Abstract IA03: A Tale of Two Biomarkers: Improving Outcomes for Cancer of Unknown Primary and Recurrent Early Stage Non-Small Cell Lung Cancer

Abstract Successful Cancer Precision Medicine requires robust, reliable, reproducible and clinically implementable biomarkers. My presentation will focus on the biomarker roadmaps for a tissue and a liquid biopsy biomarker to serve two unmet clinical needs. Cancers of Unknown Primary (CUP) are a heterogeneous group of metastatic cancers for which the primary site of the cancer remains undiagnosed. Patients with CUP have poor responses to empirical chemotherapy which, in most cases, is standard of care. Tissue biopsies can be challenging, and personalised therapies are difficult to access without knowing the tissue-of-origin (TOO). We therefore developed a liquid biopsy as an alternative approach to identify TOO. We developed CUPiD, a machine learning classifier utilising publicly available DNA methylation data, that made highly accurate TOO predictions across 29 tumour classes. CUPiD was tested on 143 cancer cfDNA samples from 13 known cancer types and on 27 non-cancer controls yielding overall sensitivity of 84.6% and TOO accuracy of 96.8%. We applied CUPiD to 41 patients with CUP and made a TOO prediction in 78% of which 72% were clinically consistent with subsequent primary tumour diagnosis or suspected/differential tumour diagnosis and 19% were for patients with no diagnostic suspicions. Combining CUPiD with cfDNA mutation data would allow us to re-classify the CUP diagnosis and/or change treatment stratification in 80% of CUP cases analysed. The turnaround of CUPiD is within 2-3 weeks considerably reducing time to improved treatment options compared to current practice. Further validation of CUPiD and testing across a wider range of tumour types is currently ongoing. Our goal is to personalise therapy for patients with CUP for improved outcomes. Community based low dose CT screening for lung cancer in the UK is leading to a substantial increase in detection of stage I Non-Small Cell Lung Cancer (NSCLC) cases. Despite curative intent surgery ∼20% of patients with stage I NSCLC will suffer from disease recurrence, with poorer prognosis for later stages. Using three independent patient cohorts, we investigated a COX-2-associated pro-tumorigenic inflammatory gene signature (PTI) that anticipates recurrence in stage I within a year of surgery in both Lung Adenocarcinoma (LUAD) and Lung Squamous Carcinoma (LUSC) patients. We translated this assay to a clinically compatible workflow using FFPE samples and confirmed the prognostic capability of PTI in real-world samples from routine care and screening programs. Our findings establishing PTI as a potential prognostic tissue-based biomarker to guide imaging surveillance frequency and adjuvant therapy decisions, supporting escalation in PTIhigh and de-escalation in PTIlow stage I patients. The potential for a COX-2 targeting anti-inflammatory intervention to prevent or delay disease recurrence now warrants further study. Citation Format: Caroline Dive. A Tale of Two Biomarkers: Improving Outcomes for Cancer of Unknown Primary and Recurrent Early Stage Non-Small Cell Lung Cancer [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2025 Oct 22-26; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2025;24(10 Suppl):Abstract nr IA03.