Persistent long-term subclinical myocardial changes in multisystem inflammatory syndrome in children revealed by cardiac magnetic resonance and galectin-3

Abstract

BACKGROUND

Multisystem inflammatory syndrome in children (MIS-C), a post-infectious complication of severe acute respiratory distress syndrome corona virus-2, commonly presents with acute myocardial dysfunction. Although early recovery is frequently observed, the long-term myocardial health, particularly subclinical changes detectable by cardiac magnetic resonance imaging (CMRI), remains unclear. Biomarkers such as Galectin-3 may offer insight into persistent inflammation or fibrosis.

AIM

To evaluate long-term myocardial structure and function in children following MIS-C using CMRI and two-dimensional echocardiography (2D-ECHO), and to assess serum galectin-3 levels and their correlation with imaging findings.

METHODS

Twelve children diagnosed with MIS-C were prospectively evaluated ≥ 6 months post-illness (mean follow-up: 23.6 months). All participants underwent the same-day 2D-ECHO and 3T CMRI, including cine imaging, T1/T2 mapping, and late gadolinium enhancement. Serum galectin-3 levels were measured by enzyme-linked immunosorbent assay and compared with those of age-matched controls.

RESULTS

2D-ECHO demonstrated a normal ejection fraction (EF) (mean EF: 62.3%) and normal coronary artery Z-scores in all patients. However, CMRI revealed reduced EF (< 55%) in 4/12 (33.3%) patients and low stroke volume in 5/12 (41.7%). Elevated native T1 values, suggestive of diffuse myocardial fibrosis, were observed in 2/12 (16.7%) patients. No late gadolinium enhancement, regional wall motion abnormalities (RWMA), or coronary changes were detected. Serum galectin-3 levels were higher in MIS-C patients compared to controls (13.49 ± 7.63 ng/mL vs 7.48 ± 3.13 ng/mL); however, this difference was not statistically significant.

CONCLUSION

At a mean follow-up of 23.6 months, CMRI identified subclinical myocardial abnormalities in a subset of children with prior MIS-C, including reduced EF (33.3%), low stroke volume (41.7%), and elevated native T1 values (16.7%), despite normal findings on 2D-ECHO. These findings suggest greater CMRI sensitivity than 2D-ECHO for long-term myocardial assessment. Although serum galectin-3 levels were higher in patients compared to controls, the difference was not statistically significant, and its clinical relevance remains to be established. Given the small sample size, larger studies are required to confirm these observations.

Key Words: Multisystem inflammatory syndrome in children; Echocardiography; Cardiac magnetic resonance imaging; Serum galectin; Myocardium

Core Tip: Children recovering from multisystem inflammatory syndrome in children may appear to have complete cardiac recovery on routine echocardiography, yet subtle myocardial abnormalities can persist undetected. Cardiac magnetic resonance imaging performed nearly two years post-illness revealed reduced ejection fraction and elevated native T1 values in a subset of patients, suggesting subclinical myocardial dysfunction. Although serum galectin-3 levels were mildly elevated, they did not correlate with imaging markers. These findings highlight the importance of incorporating cardiac magnetic resonance imaging into long-term follow-up to identify silent myocardial changes and guide future surveillance strategies in multisystem inflammatory syndrome in children survivors.

INTRODUCTION

Multisystem inflammatory syndrome in children (MIS-C) is a post-infectious hyperinflammatory condition temporally associated with severe acute respiratory distress syndrome corona virus-2 (SARS-CoV-2) infection, first recognized in April 2020 during the coronavirus disease 2019 pandemic[1]. MIS-C typically manifests 2-6 weeks after SARS-CoV-2 exposure and shares clinical features with Kawasaki disease (KD), toxic shock syndrome, and macrophage activation syndromes[2,3]. It is characterised by a cytokine storm resulting in multisystem involvement and frequent cardiovascular dysfunction[2,3].

Cardiac involvement is a major contributor to morbidity in MIS-C, with approximately 80% of affected children showing manifestations such as ventricular dysfunction, valvulitis, pericardial effusion, coronary artery abnormalities (CAAs), and arrhythmias[4]. Studies using 2D-echocardiography (2D-ECHO) have shown high rates of left ventricular (LV) systolic and diastolic dysfunction during the acute phase, with improvement noted on follow-up[5,6]. However, 2D-ECHO may not detect subclinical or diffuse myocardial injury.

