The present study substantiates the prognostic implication of CMR based myocardial T2 Mapping in patients with sAMC. This may improve individual risk stratification in this patient cohort. The core findings are: (a) Elevated global and regional T2 values in patients with sAMC predict MACE and rehospitalisation (b) Patients with normal T2 times compared to healthy controls are at low risk for adverse cardiac events (c) T2 values decline with clinical recovery of over time.
Predictors of clinical outcome in acute myocarditis
Several studies reported that native T2 time is increased in patients with clinical diagnosis of acute myocarditis [
8,
9,
19,
20]. Only recently, we demonstrated that T2 Mapping increases diagnostic accuracy for bpAMC [
10].
The presented study demonstrates that measurement of global T2 time provides additional prognostic information and facilitates risk stratification in patients presenting with myocarditis. Kaplan-Meier curves generated among different groups of patients (T2 time <2 SD, >2 SD, <4 SD and >4 SD compared to controls) showed a significantly shorter event-free interval in patients with global T2 time exceeding 4 SD. Furthermore, the percentage of myocardial fraction with T2 time above 80 ms was higher in patients, who reached the combined endpoint. On the other hand patients with normal T2 values were at low risk for following events.
The fact that global T2 time at first presentation was significantly elevated in patients who experienced an adverse outcome (cardiac death, heart transplantation, ventricular assist device implantation) highlights the potential of T2 Mapping strategies to identify patients at risk. Assuming that an acute myocarditis is characterized by widespread cell damage, inflammation and edema, global T2 time yielded a higher odds ratio for predicting combined endpoint than the regional approach did. Furthermore, this study showed that an adverse clinical outcome was associated with the grade of elevation of global T2 time and amount of myocardial area with increased T2 values.
T1 Mapping provides high diagnostic accuracy in confirmation and exclusion of myocarditis and also showed its ability to differentiate between different stages of disease activity [
12,
21]. Recently, Puntmann et al. substantiated the prognostic value of T1 Mapping in a large cohort of patients with non-ischemic dilated cardiomyopathy [
22]. However, a prognostic value of T2 Mapping in acute myocarditis has not been demonstrated yet.
Lurz et al. demonstrated that T1 Mapping did not provide additional diagnostic information in patients with chronic symptoms, which could be explained by histological pathology [
13]. Acute inflammatory responses, such as necrosis, edema and hyperemia, are believed to regress as expansion of the extracellular space due to diffuse fibrosis processes. As a result, T1 relaxation time decreases over time [
21,
23]. It remains questionable, whether T1 Mapping in early stages is solely increased due to edema and hyperemia. If this is the case, T1 Mapping might suffer from decreased sensitivity compared to T2 Mapping, which could explain the lack of prognostic value. However, this remains speculative at the moment.
Recent data also confirm the value of ventricular function as an independent predictor of outcome in patients with acute myocarditis [
4,
24]. Biventricular dysfunction at initial presentation has already been identified as a predictor for death and cardiac transplantation [
25]. Moreover, baseline left ventricular function was a marker of prognosis regardless of the clinical pattern of disease onset [
16]. The presented study yielded similar results (Table
3). The alteration of LVEF at the initial CMR was a strong predictor of outcome.
In addition, presence of LGE is a controversial predictor of outcome in acute myocarditis. Grün et al. investigated a large population of patients with biopsy-proven viral myocarditis who also underwent CMR. In their study LGE was the best predictor for all-cause- and cardiac mortality [
4]. In line with this, we confirm that 91% of patients who reached combined endpoint had presence of LGE on first CMR. Note that only one patient without presence of LGE experienced the endpoint. However specificity was low (29%) rendering individual patient counselling impractical (Table
3). Similarly, Sanguineti et al. investigated 203 consecutive patients with an initial CMR-based diagnosis of acute myocarditis and described an initial alteration of LVEF as the only independent CMR predictor of adverse clinical outcome. In this context LGE was not predictive for outcome [
24].
In a previous published study myocardial edema assessment by T2 weighted Imaging was a strong predictor for left ventricular function recovery [
26]. Vermes et al. hypothesized that observed improvement of systolic function reflects recovery of reversible injured myocardium. Our results show that T2 weighted Imaging was neither sensitive nor specific in predicting clinical outcome. T2 weighted Imaging suffers from several limitations like motion artefacts and low sensitivity, which impair diagnostic accuracy [
7,
19]. Summarizing our results did not reveal a predictive value for T2 weighted Imaging in this patient cohort.
T2 Mapping in the course of disease
In a subgroup analysis (51% of sAMC patients), we accomplished follow-up CMR and demonstrated that T2 time was significantly elevated at initial presentation possibly reflecting acute stage of disease and declined during the healing process of myocarditis (Fig.
3). In detail, we discovered a significant decrease of T2 values in patients who experienced recovery, however being still elevated compared to healthy controls. Recently, Luetkens et al. published similar results in a small patients cohort. [
27] In their study 24 patients with suspected acute myocarditis underwent several CMR examinations during follow-up. CMR markers of myocardial inflammation including T1 and T2 relaxation times demonstrated a rapid and continuous decrease over time. In the presented study, patients with on-going symptoms and persistent cardiac dysfunction displayed persistently elevated T2 values at follow-up, giving rise to a discriminative value for a CMR-based approach to distinguish patients with active myocarditis from healed stages of disease. These results might be explained by a resolution of myocardial edema in the majority of patients at follow-up examination, while those with persistent left ventricular dysfunction may still have suffered from on-going myocardial inflammation.