Several strategies have been used to determine biomarkers for Chronic obstructive pulmonary disease (COPD) diagnosis and phenotypes? identification. However, although omics may initially produce non-hypothesis-biased results, the typical subsequent statistical processing is not blind, as it is based on previously defined phenotypes.

Objectives: To evaluate the potential of using completely blind proteomic analysis complemented with hypothesis-driven multiplex to identify the presence of both COPD and exacerbations

Methods: The plasma of 34 individuals from H. Mar (24 stable & 10 exacerbated COPD and 10 controls) was analysed using liquid chromatography-tandem mass spectrometry (LC-MS/MS) and multiplex immuno-based. The analysis of the results was clinically blind, based solely on a clustering approach. Only at the end the blinded generated clusters were confronted with clinical groups.

Results: Clusters did not help in COPD diagnosis, but those from LC-MS/MS differentiated exacerbations from the stable phase (p=0.01), showing a specificity of 79±25% and a sensitivity of 80±25% (accuracy: 79±25%). The differentially abundant proteins involved were associated with inflammatory and antibody-mediated immune responses, blood coagulation, lipid profile modulation, and complement pathways. Clusters based on the multiplex did not help identify acute episodes.

Conclusions: A clinically blinded LC-MS/MS strategy demonstrated that extended proteomic profiling could help identify COPD exacerbation.

FUNDING: FIS (PI21/00785, BA22/00009 & PFIS FI22/000003) & CIBERES, ISCIII-EU; SEPAR Grants 2019 & 2021; Catalan Government (2021 SGR 00100).