Repurposing schirmer strips for tear biomarker profiling using infrared spectroscopy

To develop and validate a label-free analytical platform using attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy on Schirmer strips for tear biomarker profiling, and assess its diagnostic utility in differentiating aqueous-deficient dry eye disease (DED) subtypes.

Tear samples were collected using Schirmer strips from individuals undergoing routine ophthalmic examinations and from patients with aqueous-deficient DED, with or without primary Sjögren's syndrome. The wetted regions of the strips were analyzed using ATR-FTIR spectroscopy. Stepwise linear regression was employed to model the spectral contributions of individual proteins to the simulated tear spectrum. Spectral differences were quantified using Euclidean distances and correlated with biomarker concentrations and clinical parameters using Spearman correlation. Sparse partial least squares discriminant analysis (sPLS-DA) was then applied to classify DED subtypes.

Prominent spectral features were observed in the 800-1700 cm^-1 and ∼3000 cm^-1 regions. Model fit for the simulated tear spectrum progressively improved with the sequential addition of protein spectra. Spectral Euclidean distance effectively captured the biomarker concentration gradient, as exemplified by lactoferrin. It demonstrated stronger correlations with tear film breakup time and lipid layer thickness than Schirmer wetting length and remained stable after elution. sPLS-DA accurately classified DED subtypes with an AUC of 0.963, driven primarily by protein-related spectral features.

The integrated tear analysis platform based on ATR-FTIR spectroscopy of Schirmer strips enables accurate, cost-effective differentiation of DED subtypes by capturing tear molecular fingerprints, laying the foundation for integrating tear-based precision medicine into routine clinical workflows.