for MRD detection following curative-intent treatment in stage II/III CRC, with potential to guide adjuvant therapy decisions and enable proactive, ctDNA-informed monitoring.

6075 Background: The rising incidence of oropharyngeal squamous cell carcinoma (OPSCC) is largely attributable to human papillomavirus associated (HPV+) disease, which accounts for ~70% of OPSCC cases. Circulating tumor (ct)DNA has the potential to enable more accurate treatment response assessment, guide response-adaptive management, and detect minimal residual disease to indicate persistence or recurrence. Both mutation-based tumor-informed ctDNA and ctHPV-DNA testing have demonstrated utility in HPV+ disease, but prospective intrapatient evaluations remain limited. A direct comparison of these approaches is essential to determine redundancy versus complementarity and to guide optimal integration into OPSCC patient management. Methods: In an ongoing prospective study, serial plasma samples were obtained from patients with stages I-IV OPSCC undergoing curative intent treatment. Up to 50 patient specific somatic variants were selected based on tumor whole exome sequencing to develop a personalized tumor-informed next generation sequencing (NGS) ctDNA assay (Haystack MRD) for plasma analysis. In patients with HPV+ disease (determined via ISH, IHC, and/or NGS), plasma was also analyzed using an NGS-based assay interrogating 13 high-risk HPV strains (Haystack HPV). Paired intrapatient samples were analyzed using percent agreement with 95% confidence intervals and Cohen’s kappa; concordance of dynamic changes was assessed using Spearman’s correlation. Results: As of January 2026, ctDNA results were available for 111 serial timepoints from 26 patients. The median number of timepoints per patient was 4 (range 1-9). Seventeen patients (65%) had HPV+, and 9 (35%) had HPV− disease. In HPV+ patients, across 85 longitudinal samples collected during multimodal treatment and post-treatment surveillance, mutation-based ctDNA and ctHPV demonstrated high concordance (91%; 95% CI, 82.5–95.2; κ=0.80). Of 30 ctDNA+ samples, 28 were ctHPV+ (93%; 95% CI, 78.7–98.2%), while 49 of 55 ctDNA- samples were ctHPV- (89%; 95% CI, 78.2–94.9%). Discordance was infrequent (8/85, 9.4%), predominantly ctHPV+/ctDNA- (6/85, 7.1%). All ctHPV+/ctDNA- cases occurred during neoadjuvant treatment monitoring and reflected earlier clearance of ctDNA, with ctHPV clearance lagging by several weeks to months. Two low-level (<100 parts per million) ctDNA+/ctHPV- cases were observed in the adjuvant setting. When both analytes were present, dynamic changes in ctDNA and ctHPV levels were highly concordant (Spearman’s ρ=0.94), although ctHPV was consistently detected at higher absolute levels. Conclusions: In HPV-driven OPSCC, tumor-informed ctDNA and ctHPV show high longitudinal concordance and distinct clearance kinetics, with earlier ctDNA clearance. Ongoing analyses will define how these assays can be optimally integrated into response assessment, treatment adaptation, and surveillance strategies.

