Troponin elevation and vascular involvement post-COVID-19

Abstract Introduction COVID-19 has been linked to cardiovascular complications, including myocarditis and vascular dysfunction. Cardiac troponin elevation is relatively common in hospitalised patients with COVID-19, but not usually associated with myocarditis. This prevalent rise in cardiac troponin remains largely unexplained. Purpose We hypothesised cardiac troponin elevation seen in COVID-19 is associated with systemic vascular injury. Methods Convalescent COVID-19 patients underwent comprehensive phenotyping, including blood biomarker analysis (n=208) and imaging (n=159, MRI). Myocarditis was adjudicated using the Modified Lake Louise Criteria. Plasma proteomics analysed 7,288 proteins at enrolment and 28-60 days. Differential expression with Benjamini-Hochberg adjustment and pathway analysis (GO-CC, Hallmark) were performed for participants with adjudication and proteomics at enrolment (n=155, V1) and 28-60 days (n=154, V2). Gluteal subcutaneous biopsies were obtained from a sub-study of 37 participants (27 COVID-19, 10 controls; mean age 57, 48% women, 41% cardiovascular disease) to evaluate small resistance arteries using wire myography with Fasudil, a rho-kinase inhibitor, and immunohistochemistry. Results Among participants with plasma proteomics available (n=155, mean age: 55 years; 43% female), 72 (47%) had a history of cardiovascular disease. A high likelihood of myocarditis was adjudicated in 12.9% (Very, n=20;Likely, n=63; Unlikely, n=55; Not, n=17). Peak admission troponin (n=192, median=4.00ng/L, IQR=9.0), ranged from 0.5ng/L to 40,404ng/L. Actin and myosin-related pathways were downregulated (Figure 1A). Myosin light (MYL) proteins (MYL3, MYL5, MYL6, MYL6B, MYL9, MYL11, MYL12A, MYL12B), expressed by mainly skeletal and vascular smooth muscle cells (Human Protein Atlas v24.0, Figure 1B), were downregulated at follow-up (Figure 1C). Principal component analysis of actin and myosin proteins (n=209 proteins) showed distinct separation between visits (Figure 1D). At enrolment, differentially expressed MYL-proteins correlated positively with high-sensitivity troponin (n=208, all p0.02) but not with NT-proBNP (n=208). No correlations were observed with myocardial contractility from strain-encoded cardiovascular magnetic resonance imaging or parameters of cardiac remodelling (n=153–155), adjudicated myocarditis status (n=83 (53.5%) Very/Likely), or steroid treatment (n=88). U46619-induced vasoconstriction (Throboxane-A2 analogue) was elevated in COVID-19 arteries (p=0.007) and mitigated by Fasudil. Phosphorylated myosin light chain antibody uptake was greater in COVID-19 arterioles (40.1%) versus controls (10.0%, p0.001) (Figure 1E). Conclusion These findings highlight the complexity of troponin leakage, suggesting an underlying systemic vascular aetiology. We provide novel insights to explain the commonly observed troponin elevations not associated with clinically significant sequelae.