模式特异性相干干涉振动和频产生成像用于检测肺肿瘤及临床FFPE存档中的纳米级胶原重塑

Background: Nanoscale collagen remodeling is a pivotal physical signature of tumor development and metastasis. Despite its importance, there has been a critical technical gap inresolving these subtle structural cues in a clinical setting without exogenous labels. Here, we report a first-of-its-kind label-free imaging mechanism based on bond-specific coherent interference in vibrational sum-frequency generation (VSFG) microscopy, providing unprecedented sensitivity to tumor-associated nanostructural variations. Methods: We utilized hyperspectral VSFG microscopy to investigate collagen I remodeling inlung tumor tissues. A novel biophysical model was developed to translate mode-specific interference patterns-specifically the I(NHs)/I(CH2,ss) intensity ratio-into quantitative collagen interfibrillar distances at the 20-50 nm level. To ensure immediate clinical impact, we conducted a timely validation of this mechanism on archived formalin-fixed paraffin-embedded(FFPE) tissues, benchmarking the results against optimal cutting temperature (OCT) cryosections and atomic force microscopy (AFM) nanomechanical mapping. Results: Our findings reveal that metastatic lung tumors exhibit dramatic spectral shifts driven bydistinctive interferences between vibrational modes. We demonstrate, for the first time, that thesespectral signatures can serve as a direct readout for collagen packing density at the sub-50 nmscale, which directly correlates with the increased tissue stiffness observed in tumor progression.Crucially, we provide the first evidence that the molecular-level structural cues detected by VSFGare preserved through harsh clinical fixation and embedding processes. The diagnostic metrics obtained from deparaffinized FFPE samples were statistically indistinguishable from fresh-frozenOCT controls, effectively removing the major barrier to applying this technology to clinical pathology. Conclusions: This study establishes a high-priority diagnostic platform that bridges the gapbetween nanoscale biophysics and clinical oncology. By demonstrating that VSFG can extract high-fidelity structural signatures from both fresh tissues and the vast global archives of FFPE samples, this work enables large-scale retrospective prognostic studies that were previously impossible. This discovery provides a powerful, label-free tool for future pathology and represents a significant advancement in our ability to monitor and understand the tumor microenvironment at the nanoscale.