30 Background: While many patients with LARC experience durable responses after TNT, relapse risk remains a concern and is not reliably predicted by standard clinicopathologic measures. ctDNA is a promising biomarker for minimal residual disease (MRD) that could improve post-TNT risk-stratification and guide postoperative surveillance; however, reduced sensitivity demonstrated in previous studies has limited its clinical utility in this setting. Methods: In an ongoing prospective study of patients with LARC receiving TNT, plasma was collected pre-treatment, after TNT but before surgery, post-surgery (“landmark”), and every 3 months for 1 year. ctDNA analysis was performed using a tumor-informed MRD assay interrogating up to 50 personalized variants (Haystack MRD), and results were evaluated alongside clinical outcomes. Results: Forty patients treated with TNT followed by surgery were included in this analysis. Pre-treatment ctDNA was detected in 13/13 (100%) patients. Among 34 patients with ctDNA results post-TNT, 32/34 (94%) had residual disease and 2/34 (6%) achieved pathologic complete response (pCR). ctDNA was detected post-TNT in 22/32 (69%) patients with residual disease, while both patients with pCR were ctDNA negative. Post-TNT sensitivity was higher in patients with pathologic stage III/IV versus I/II disease (15/18 [83%] vs 7/14 [50%]). Out of 20 patients with ctDNA results at the landmark post-surgical time point, all 5 (100%) patients with positive ctDNA relapsed, while 1/15 (7%) ctDNA-negative patients relapsed (this patient had detectable ctDNA at 6-months post-surgery), yielding 83% sensitivity and 93% negative predictive value for disease relapse at the landmark timepoint, and 100% sensitivity within 6 months of surgery. Testing of additional patients and time points is ongoing. Conclusions: A next-generation tumor-informed MRD assay shows excellent sensitivity for ctDNA detection in LARC patients pre-treatment (100%) and post-surgery (83% at landmark, 100% at 6 months). Post-TNT sensitivity was 69% overall and 83% in stage III/IV disease, representing an improvement over previous reports. These initial findings support integrating ctDNA into post-TNT risk stratification and postoperative surveillance in LARC; testing of additional patients/time points and further follow up is ongoing to validate these results.

4077 Background: Despite curative-intent treatment, patients with esophageal cancer experience a high risk of recurrence, and optimal patient management is limited by poor risk stratification and treatment response assessment strategies. ctDNA has demonstrated value as a prognostic biomarker in esophageal cancer; however, improved analytical and clinical performance of ctDNA analysis is necessary to reliably support patient management decisions in clinical practice. Methods: Between September 2017 and June 2023, plasma samples were collected from patients with esophageal cancer before, during, and after standard of care treatment in the routine care setting. In this retrospective analysis, ctDNA testing was performed on a subset of patients using a next-generation tumor-informed assay interrogating up to 50 personalized variants (Haystack MRD, Quest Diagnostics). Results: ctDNA was assessed in 149 samples from 51 patients with stage I-III esophageal adenocarcinoma (n= 40) or squamous cell carcinoma (n= 11). Fifteen patients with clinical follow-up (FU) available at the time of analysis had at least one sample collected after curative-intent treatment [neoadjuvant chemoradiotherapy (nCRT) and surgery (n=10) or definitive CRT (n=5)]. ctDNA was detected (ctDNA+) following curative-intent treatment in 5/15 (33%) patients, all 5 (100%) of whom experienced disease recurrence or were deceased at FU (median time from ctDNA+ result to FU: 14.4 months, range: 0.1-24.5). Of the 10 (67%) patients with no ctDNA detected (ctDNA-) following curative-intent treatment, 7 (70%) were disease-free at FU (median time from ctDNA- result to FU: 46.7 months, range: 6.3-65.5). In the neoadjuvant setting, paired pre- and post-nCRT samples were evaluated in 18 patients, demonstrating ctDNA detection in 18/18 (100%) patients prior to nCRT versus 8/18 (44%) following nCRT. ctDNA positivity following nCRT was strongly associated with poor pathological response (p=0.0026), and ctDNA dynamics observed longitudinally during nCRT served as a robust indicator of response. Of note, one patient experienced metastatic progression during nCRT, discovered at surgery, and ctDNA levels in this patient increased 550-fold while on nCRT. Conclusions: Evaluation of ctDNA using a next-generation tumor-informed platform supports improved response assessment to nCRT as well as accurate risk stratification following curative-intent treatment in patients with esophageal cancer. ctDNA positivity following curative-intent treatment predicted disease recurrence with a lead time of up to 22 months. Additional analyses are ongoing to further validate these findings.