Cardiac magnetic resonance imaging (CMRI) provides a more sensitive and detailed evaluation of myocardial structure and function. Advanced CMRI techniques, including T1 and T2 mapping and late gadolinium enhancement (LGE), can detect myocardial edema, inflammation, and fibrosis[5]. Although LGE is uncommon in MIS-C, persistent subclinical myocardial changes have been reported. The long-term cardiac sequelae of MIS-C remains unclear, particularly in children who appear clinically recovered[7].

Emerging evidence suggests that serum biomarkers, such as galectin-3 - a mediator of myocardial fibrosis and inflammation - may reflect underlying cardiac pathology. Elevated levels have been reported in conditions like KD with adverse cardiovascular outcomes[8].

Although short-term echocardiographic recovery following MIS-C has been well documented, data on long-term subclinical myocardial changes are limited, particularly using advanced tissue characterisation techniques. Furthermore, the role of fibrosis biomarkers such as galectin-3 in MIS-C remains largely unexplored. This study uniquely combines long-term 3T CMRI with serum galectin-3 assessment to evaluate persistent myocardial changes in clinically recovered children. Additionally, we sought to compare CMRI findings with 2D-ECHO to evaluate the utility of multimodal cardiac assessment in this population.

MATERIALS AND METHODS

This prospective observational study was conducted in the Allergy Immunology Unit, Advanced Pediatrics Centre, Postgraduate Institute of Medical Education and Research, Chandigarh, between January 2022 and December 2023. Twelve children previously diagnosed with MIS-C, based on World Health Organization criteria, were enrolled at least six months after diagnosis[9]. Exclusion criteria included the need for sedation for MRI, contraindications to magnetic resonance imaging (MRI) (e.g., non-cooperative children requiring sedation, non-compatible implants, renal dysfunction with serum urea > 35 mg/dL or creatinine > 0.8 mg/dL), or lack of consent. The study was approved by the Institute’s thesis committee and Institutional Ethics Committee, approval No. INT/IEC/2022/MD-454. Written informed consent and assent were obtained from parents and patients, respectively.

Data collection

Demographic and clinical details were recorded using a standardised proforma. All patients underwent both 2D-ECHO and CMRI to assess cardiac structure and function, with a particular focus on CAAs and myocardial tissue characterisation.

CMRI

CMRI was performed using a 3T scanner (Magnetom Vida, Siemens Healthineers, Erlangen, Germany) using a phased-array surface coil. Imaging protocols included cine sequences, LGE, and T1/T2 mapping (Siemens MyoMaps), acquired during breath-hold with electrocardiogram gating. Parameters analyzed included LV volumes, mass, ejection fraction (EF), segmental wall thickness (per American Heart Association 17-segment model), coronary artery anatomy, wall motion, presence of thrombus, delayed enhancement, and native T1/T2 values. All images were interpreted by an experienced cardiac radiologist blinded to the clinical data.

2D-ECHO

2D-ECHO was performed using the Philips EPIQ 7 system on the same day as CMRI by a cardiologist blinded to clinical details and CMRI findings. Assessed parameters included EF, M-mode findings, regional wall motion, valvular anomalies, and CAAs. CAAs were classified based on Z-scores, with aneurysms categorised by luminal diameter, including criteria for giant aneurysms[10].

Serum galectin-3 measurement

Serum galectin-3 levels were measured in all patients and age-matched healthy controls using a commercially available enzyme-linked immunosorbent assay kit (R&D Systems, United States; Cat. No. DGAL30), according to the manufacturer’s instructions.

Blood samples were collected under aseptic conditions, allowed to clot at room temperature, and centrifuged to separate serum. Samples were aliquoted to avoid repeated freeze–thaw cycles and stored at -80 °C until analysis. Patient and control samples were processed and analysed simultaneously using identical protocols to minimise analytical variability.

All samples and standards were assayed in duplicate. Optical density was measured using a microplate enzyme-linked immunosorbent assay reader (Tecan, Switzerland), operated with appropriate wavelength settings as per the assay recommendations. Calibration and quality control checks were performed in accordance with standard laboratory protocols. Galectin-3 concentrations were calculated from a standard calibration curve and expressed in ng/mL.

Statistical analysis

Data were compiled using Microsoft Excel and analysed in SPSS v23 (IBM Corp.). Continuous variables were expressed as mean ± SD or median with interquartile range (IQR), as appropriate, while categorical variables were presented as frequencies and percentages. Comparisons between groups were performed using paired t-tests or Wilcoxon signed-rank tests, depending on data distribution. Correlations were assessed using Pearson’s or Spearman’s correlation coefficients. A P value < 0.05 was considered statistically significant.

