A collaboration between U.S.-based Bio-Techne and Israeli AI specialist Nucleai has, for the first time in a clinical trial, identified spatial immune interactions within tumor tissue as predictors of treatment success in metastatic melanoma. Data from the SECOMBIT trial will be presented at the Society for Immunotherapy of Cancer (SITC) 2025 Annual Meeting as one of the top 150 abstracts selected from over 1,200 submissions (Poster #528).
Researchers analyzed 42 pre-treatment biopsies using Bio-Techne’s COMET™ platform—a 28-plex multiplex immunofluorescence (mIF) panel—and integrated the imaging data with histopathology and clinical outcomes via Nucleai’s multimodal AI-powered spatial operating system. The results show that distinct cell-to-cell interactions in the tumor microenvironment predict progression-free survival (PFS), overall survival (OS), and clinical benefit across three treatment arms involving immune checkpoint blockade (ICB) and targeted therapy.
- Arm A (targeted therapy ? ICB): PD-L1+ CD8 T-cells and ICOS+ CD4 T-cells were associated with improved outcomes.
- Arm B (ICB ? targeted therapy): PD-1+ CD8 T-cells at the tumor invasive margin and their proximity to PD-L1+ CD4 T-cells correlated with longer survival.
- Arm C (targeted ? ICB ? targeted): APC-T-cell interactions at the tumor margin indicated better prognosis; macrophage interactions in the outer tumor microenvironment (TME) signaled poorer outcomes.
The study, led by Professor Paolo Ascierto (University of Naples Federico II), demonstrates that not just the presence, but the location and interaction of immune cells within the tumor critically determine therapeutic response.
Background: The Evolution of Personalized Melanoma Therapy
Since 2011, five-year survival rates for metastatic melanoma have risen from under 10% to over 50%, driven by BRAF/MEK inhibitors and PD-1/CTLA-4 blockers. Yet 40–50% of patients fail to respond initially, and resistance frequently emerges. Traditional biomarkers—such as PD-L1 expression or tumor mutation burden—explain only a fraction of treatment variability.
Spatial proteomics fills this gap: high-resolution multiplex imaging maps dozens of markers simultaneously, revealing whether killer T-cells infiltrate tumors or remain trapped in surrounding stroma. AI algorithms detect subtle patterns invisible to the human eye. The SECOMBIT trial (NCT02631447), initiated in 2016, compared three sequencing strategies using encorafenib/binimetinib and ipilimumab/nivolumab; long-term 2022 data had already identified Arms B and C as superior.
The newly reported spatial signatures could soon guide pre-treatment decisions—matching patients to the optimal regimen and avoiding overtreatment or undertreatment. Bio-Techne and Nucleai are showcasing new assays (ProximityScope™, expanded SPYRE™ portfolio) at Booth #603, enabling such analyses in hours rather than weeks.
Experts view this as a leap toward companion diagnostics: “Spatial biology plus AI makes precision medicine explainable and scalable,” said Nucleai CEO Avi Veidman. Larger-scale clinical validation is underway; first companion tests could reach the market by 2027.