Introduction: Tumor-informed minimal residual disease (MRD) monitoring assays based on plasma circulating tumor DNA (ctDNA) are increasingly integrated into the clinical management of patients with cancer. In the post-surgical curative intent setting and as an adjunct to radiographic imaging for response assessment and surveillance, timely identification of MRD may better guide therapeutic decision-making. The increasing clinical adoption of MRD testing, supported by a growing body of evidence demonstrating its potential utility at critical decision points across diverse tumor histology, has brought attention to the variability in the analytical performance of available ctDNA assays. This variability is becoming increasingly appreciated as a key factor influencing clinical performance. Case Presentations: Here we report a case series in breast and rectal cancer involving treatment monitoring with a novel advanced MRD assay, illustrating its ability to identify subclinical metastasis and disease resolution below the validated limit of detection of a commercially available ctDNA assay in these cases. Conclusion: Results aided medical decision-making and underscored the need for highly sensitive assays in MRD detection. The differences in sensitivity, driven primarily by analytical variables, highlight the importance of selecting an assay that is not only analytically robust but also appropriately matched to the patient’s specific clinical context, to help ensure optimal utility and minimize the risk of misinterpretation.

Importance:In a randomized clinical trial, treatment guided by tumor-informed circulating tumor (ct)DNA testing reduced adjuvant chemotherapy use without compromising recurrence-free survival in patients with stage II colon cancer. The potential effects of adopting ctDNA testing into routine patient care is unknown. Objective:To compare the total cost of patient care scenarios with and without the adoption of ctDNA testing. Design, Setting, and Participants:This budget impact analysis was conducted from the perspectives of US commercial health and Medicare Advantage payers. A decision-analytical model was populated with age-specific incidence of colon cancer, use of adjuvant chemotherapy, and use of single-agent or multiagent regimens. Total cost was estimated with the costs of ctDNA testing, drug acquisition, administration, surveillance, and adverse events. The analysis was conducted from September 2023 to January 2024. Exposures:The adoption of ctDNA testing. Main Outcomes and Measures:The incremental cost in the first year following the adoption of ctDNA testing, where testing will affect patient treatment and costs. Results:In hypothetical plans with 1 million individuals covered, 35 commercial health plan members and 102 Medicare Advantage members aged 75 years and younger were eligible for ctDNA testing. In the base case with a 50% adoption rate, total cost savings were $221 684 (equivalent to $0.02 per member per month [PMPM]) for a commercial payer and $116 720 (equivalent to $0.01 PMPM) for a Medicare Advantage payer. Cost savings were robust to variations in assumptions of all parameters in the commercial population but sensitive to variations in assumptions of adjuvant chemotherapy use rates in the Medicare Advantage population. The number needed to test to avoid 1 patient receiving adjuvant chemotherapy was 4 in the commercial population and 10 in the Medicare Advantage population. The budget-neutral cost for ctDNA testing was $16 202 for a commercial payer and $5793 for a Medicare Advantage payer. Conclusions and Relevance:Use of tumor-informed ctDNA testing to guide adjuvant chemotherapy in postsurgery patients with stage II colon cancer was projected to result in cost savings for both commercial and Medicare Advantage payers. Adoption of ctDNA testing is therefore advantageous from a budgetary perspective.

Supplementary Figure 2 from Detection of Tumor DNA at the Margins of Colorectal Cancer Liver Metastasis

Supplementary Tables 1-2 from Detection of Tumor DNA at the Margins of Colorectal Cancer Liver Metastasis