RESULTS

Patient demographics and clinical characteristics

A total of 12 patients with MIS-C (8 boys; 4 girls) were enrolled, with a mean age at enrolment of 11.33 ± 1.37 years. The mean age at diagnosis was 9.50 ± 0.90 years [median (IQR): 10 (9.00-10.00)], and the mean interval between diagnosis of MIS-C and cardiac imaging was 23.58 ± 5.65 months. All 12 (100%) patients presented with fever. Gastrointestinal symptoms were observed in 9 (75%) patients, including abdominal pain in 3 (25%), vomiting in 4 (33.3%), and loose stools in 2 (16.7%). Hypotension was present in 7 (58.3%) patients. Respiratory symptoms were noted in 6 (50%) patients, including cough in 5 (41.7%) and respiratory distress in 3 (25%). Rash was present in 7 (58.3%), conjunctival injection in 5 (41.7%), skin peeling in 3 (25%), and oral mucosal changes in 1 (8.3%). Seven (58.3%) patients developed cardiogenic shock requiring vasoactive support: 3 (25%) required a single inotrope, 3 (25%) required two inotropes, and 1 (8.3%) required more than two inotropes (Table 1). The mean EF ± SD at presentation was 46.25% ± 11.31% of these 7 patients. The remaining 5 (41.7%) did not require vasoactive support. All patients received intravenous immunoglobulin (2 g/kg) and methylprednisolone (30 mg/kg for 3-5 days). At presentation, 3 (25%) patients had CAAs, including a medium-sized right coronary artery (RCA) aneurysm (> 5 Z-score) in 1 (8.3%), a small left main coronary artery (LMCA) aneurysm in 2 (16.7%), and left anterior descending coronary artery (LAD) dilatation (2.5-5 Z-score) in 1 (8.3%). One patient (8.3%) had aneurysms involving both the RCA and LMCA (Table 2).

Table 1 Clinical features observed during acute presentation of multisystem inflammatory syndrome in children (n = 12).

Clinical characteristics		n (%)
Fever		12 (100.0)
GI manifestations	Abdominal pain	3 (25.0)
	Vomiting	4 (33.3)
	Loose stools	2 (16.7)
Cough		5 (41.1)
Respiratory distress		3 (25.0)
Conjunctival injection		5 (41.7)
Oral mucosal changes		1 (8.3)
Rash		7 (58.3)
Seizure		1 (8.3)
Skin peeling		3 (25.0)
Hypotension		7 (58.3)

GI: Gastrointestinal.

Table 2 Coronary artery dimensions and left ventricular ejection fraction assessed by 2D-echocardiography during acute presentation of multisystem inflammatory syndrome in children.

LMCA diameter in mm (Z score)	LAD diameter in mm (Z score)	LCx diameter in mm (Z score)	RCA diameter in mm (Z score)	EF (%) (low: < 55%)
1.91 mm (-0.46Z)	2.12 mm (+0.03Z)	NA	2.1 mm (+0.21Z)	52
2.8 mm (-0.12Z)	2 mm (-0.5Z)	2.2 mm (-0.59Z)	2.4 mm (-0.33Z)	55
2.43 mm (+0.76Z)	2.01 mm (+0.24Z)	NA	1.88 mm (-1.41Z)	30
2.89 mm (-0.45Z)	2.45 mm (+0.13Z)	1.86 mm (-0.82Z)	2.37 mm (-0.85Z)	40
5.2 mm (+4.15Z)	3 mm (1.17Z)	NA	5 mm (+5.49Z)	35
2.5 mm (+0.3Z)	2 mm (+0.5Z)	1.5 mm (-0.63Z)	2.3 mm (+0.49Z)	50
2.8 mm (-0.62Z)	2.6 mm (+0.5Z)	1.01 mm (-2.02Z)	2.98 mm (+0.43Z)	55
3.3 mm (+2.96Z)	1.6 mm (-1.11Z)	NA	2.6 mm (+0.58Z)	35
2.30 mm (-1.5Z)	1.94 mm (-0.86Z)	1.86 mm (-0.97Z)	2.33 mm (-0.72Z)	60
2.6 mm (-1.21Z)	2.4 mm (-0.11Z)	2.1 mm (-0.63Z)	2.5 mm (-0.71Z)	55
2.17 mm (-0.75Z)	1.98 mm (+0.18Z)	2.11 mm (+0.98Z)	2.47 mm (+0.64Z)	55
3.3 mm (+1.94Z)	3.1 mm (+3.86Z)	2.6 mm (+1.62Z)	2.8 mm (+2.3Z)	30

Z-scores were calculated using Dallaire’s equations. Ejection fraction < 55% was considered reduced. EF: Ejection fraction; RCA: Right coronary artery; LMCA: Left main coronary artery; LAD: Left anterior descending artery; LCx: Left circumflex artery; NA: Not available.

