No Spatially Resolvable Perivascular Monocyte Gradient in Healthy Human Lung: A Spatial-Transcriptomic Test of the Sustained-Extravasation Model of Interstitial Macrophage Renewal

Background: Whether lung interstitial macrophages (IMs) are continuously replenished by blood-monocyte extravasation remains unresolved. We performed a hypothesis-driven secondary analysis of four public datasets. Methods: Dual-module scoring (MonoScore/ResidScore) was combined with vessel-distance gradient tests across the Madissoon 2023 healthy-lung atlas, the 10x PBMC 10k reference, and the Natri 2024 lung-disease cohort, using donor or slide as the statistical unit, spatially aware permutation nulls, and multiple-testing correction. Results: Among 40,634 healthy lung myeloid cells (34,912 in six axis populations), a continuous monocyte-to-resident axis placed perivascular IMs at the resident pole; batch-free intra-lung controls separated lung CD14 monocytes from perivascular IMs (AUC=0.9924), whereas cross-dataset AUCs are batch-confounded upper bounds not used for inference. Across three Visium slides (two donors), ResidScore signal per spot rose with vessel distance (3/3 slides concordant; toroidal-shift p=0.009-0.016, not surviving Bonferroni correction), while no slide showed the predicted monocyte-signal decay; block-level intervals were uninformative. In idiopathic pulmonary fibrosis (IPF), IM MonoScore was elevated (+0.088, p_adj=0.014 across 36 tests) without significant ResidScore loss or compositional change (p=0.68). Conclusions: This cross-sectional design cannot test ontogeny; it tests the spatial corollary of sustained extravasation, which steady-state adult human lung does not support — without contradicting mouse fate-mapping evidence, which addresses developmental and turnover-level origin. The perivascular niche emerges as a candidate renewal interface with a locally attenuated residency program. In IPF, IM monocytization reflects transcriptional reprogramming rather than compositional replacement.