OBJECTIVES:The aim of this pilot study was to evaluate the presence of somatic mutations in matched tumor and circulating DNA (ctDNA) samples from patients with primary head and neck squamous cell carcinoma (HNSCC) and assess the association of changes in ctDNA levels with survival. MATERIALS AND METHODS:Our study included 62 patients with stage I-IVB HNSCC treated with surgery or radical chemoradiotherapy with curative intent. Plasma samples were obtained at baseline, at the end of treatment (EOT), and at disease progression. Tumor DNA was extracted from plasma (ctDNA) and tumor tissue (tDNA). The Safe Sequencing System was used assess the presence of pathogenic variants in four genes (TP53, CDKN2A, HRAS and PI3KCA) in both ctDNA and tDNA. RESULTS:Forty-five patients had available tissue and plasma samples. Concordance of genotyping results between tDNA and ctDNA at baseline was 53.3%. TP53 mutations were most commonly identified at baseline in both ctDNA (32.6%) and tDNA (40%). The presence of mutations in this restricted set of 4 genes in tissue samples at baseline was associated with decreased overall survival (OS) [median 58.3 months for patients with mutations vs. 89 months for patients without mutations, p < 0.013]. Similarly, patients presenting with mutations in ctDNA had shorter OS [median 53.8 vs. 78.6 months, p < 0.037]. CtDNA clearance at EOT did not show any association with PFS or OS. CONCLUSIONS:Liquid biopsy enables real-time molecular characterization of HNSCC and might predict survival. Larger studies are needed to validate the utility of ctDNA as a biomarker in HNSCC.

Supplemental Tables and Figure Caption

Supplementary Figure 1 from Detection of Tumor DNA at the Margins of Colorectal Cancer Liver Metastasis

Abstract Background Human papillomavirus (HPV)-associated oropharyngeal cancer (OPC) has a favorable prognosis which has led to efforts to de-intensify treatment. Response-adaptive de-escalated treatment is promising, however improved biomarkers are needed. Quantitative cell-free HPV-DNA (cfHPV-DNA) in plasma represents an attractive non-invasive biomarker for grading treatment response and post-treatment surveillance. This prospective study evaluates dynamic changes in cfHPV-DNA during induction therapy, definitive (chemo)radiotherapy, and post-treatment surveillance in the context of risk and response-adaptive treatment for HPV + OPC. Methods Patients with locoregional HPV + OPC are stratified into two cohorts: High risk (HR) (T4, N3, $$\ge$$ ≥ 20 pack-year smoking history (PYH), or non-HPV16 subtype); Low risk (LR) (all other patients). All patients receive induction chemotherapy with three cycles of carboplatin and paclitaxel. LR with ≥ 50% response receive treatment on the single-modality arm (minimally-invasive surgery or radiation alone to 50 Gy). HR with ≥ 50% response or LR with ≥ 30% and < 50% response receive treatment on the intermediate de-escalation arm (chemoradiation to 50 Gy with cisplatin). All other patients receive treatment on the regular dose arm with chemoradiation to 70 Gy with concurrent cisplatin. Plasma cfHPV-DNA is assessed during induction, (chemo)radiation, and post-treatment surveillance. The primary endpoint is correlation of quantitative cfHPV-DNA with radiographic response. Discussion A de-escalation treatment paradigm that reduces toxicity without compromising survival outcomes is urgently needed for HPV + OPC. Response to induction chemotherapy is predictive and prognostic and can select candidates for de-escalated definitive therapy. Assessment of quantitative cfHPV-DNA in the context of response-adaptive treatment of represents a promising reliable and convenient biomarker-driven strategy to guide personalized treatment in HPV + OPC. Trial registration This trial is registered with ClinicalTrials.gov on October 1st, 2020 with Identifier: NCT04572100 .

