Association between plasma catecholamine concentration and mortality in hospitalized adults with COVID-19: a secondary analysis of a multicenter randomized trial

The pathophysiology of COVID-19 sepsis is linked to activation of the renin–angiotensin–aldosterone system (RAAS) pathway, a heightened inflammatory cascade, and thrombosis. Catecholamines are likely central to all three [1]. COVID-19 sepsis is often characterized by hypoxemia, hypotension, and shock, which are treated with exogenous catecholamines. However, some data suggest detrimental effects of exogenous catecholamine use in sepsis [2]. We investigated the relationship between endogenous catecholamines and mortality in hospitalized patients with COVID-19 and assessed the relationship between catecholamines and angiotensin II (Ang II), cytokines, and platelets to explore associated mechanisms.

We analyzed epinephrine and norepinephrine values (High Sensitive ELISA assay; Eagle Biosciences) in blood samples from patients hospitalized with COVID-19 with new hypoxemia who participated in one of two randomized trials (ACTIV-4 Host Tissue clinical trial platform: NCT04924660). [3] Exclusion criteria were COVID-19 symptoms > 14 days, hospitalization > 3 days, or vasopressor use at randomization. Therapeutic trials ran with a shared placebo arm and investigated RAAS agents, TXA-127 (synthetic angiotensin1-7) and TRV-027 (angiotensin II type 1 receptor-biased ligand). To mitigate the treatment’s confounding effects, we used baseline blood samples from the placebo and treatment arms before randomization (day 0) and serial samples (days 3 and 5 post-randomization) from the placebo arm only. We measured cytokines (IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, KC/GRO, and TNFα with a validated multiplex assay [Meso Scale Diagnostics]), renalase (electrochemiluminescence assay platform [Meso Scale Diagnostics]), Ang II, and Ang-(1-7) (distinct radioimmunoassays, Hypertension Center, Wake Forest). We built Cox regression models for mortality and bivariate mixed-effect models to evaluate within-subject correlation of biomarkers over time using SAS version 9.4.

From October- November 2022, we analyzed data from 156 patients. Ninety-day mortality was 23% (median time: 15 days [IQR: 10–26]). Survivors and decedents had similar sex, race, ethnicity, comorbidities, baseline hemodynamics, and COVID-19 medications (p > 0.05). Decedents had higher age (61 years vs 51 years; p = 0.002), preexisting oxygen use (8% vs 1%, p = 0.044), vasopressor use (42% vs 7%; p < 0.001), and ICU admissions (67% vs 18%; p < 0.001) and worse WHO severity scores (> 4; 75% vs 23%; p < 0.001).

Median baseline values were 331.4 (range: 15.8–2397.2) pg/ml for endogenous norepinephrine and 30.3 (range: 0.0–871.20) pg/ml for endogenous epinephrine. Univariate analysis showed every 100 pg/ml increase in norepinephrine was associated with increased risk of death at baseline (HR = 1.09; 95% CI: 1.01–1.17; p = 0.028) and day 5 (HR = 1.07; 95% CI: 1.04–1.10; p < 0.001). Multivariable Cox regression analysis showed endogenous norepinephrine values were independently associated with 90-day mortality after adjusting for demographics, COVID-19 severity, and exogenous norepinephrine administration. Every 100 pg/ml increase in plasma norepinephrine from baseline was associated with an 11% increased risk of death (HR = 1.11; 95% CI; 1.01–1.21, Fig. 1A), and mortality risk was 14 times higher for patients with increasing trends over 5 days than decreasing trends (HR = 14.11; 95% CI: 2.3–86.75, Fig. 1B). Sensitivity analysis showed all associations remained robust statistically (p < 0.02) after excluding patients who received exogenous norepinephrine.

Fig. 1

(a) Multivariable Cox Regression Model of Norepinephrine and Mortality in patients hospitalized with COVID-19, and (b) Kaplan–Meier Survival Curves for Mortality Events and Norepinephrine (by Trend of Change) adjusted for covariates

Norepinephrine was significantly and positively correlated with RAAS and cytokines at baseline (Ang II, Ang1-7, IL-6, TNFα, IL-1β, IL-2, and IL-8; p < 0.01), day 3 (Ang II and TNFα; p < 0.02), and day 5 (Ang II, IL-6, IL-8, and IL-10; p < 0.01). Longitudinally, at the patient level, norepinephrine was positively correlated with Ang II (r = 0.67, p < 0.01), IL-6 (r = 0.68, p < 0.01) and IL-8 (r = 0.60, p < 0.03). We did not find significant correlation between epinephrine and mortality or with other biomarkers over time (p > 0.05).

We note some limitations. We did not stabilize catecholamines at collection and post-hoc, our study lacked an external validation cohort. However, our results were strengthened by a methodologically rigorous data collection platform across 35 sites.

Our results support the hypothesis that COVID-19 is a high-endogenous catecholamine state. We report high plasma values of norepinephrine (3–10 times that of healthy historical controls) in patients admitted with moderate-to-severe COVID-19 [4], corroborating similar results in another cohort of ICU COVID-19 patients [5]. We found high endogenous norepinephrine to be independently associated with 90-day mortality. This association remained valid, even excluding patients with exogeneous vasopressor use. The bivariate serial sample correlation between endogenous norepinephrine and Ang II and IL-6 may explain how sustained elevations of endogenous norepinephrine contribute to mortality risk. Endogenous norepinephrine is implicated in the upregulation of inflammation and thrombotic systems, which causes microvascular dysfunction in severe COVID-19. Taken together, our findings support further studies investigating the role of catecholamines in COVID-19-related shock.