Abstract Rationale Dipeptidyl peptidase 3 (DPP3) is a zinc metalloprotease released into circulation upon cellular injury that degrades components of the renin-angiotensin-aldosterone system. Elevated circulating DPP3 has been linked to worse outcomes in critical illness cohorts, and a novel anti-DPP3 antibody improved cardiac and renal function in animal models of shock. We evaluated DPP3 as a prognostic biomarker and therapeutic target in severe acute hypoxemic respiratory failure (AHRF) and assessed its association with inflammatory phenotype. Methods We analyzed data from a prospective cohort of adults with AHRF requiring advanced respiratory support. We examined associations between baseline DPP3 and 28-day mortality (Cox regression), progression to intubation within 28 days among patients not ventilated at baseline (competing risk regression with death and discharge as competing events), and progression to successful extubation within 28 days among ventilated patients (competing risk regression with death as competing event). Models were adjusted for sex and baseline APACHE II. We used linear mixed-effects models with patient-level random effects to evaluate the impact of temporal changes in DPP3 from baseline to day 5. Inflammatory phenotype was determined using plasma biomarkers. We compared DPP3 values between phenotypes using a Wilcoxon rank-sum test, and measured associations between DPP3 and 28-day mortality within each phenotype. Results 281 patients from 15 US hospitals were included in this study. At baseline, mean age was 64 (SD ± 17) years, and 132 (47%) were mechanically ventilated. Median baseline DPP3 was 22 ng/ml (IQR 14-39). Compared to participants in the lowest quartile of baseline DPP3, participants in the highest quartile had a higher hazard of death (HR 3.11, 95%CI 1.67-5.79, figure), higher incidence of intubation (HR 3.8, 95%CI 1.3-11.0), and lower incidence of successful extubation (HR 0.4, 95%CI 0.2-0.9). DPP3 trajectory from baseline to day 5 was not significantly associated with mortality (P = 0.764). 76 (27%) patients were classified as hyperinflammatory, and the rest hypoinflammatory. Median DPP3 was 30 ng/ml (IQR 19-64) among hyperinflammatory patients and 20 ng/ml (IQR 13-33) among hypoinflammatory patients (P 0.001, figure). Within the smaller hyperinflammatory subset, baseline DPP3 was not significantly associated with mortality (Q4vQ1 HR 1.8, 95%CI 0.6-5.7). Within the larger hypoinflammatory subset, higher baseline DPP3 was associated with decreased survival (Q4vQ1 HR 3.3, 95%CI 1.5-7.2). Conclusions In severe AHRF patients, baseline DPP3 was strongly associated with mortality, intubation, and later extubation. Hyperinflammatory patients had higher baseline DPP3. DPP3 may represent a promising prognostic biomarker and potential therapeutic target in severe AHRF. This abstract is funded by: R35HL177135, K23HL173659, 5TL1TR001871-10
Yang et al. (2026) studied this question.