Laboratory parameters at the time of presentation in patients with MIS-C

All patients underwent evaluation of inflammatory parameters during the acute phase of MIS-C. The median total leukocyte count among these children was 7.47 × 10⁹/L (range 1.64-14.4 × 10⁹/L). Thrombocytopenia (< 150 × 10⁹/L) was observed in 8 patients. mean ± SD N-terminal pro-B-type natriuretic peptide levels were 3189 ± 9640 pg/mL. mean ± SD C-reactive protein levels were 170.68 ± 124.64 mg/L. All patients (12/12; 100%) had elevated N-terminal pro-B-type natriuretic peptide and CRP levels. Erythrocyte sedimentation rate values were available in 6 patients, of whom 5 had elevated levels.

2D-ECHO findings

At enrollment (follow-up), all patients underwent 2D-ECHO (Table 3). Mean coronary artery Z-scores were -0.09 ± 1.28 for RCA, -0.17 ± 1.12 for LMCA, 0.33 ± 0.91 for LAD, and 0.00 ± 1.03 for the left circumflex (LCx) artery. Coronary Z-scores were within normal limits in all patients. The parasternal long-axis measurements, including the interventricular septum, LV, and posterior wall, were within normal limits in both end-diastole and end-systole. The mean EF was 62.33% ± 5.28%, and the mean fractional shortening was 34.83% ± 2.72%. EF was normal (> 55%) in all patients. Valvular abnormalities were observed in 3 (25%) patients, including trivial tricuspid regurgitation in 3 (25%) and trivial mitral regurgitation in 1 (8.3%). One patient (8.3%) had both tricuspid regurgitation and mitral regurgitation.

Table 3 Coronary artery dimensions assessed by 2D-echocardiography at follow-up enrolment in patients with multisystem inflammatory syndrome in children.

RCA diameter in mm (Z score)	LMCA diameter in mm (Z score)	LAD diameter in mm (Z score)	LCx diameter in mm (Z score)
2.7 mm (0.05Z)	3.2 mm (0.48Z)	2.5 mm (0.47Z)	2.2 mm (0.22Z)
2.9 mm (0.21Z)	3 mm (-0.2Z)	2.4 mm (-0Z)	2.2 mm (-0.3Z)
2.5 mm (-0Z)	2.6 mm (-0.5Z)	2.1 mm (-0.2Z)	2.7 mm (-0.6Z)
1.36 mm (-3.2Z)	2.5 mm (-1.7Z)	2.3 mm (-0.6Z)	2.1 mm (-0.8Z))
2.2 mm (-1.3Z)	2.5 mm (-1.5Z)	2.4 mm (-0.1Z)	2.6 mm (0.6Z)
2.1 mm (-0.5Z)	2.4 mm (-0.5Z)	2.3 mm (0.7Z)	1.6 mm (0.8Z)
3.8 mm (1.54Z)	4 mm (1.35Z)	3.3 mm (1.6Z)	3.1 mm (1.03Z)
3 mm (1.22Z)	2.7 mm (1.66Z)	2.8 mm (1.7Z)	2.8 mm (1.59Z)
3.2 mm (0.73Z)	3.2 mm (0.09Z)	3 mm (1.29Z)	2.5 mm (0.15Z)
2.85 mm (-0.5Z)	2.96 mm (-1Z)	2.33 mm (-0.7Z)	1.59 mm (-1.9Z)
2.6 mm (-0.2Z)	2.86 mm (-1.3Z)	2.7 mm (-1Z)	1.8 mm (-1.4Z)
2.7 mm (0.98Z)	3 mm (1.06Z)	2.3 mm (0.84Z)	2.2 mm (0.63Z)

All coronary Z-scores were within normal range at follow-up. RCA: Right coronary artery; LMCA: Left main coronary artery; LAD: Left anterior descending artery; LCx: Left circumflex artery.