Purpose/Objective(s)Human papillomavirus (HPV) associated oropharyngeal cancer (HPV+OPC) has a favorable prognosis. Current response-adapted strategies utilize RECIST to guide therapy. In the present study, we examined the utility of cell-free HPV DNA (cfHPV DNA) kinetic responses as an additional parameter to track therapy efficacy in de-escalation strategies. Here, we report a comparison of cfHPV DNA kinetics by RECIST and tumor volume (TV) from a prospective response-adaptive de-escalation trial.Materials/MethodsPatients enrolled on the prospective phase II OPTIMA 2 (NCT03107182) de-escalation trial with available cfHPV DNA information were included for analysis. OPTIMA 2 enrolled locoregionally advanced HPV+OPC including low and high risk patients. High risk patients were defined as T4, N2c-N3 (AJCC 7th edition), > 20 pack year smoking history, or non-HPV16 subtype. Neck CT or MRI as well as cfHPV DNA was obtained at baseline and following induction chemoimmunotherapy. Target lesions were assessed via RECIST 1.1 at baseline and post-induction therapy per protocol; patients with ≥ 50% response by RECIST received de-escalated radiation dose and volume. cfHPV DNA was detected and quantified using a CLIA-certified cfHPV sequencing assay (HPV-SEQ, Sysmex Inostics, Inc.). For this analysis, target lesions were delineated post hoc to obtain TV metrics. Linear regression was used to examine correlation between cfHPV DNA, RECIST, and TV response. Receiver Operating Characteristic (ROC) analysis was used to identify an optimal TV response to predict ≥ 95% cfHPV DNA response.Results39 patients of which 31 had follow up cfHPV DNA information were evaluable. Median RECIST was 4.9 cm at baseline and 2.0 cm at follow up. Median TV was 29 cm3 at baseline and 6 cm3 at follow-up. 39 (100%) patients had detectable cfHPV DNA at baseline, with median 180 [IQR: 15, 966] copy number/mL. Of 31 patients with paired cfHPV DNA information at baseline, 29 (94%) had a ≥ 95% cfHPV DNA reduction. Of those, cfHPV DNA was undetectable in 21 (71%) cases. Baseline TV was associated with cfHPV DNA (p=0.04), and the median cfHPV DNA / TV ratio was 5.5 (copy number / cm3). No differences were observed between low and high-risk cohorts with respect to cfHPV DNA or radiographic metrics. RECIST was not associated with cfHPV DNA metrics. When adjusting for risk, TV response predicted cfHPV DNA response (p = 0.003). ROC analysis identified a TV response of 63% (AUC 0.88) to predict ≥ 95% cfHPV DNA response; 74% of patients had a TV response ≥ 63%.ConclusioncfHPV DNA kinetics were more pronounced than radiographic response to systemic therapy in both high and low risk groups. cfHPV DNA was predicted by TV response to chemoimmunotherapy, but was not predicted by traditional RECIST dynamics. We identified potential TV responses to predict cfHPV DNA response. This represents the first direct correlation of circulating tumor DNA levels to tumor volume and has implications for response assessment in the neoadjuvant setting as well as in other solid tumors.

Abstract Background: Access to molecular testing for metastatic NSCLC (mNSCLC) patients has been improved by circulating tumor DNA (ctDNA) based liquid biopsies (LB), which can obviate invasive procedures, expedite results, and enable serial testing. Emerging clinical applications for LB include therapeutic efficacy monitoring and minimal residual disease (MRD) detection, which demand a ctDNA assay with high sensitivity, specificity, and appropriate genomic coverage. Here we demonstrate that SafeSEQ next-generation sequencing (NGS) LB delivers equivalent performance to OncoBEAM digital PCR, a sensitive ctDNA approach that has been extensively clinically validated in pivotal trials (e.g., AURA, TIGER-X). Importantly, SafeSEQ delivers significantly expanded genomic coverage to address the need for expanding targeted therapy indications, monitoring treatment response, and detecting MRD. Methods: Whole blood samples (n=176) were collected from mNSCLC patients prior to/during treatment or at disease progression and transported to a CLIA laboratory for OncoBEAM analysis to detect mutations in EGFR (exon 19 del, L858R, T790M, C797S), KRAS (codons 12, 13, 61) and BRAF V600E. Replicate plasma aliquots were analyzed with SafeSEQ to interrogate clinically relevant regions in BRAF, EGFR, ERBB2, KRAS, MET, NRAS, PIK3CA, and TP53. Mutation level concordance between the methods was assessed, where only genomic alterations interrogated by both platforms were considered. For mutations detected by both methods, correlation analysis of mutant allelic frequency (MAF) was performed. Results: Concordance analysis of the mutation results from OncoBEAM and SafeSEQ testing of 176 replicate patient samples demonstrated an overall percent agreement (OPA) of 99.6%, with a positive percent agreement (PPA) of 78.1% and a negative percent agreement (NPA) of 99.9%. The mean MAF for discordant mutations (n=16) was 0.06% (range: 0.04-0.12%). When considering mutations with MAF >0.1%, PPA and OPA increased to 96.0% and 99.9%, respectively. MAF levels for mutations detected by both methods demonstrated a strong linear correlation (R2=0.98). Of 124 patient samples having no mutation detected by OncoBEAM, 75 (60%) showed ≥1 alteration with SafeSEQ. Panel-wide, 76% and 30% of all mutations were detected at <1% and <0.1% MAF, respectively. Conclusions: SafeSEQ demonstrates clinical sensitivity comparable to OncoBEAM, with a strong positive correlation between MAF values across a broad dynamic range. SafeSEQ also provides expanded coverage across broader genomic regions, which - when combined with robust clinical performance - should better inform treatment selection, improve high resolution monitoring of therapeutic efficacy, and enable MRD detection and surveillance for NSCLC patients. Citation Format: Hillary Sloane, Priya Sathyanarayan, Daniel Edelstein, Frederick Jones, Jennifer Preston, Sam Wu, Jenna Los, Lara Duchstein, Johannes Fredebohm, Katharina Wichner, Denise Heim, Frank Holtrup, Hannah Quinn, David Feller-Kopman. Clinical evaluation of NGS-based liquid biopsy genotyping in non-small cell lung cancer (NSCLC) patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr LB053.

