The Missing Antigens of Neuroendocrine Tumors: A Null-Hypothesis Framework for Testing Microbial Control of Spatial Immune Editing

Abstract

Well-differentiated gastroenteropancreatic neuroendocrine tumors (GEP-NETs) are commonly regarded as immunologically “cold,” yet this label does not explain their marked spatial heterogeneity in immune infiltration, antigenpresentation machinery, stromal organization, and malignant-cell state. Intratumoral bacterial signals have now been reported in neuroendocrine neoplasms, including recent pancreatic NET cohorts, while work in melanoma, glioblastoma, and colorectal cancer has shown that tumor-resident bacteria can contribute HLA-bound peptides and generate cognate T-cell responses. Whether these observations intersect in GEP-NETs remains unknown. This Perspective proposes a falsifiable, direction-agnostic framework rather than reporting an ongoing or completed study. The global null hypothesis is that verified intratumoral microbial residency does not culminate in a reproducible mechanistic chain linking microbial gene or protein, microbe-derived HLA-bound peptide, spatially concordant cognate T-cell clone, functional recognition, and immune selection. Nested sub-null hypotheses separately test residency, antigen presentation, cognate recognition, spatial immune association, and immune editing, so that partial evidence cannot be retrospectively promoted to global causal inference. The proposed design integrates multiregion histopathology, contamination-aware spatial mapping, orthogonal microbial detection, metagenomic/metatranscriptomic profiling, HLA-I/HLA-II immunopeptidomics, multiplex immune mapping, T-cell receptor sequencing, and functional perturbation. It additionally prespecifies gastrointestinal/ductal contamination models, microbial routes of entry, tiered tissue allocation, mixed-effects spatial statistics, multiplicity control, and confirmatory-versus-exploratory analyses. A positive result would not imply that microbes cause NET immune coldness or that microbial depletion would make a cold tumor hot. The decisive question is narrower: whether microbial antigens become part of the antigenic phenotype on which spatial immune selection acts.