Spatial multiomics to inform immunocytokine engineering: knowledge base, gaps, and QC solutions.

Systemic pro-inflammatory cytokine therapies (e.g., IL-2) represented early milestones in immunotherapy, but their use is hampered by low response rates and severe off-target toxicity. In contrast, immunocytokines deliver cytokines directly to tumors, reducing systemic toxicity and enhancing efficacy. Spatial-omics provides deep insights into the tumor microenvironment (TME), enabling identification of druggable targets and accelerating development of novel antibody platforms and cytokine payloads. However, variability in patient sample quality affects data integrity, and platform differences require distinct preprocessing workflows. Spatio-temporal data demand spatial clustering to define disease-relevant niches, yet a lack of consensus about what constitutes a niche complicates interpretation and reproducibility. To overcome these challenges, effort needs to be made to improve sample collection and processing, and to reconcile the diversity of platforms with their technical limitations in niche identification. By combining knowledge of key TME cell types and marker expression with cytokines identified from autoimmune datasets, innovative immunocytokines can be designed to improve targeting, effectiveness, and patient outcomes.