BACKGROUND:The development of secondary resistance (SR) in metastatic colorectal cancer (mCRC) treated with anti-epidermal growth factor receptor (anti-EGFR) antibodies is not fully understood at the molecular level. Here we tested in vivo selection of anti-EGFR SR tumors in CRC patient-derived xenograft (PDX) models as a strategy for a molecular dissection of SR mechanisms.METHODS:We analyzed 21 KRAS, NRAS, BRAF, and PI3K wildtype CRC patient-derived xenograft (PDX) models for their anti-EGFR sensitivity. Furthermore, 31 anti-EGFR SR tumors were generated via chronic in vivo treatment with cetuximab. A multi-omics approach was employed to address molecular primary and secondary resistance mechanisms. Gene set enrichment analyses were used to uncover SR pathways. Targeted therapy of SR PDX models was applied to validate selected SR pathways.RESULTS:In vivo anti-EGFR SR could be established with high efficiency. Chronic anti-EGFR treatment of CRC PDX tumors induced parallel evolution of multiple resistant lesions with independent molecular SR mechanisms. Mutations in driver genes explained SR development in a subgroup of CRC PDX models, only. Transcriptional reprogramming inducing anti-EGFR SR was discovered as a common mechanism in CRC PDX models frequently leading to RAS signaling pathway activation. We identified cAMP and STAT3 signaling activation, as well as paracrine and autocrine signaling via growth factors as novel anti-EGFR secondary resistance mechanisms. Secondary resistant xenograft tumors could successfully be treated by addressing identified transcriptional changes by tailored targeted therapies.CONCLUSIONS:Our study demonstrates that SR PDX tumors provide a unique platform to study molecular SR mechanisms and allow testing of multiple treatments for efficient targeting of SR mechanisms, not possible in the patient. Importantly, it suggests that the development of anti-EGFR tolerant cells via transcriptional reprogramming as a cause of anti-EGFR SR in CRC is likely more prevalent than previously anticipated. It emphasizes the need for analyses of SR tumor tissues at a multi-omics level for a comprehensive molecular understanding of anti-EGFR SR in CRC.

