An Evaluation of MACSima and PhenoCycler-Fusion for Spatial Image Analysis in a Medulloblastoma Xenograft Model

In spatial proteomics, multiplexed immunofluorescence (mIF) enables high-plex visualization of protein expression in preserved tissue, offering insights into tumor heterogeneity and the microenvironment. While MACSima and PhenoCycler-Fusion employ distinct strategies, direct comparisons under biologically controlled in vivo conditions remain limited. We applied both platforms to sagittal formalin-fixed, paraffin-embedded (FFPE) sections from an orthotopic xenograft mouse model of human medulloblastoma (MB), featuring leptomeningeal seeding (LMS). These longitudinal sections spanning brain and spinal cord allowed simultaneous assessment of areas with distinct cellular architecture. Fifteen-marker mIF was performed. MACSima utilized MICS technology with MACS iQ View for automated workflows; PhenoCycler-Fusion used a DNA-barcoded antibody system and QuPath for open-ended image processing. Segmentation was evaluated using identical MACSima data. Both platforms enabled high-plex imaging mIF while preserving tissue morphology. DAPI, Ki-67, and Actin were consistently detected across both systems. Ki-67 expression localized to densely packed tumor regions and was also observed in lower-density areas. Analyzing MACSima data, MACS iQ View detected fewer cells but a higher Ki-67 positive rate in dense regions; conversely, QuPath detected more cells but with a lower positivity rate. In low-density areas, both tools yielded similar results. These differences reflect distinct segmentation algorithms and thresholding strategies. This study confirms both platforms support mIF-based spatial proteomic analysis in complex, heterogeneous tissues. However, analysis tools influence quantification. Therefore, standardization of algorithmic settings and additional validation are crucial for precise data interpretation. This research provides practical insights for platform selection in basic, translational, and clinical applications by directly evaluating staining, image acquisition, and analysis pipelines.