Abstract 2743: Molecular characterization and prognostic evaluation of disseminated cancer cell subpopulations detected in bone marrow of non-small cell lung cancer patients

Abstract Introduction: Disseminated cancer cells (DCC) can be detected years before metastatic manifestation in bone marrow of lung cancer patients. Although a consensus protocol has been established, the impact of markers and methods for DCC detection are still not settled. We recently found that EpCAM+ DCC detected by immunofluorescence (IF) outperform the classical immunocytologic (IC) detection by cytokeratin (CK) antibodies regarding prognostic impact. However, it remained unclear whether this difference resulted from biological traits associated with EpCAM+ vs CK+ DCC or from methodological reasons. We therefore performed a side-by-side comparison of IC vs IF for CK and EpCAM. Methods: We prospectively collected bone marrow (BM) aspirates from 56 non-metastasized NSCLC patients that underwent tumor resection with curative intention. After density gradient centrifugation, we analyzed the mononucleated cells by (i) IC using the anti-CK antibody A45-B/B3 and (ii) by IF using the anti-CK antibody A45-B/B3 and the anti-EpCAM antibody HEA-125. IF signals were captured and objectively quantified using ImageJ, whereas IC was evaluated by two independent observers. Marker-positive cells were isolated and subjected to Whole Genome Amplification (WGA) and whole genome sequencing for assessment of copy number alterations (CNA). Survival analysis was performed for progression-free survival (PFS), tumor-specific survival (TSS) and overall-survival (OS) after a median follow-up time of 72. 5, 78. 7, 78. 7 months, respectively. Results: IC and IF revealed the presence of CK+ DCCs in 37. 5% and 41. 1% of cases, respectively (P=0. 34). IF identified EpCAM+ DCCs in 55. 4% of patients, resulting in a 63. 8% DCC-positivity rate found by IF, when combined with CK. The average number of DCCs per million MNC, the DCC density (DCCD), was 0. 92, 0. 93 and 1. 44 for CKIC and CKIF and for EpCAMIF, respectively. No association was found between CKIC + DCCs and PFS, TSS or OS. In contrast, detection of EpCAMIF + and/or CKIF + DCCs showed a trend towards reduced PFS (P=0. 055), TSS (P=0. 225) and OS (P=0. 155). Subgroup analysis revealed that both EpCAMIF + and CKIF + cells were correlated with worse PFS (P=0. 016 and P=0. 040, respectively) and TSS (P=0. 033 and P=0. 008, respectively). CKIF-positivity was also associated with worse OS (P=0. 014), whereas EpCAM was not. CKIF + DCCs remained an independent prognostic variable for TSS and OS upon multivariate testing (hazard ratio: 6. 45 and 3. 31, respectively), while EpCAMIF + was not. The IF-marker-defined cell phenotypes exhibited comparable frequencies and distributions of CNAs across the entire genome (~50% of cells). Conclusions: Detection of CK+ DCC by quantified IF analysis apparently outperforms CK detection by IC in survival analysis. The impact of CK and EpCAM expression levels on outcome are under investigation. Citation Format: Daniel Spitzl, Tobias Mederer, Julius Zank, Severin Gütter, Michael Ried, Hans-Stefan Hofmann, Christoph A. Klein. Molecular characterization and prognostic evaluation of disseminated cancer cell subpopulations detected in bone marrow of non-small cell lung cancer patients 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 (6Suppl): Abstract nr 2743.