The activation of evolutionary-conserved inflammatory cascades, such as the pathogen-receptor recognition TLR4 pathway, triggered by multiple acute inflammatory conditions such as sepsis (viral, bacterial, fungal), acute pancreatitis, and trauma directly contribute to the multi-organ failure and high mortality associated with ARDS23,24. In addition to the intense inflammatory cascade that characterizes ARDS pathogenesis, endothelial and epithelial cell injury, dysregulated coagulation, apoptosis, and fibrosis are prominent features25. Accordingly, numerous potential biomarkers are generated by the acute dysregulation of multiple biochemical and cellular pathways characteristic of ARDS pathobiology25,26. Extracellular NAMPT was identified as one such potential biomarker and druggable target based on extensive research including preclinical mechanistic, genomic, and multi-species ARDS models5,12,27,28. The transcriptional regulation and blood//lung protein expression of NAMPT are highly induced by damage-associated molecular pattern proteins (DAMP) and multiple ARDS-relevant stimuli including bacterial infection, shock, trauma, hypoxia, and excessive mechanical stress29–32. Upon binding to TLR4, eNAMPT elicits a profound cytokine release that is mediated by proinflammatory transcription factors such as NFκB and leads to increased vascular permeability and ultimately multi-organ dysfunction12,29. The role of eNAMPT as a potential standalone diagnostic or prognostic biomarker in ARDS has not previously been demonstrated although eNAMPT was among a panel of six biomarkers that predicted mortality in ARDS13. An ideal diagnostic biomarker for ARDS would identify early stages of the syndrome, minimize disease heterogeneity, reflect the natural history of the syndrome, and be a potential target for a clinical trial4. A prognostic biomarker would provide information that addresses the overall outcome of ARDS and be potentially useful in stratifying patients for enrollment in clinical trials, thus enhancing the ability to detect beneficial effects from novel therapies4.
We have confirmed and validated, for the first time, that circulating eNAMPT is a potential diagnostic biomarker in ARDS and several ARDS-predisposing systemic acute inflammatory conditions including acute pancreatitis, sepsis, septic shock, and trauma. In addition to demonstrating that median circulating eNAMPT levels were significantly higher in ARDS and these ARDS-inducing acute inflammatory conditions when compared to healthy controls, we identified the range of eNAMPT values between 26 ng/ml and 33 ng/ml based on our in-house colorimetric ELISA assay to represent the best possible cut-off for distinguishing patients from healthy controls.
The translational utility of circulating eNAMPT as a diagnostic biomarker in ARDS include early diagnosis, minimizing disease heterogeneity, and facilitating the selection of subjects for enrollment in clinical trials, especially for therapies targeting this pathway28,33. An interesting finding in our analysis was the positive correlation between circulating eNAMPT levels and severity of acute pancreatitis. This also highlights the potential prognostic utility of eNAMPT as a biomarker. We did not replicate this positive correlation with disease severity in the sepsis, trauma, and ARDS cohorts, likely because of suboptimal classification of disease severity, non-standardized timing of specimen collection, and heterogeneity of cohorts.
The strengths of our report include, the inclusion of a large and diverse population of patients and controls in the discovery cohort, a robust validation with an independent cohort (patients with ARDS and sepsis), and use of two novel complementary ELISA assays, the MSD-Uplex assay electrochemiluminescent immunoassay, MesoScale (Meso Scale Discovery, MSD®)22 and our validated in-house ELISA assay30. Our findings are also strengthened by the consistency across all the cohorts where circulating eNAMPT levels were significantly higher in patients with acute systemic inflammatory states (sepsis, trauma, acute pancreatitis, and ARDS) when compared to controls.
A limitation of our analysis, which is inherent with the use of previously bio-banked specimens, is the heterogeneity of our cohorts as shown by variability in mortality rates. We did not compare the diagnostic performance of circulating eNAMPT to other diagnostic biomarkers in ARDS, such as soluble receptor for advanced glycation end products (sRAGE)34 and Angiopoeitin-235,36. We had limited information on generic severity of illness scales such as APACHE IV scores, sequential organ assessment (SOFA) scores or multiple organs dysfunction (MOD) scores, which limited the ability to better classify the patients within each cohort and further refine the diagnostic value of eNAMPT.