Abstract
Introduction: The diagnosis of myocarditis is challenging and often imprecise despite the availability of contemporary diagnostics such as cardiovascular magnetic resonance (CMR) imaging. This study aimed to detect associations between baseline clinical characteristics and CMR features in patients with acute myocarditis.
Methods: CMR reports from the Barts BioResource were interrogated with a natural language processing (NLP) algorithm to identify individuals with elevated myocardial T2 signal. Cases not meeting the clinical definition of acute myocarditis were excluded. Data on age, sex, peak troponin T, peak CRP, and symptoms were collected. Multivariate analyses evaluated associations between baseline variables and CMR outcomes, specifically left ventricular ejection fraction (LVEF) and the extent of late gadolinium enhancement (LGE).
Results: A total of 127 patients were identified. Peak troponin was independently associated with higher LGE (beta=1.90, P = 0.029) and lower LVEF (beta=-6.96%, P = 0.02). Only 18% of patients had a preceding viral syndrome. Age distribution was bimodal for women.
Conclusions: Peak troponin levels are independently associated with markers of adverse CMR outcomes, highlighting its prognostic role in acute myocarditis.
Introduction
Myocarditis, an inflammation of the heart muscle, is often described as a "clinical chameleon" due to its diverse and frequently non-specific presentation. While Cardiovascular Magnetic Resonance (CMR) imaging has emerged as the gold standard for non-invasive diagnosis—enabling the detection of myocardial edema and scarring—predicting the severity of injury from initial clinical data remains a significant challenge. This study investigates how baseline clinical characteristics and blood biomarkers correlate with structural and functional damage visible on CMR scans.
Methods
The researchers conducted a retrospective study at the Barts BioResource, a large tertiary referral center. To efficiently identify relevant cases, they employed a Natural Language Processing (NLP) algorithm to screen thousands of CMR reports for "elevated myocardial T2 signal," a hallmark of acute tissue edema.
Clinical data, including age, sex, body mass index (BMI), and symptoms, were collected alongside key biomarkers: peak high-sensitivity Troponin T (a marker of myocardial damage) and C-reactive protein (CRP). The primary imaging outcomes were Left Ventricular Ejection Fraction (LVEF) and the extent of Late Gadolinium Enhancement (LGE), quantified by the number of affected segments. Multivariate regression analyses were performed to identify independent predictors of these CMR findings, adjusting for age, sex, and time to scan.
Results
The study cohort comprised 127 patients with confirmed acute myocarditis. The majority were male (64%, mean age 36 years), while female patients exhibited a distinct bimodal age distribution.
Key findings from the clinical and imaging analysis include:
- Symptomatology: While 93% of patients presented with chest pain or shortness of breath, only 18% reported a preceding viral syndrome, suggesting that the absence of a recent infection should not rule out myocarditis.
- Imaging Prevalence: LGE was found in 93% of patients, with 98% showing a non-ischaemic pattern.
- Troponin as a Quantitative Marker: Peak troponin levels were strongly associated with the severity of myocardial injury. Patients in the highest tertile of peak troponin had significantly more LGE segments (beta=1.90) and a lower LVEF (beta=-6.96%) compared to those in the lowest tertile.
Discussion and Conclusion
This research demonstrates that peak troponin is not merely a diagnostic indicator but a valuable quantitative proxy for the extent of myocardial damage in acute myocarditis. The strong correlation between troponin levels and both functional impairment (LVEF) and tissue scarring (LGE) underscores its prognostic utility during the acute phase of the disease.
Furthermore, the study highlights the clinical heterogeneity of myocarditis. The low prevalence of preceding viral symptoms and the bimodal age distribution in women suggest that the triggers and pathophysiology of the disease may vary significantly across the patient population. These findings emphasize the need for better patient stratification and further research into long-term outcomes to improve the management of this complex condition.
Summary updated