BACKGROUND AND OBJECTIVES:The use of ultra-sensitive diagnostic tests to detect clinically actionable somatic alterations within the gene encoding the epidermal growth factor receptor (EGFR) within circulating cell-free DNA is an important first step in determining the eligibility of patients with non-small cell lung cancer to receive tyrosine kinase inhibitors. METHODS:We present the clinical validation (accuracy, sensitivity, and specificity) of a highly sensitive OncoBEAMTM EGFR V2 test, which we compare to a custom next-generation sequencing assay, for the treatment of patients with non-small cell lung cancer with EGFR tyrosine kinase inhibitor therapies. The OncoBEAMTM digital-polymerase chain reaction method detects 36 different EGFR alterations in circulating cell-free DNA, whereas the next-generation sequencing assay covers major solid tumor oncodrivers. Of the 540 samples analyzed with the OncoBEAMTM EGFR V2 test, 42.4% of patients had undergone molecular testing at diagnosis (N = 229/540) and 57.7% of patients during disease progression (N = 311/540). RESULTS:The sensitivity and specificity were measured for this BEAMing assay. The number of mutant beads and mutant allelic fraction were measured for each EGFR alteration and the level of detection was established at 0.1% for a median of 2861 genome equivalent (GE) in each reaction using HD780 horizon control DNA, as well as by an internal quality reference standard. Approximately 10%, 27%, and 63% of the 540 samples contained < 1500 GE, a range of 1500-3000 GE, and > 3000 GE, which corresponded to a maximal assay sensitivity of 2.0%, 0.5-0.1%, and 0.1-0.05% mutant allelic fraction, respectively. In a routine hospital setting, 11.4% of non-small cell lung cancer tumors were positive at diagnosis for EGFR alterations, while 43.7% samples harbored EGFR mutations at progression, among which 40.3% expressed EGFR resistance mutations after first-line tyrosine kinase inhibitor treatment with first- and second-generation drugs. CONCLUSIONS:The OncoBEAMTM EGFR V2 is a sensitive, robust, and accurate assay that delivers reproducible results. Next-generation sequencing and BEAMing technologies act complementarily in the routine molecular screening. We show that using a next-generation sequencing assay, despite its lower sensitivity, enables the identification of rare EGFR alterations or resistance mechanisms (mutation, deletion, insertion, and copy number variation) to orient first- and second-line treatments.

Abstract Purpose: Expanded RAS/BRAF mutations have not been assessed as predictive for single-agent cetuximab in metastatic colorectal cancer (mCRC), and low mutant allele frequency (MAF) mutations are of unclear significance. We aimed to establish cetuximab efficacy in optimally selected patients using highly sensitive beads, emulsion, amplification, and magnetics (BEAMing) analysis, capable of detecting alterations below standard clinical assays. Patients and Methods: CO.17 trial compared cetuximab versus best supportive care (BSC) in RAS/BRAF-unselected mCRC. We performed RAS/BRAF analysis on microdissected tissue of 242 patients in CO.17 trial using BEAMing for KRAS/NRAS (codons 12/13/59/61/117/146) and BRAF V600E. Patients without BEAMing but with previous Sanger sequencing–detected mutations were included. Results: KRAS, NRAS, and BRAF mutations were present in 53%, 4%, and 3% of tumors, respectively. Cetuximab improved overall survival [OS; HR, 0.51; 95% confidence interval (CI), 0.32–0.81; P = 0.004] and progression-free survival (PFS; HR, 0.25; 95% CI, 0.15–0.41; P < 0.0001) compared with BSC in RAS/BRAF wild-type patients. Cetuximab did not improve OS/PFS for KRAS-, NRAS-, or BRAF-mutated tumors, and tests of interaction confirmed expanded KRAS (P = 0.0002) and NRAS (P = 0.006) as predictive, while BRAF mutations were not (P = 0.089). BEAMing identified 14% more tumors as RAS mutant than Sanger sequencing, and cetuximab lacked activity in these patients. Mutations at MAF < 5% were noted in 6 of 242 patients (2%). One patient with a KRAS A59T mutation (MAF = 2%) responded to cetuximab. More NRAS than KRAS mutations were low MAF (OR, 20.50; 95% CI, 3.88—96.85; P = 0.0038). Conclusions: We establish single-agent cetuximab efficacy in optimally selected patients and show that subclonal RAS/BRAF alterations are uncommon and remain of indeterminate significance.