CMRI findings

CMRI was performed in all 12 patients at a mean follow-up of 23.6 months (range: 12-32 months) (Table 4). A reduced EF (< 55%) was observed in 4/12 (33.3%) patients. The median EF on CMRI was 58.0% (range, 47%-70%). End-diastolic volume ≤ -2 Z-score (adjusted for body surface area) was observed in 5 (41.7%) patients, while end-systolic volume ≤ -2 Z-score was noted in 1 (8.3%) patient. Reduced stroke volume (≤ -2 Z-score) was present in 5 (41.7%) patients, and low cardiac output was observed in 2 (16.7%) patients.

Table 4 Cardiac magnetic resonance imaging parameters at follow-up (mean 23.6 months) in children with multi-system inflammatory syndrome in children.

Parameters on MRI	mean ± SD, median (IQR)	Remarks
EDV (mL)	65.80 ± 16.07	-
	64.85 (51.62-76.33)
EDV Z scores	-1.72 ± 0.73	5 (42%) patients had EDV ≤ 2 Z score as per BSA
	-1.91
ESV (mL)	26.43 ± 5.92	-
	26.75 (20.4-30.25)
ESV Z scores	-0.56 ± 0.65	1 (8.3%) patient had ESV ≤ 2 Z score as per BSA
	-0.32
SV (mL)	41.27 ± 15.19	-
	33 (30.05-54.87)
SV Z scores	-1.52 ± 0.83	7 (58%) patients had SV ≤ 2 Z score as per BSA
	-2.02
CO (L/minute/m2)	3.54 ± 1.46	7 (58%) patients had CO lower than normal range (3.5-5 L/minute/m2)
	2.97 (2.42-4.8)
EF	58.34 ± 7.75	4 (33%) patients had EF < 55%
	58 (52.25-63.77)
T1 values (millisecond)	1173 ± 115	2 (16%) patients had T1 values > 1240 normal range (1021-1240)
	1195 (1171.25-1211.75)
T2 values (millisecond)	37.9 ± 2.64	-
	37 (35.50-40.00)
RWMA	-	All had no RWMA
LGE	-	No LGE were detected in any patients

MRI: Magnetic resonance imaging; IQR: Interquartile range; EDV: End diastolic volume; ESV: End systolic volume; SV: Stroke volume; CO: Cardiac output; BSA: Body surface area; EF: Ejection fraction; RWMA: Regional wall motion abnormality; LGE: Late gadolinium enhancement.

Tissue characterisation revealed elevated native T1 mapping values in 2/12 (16.7%) patients (Figure 1). No LGE was detected in any patient, suggesting diffuse myocardial fibrosis rather than focal fibrosis or necrosis.

Figure 1 Cardiac magnetic resonance imaging short-axis views at mid-ventricular level in a representative patient at follow-up. A: Late gadolinium enhancement image showing no myocardial enhancement; B: T2 parametric map demonstrating normal myocardial T2 values, indicating the absence of active myocardial edema; C: Native T1 parametric map demonstrating elevated T1 values, suggesting diffuse myocardial changes. These findings illustrate subclinical myocardial abnormalities detected on cardiac magnetic resonance imaging despite normal echocardiographic findings.

No RWMA were observed. Coronary artery assessment on CMRI showed no evidence of coronary artery dilatation, aneurysm, stenosis, or thrombosis. No structural abnormalities were identified in any patient.

Correlation analysis

A significant negative correlation was observed between EF (%) on 2D-ECHO at presentation and pro-BNP levels at presentation (ρ = -0.72, P = 0.008). In contrast, EF (%) measured by CMRI at follow-up showed no significant correlation with pro-BNP levels at presentation (ρ = -0.13, P = 0.696).

Serum galectin-3

Serum galectin-3 levels were higher in the patients compared with the controls [mean ± SD: 13.49 ± 7.63 ng/mL vs 7.48 ± 3.13 ng/mL; median (IQR): 11.00 (9.96-13.05) vs 6.54 (5.03-11.09) ng/mL]. However, the difference did not reach statistical significance (P = 0.063) (Figure 2).

Figure 2 Comparison of serum galectin-3 levels (ng/mL) in patients with prior multisystem inflammatory syndrome in children at a mean follow-up of 23.58 months and in healthy controls. Although galectin-3 levels were higher in the multisystem inflammatory syndrome in children group, the difference did not reach statistical significance (P = 0.063).

DISCUSSION

MIS-C is a severe hyperinflammatory condition temporally associated with SARS-CoV-2, characterised by prominent cardiovascular involvement, including myocardial dysfunction and CAAs, especially during the acute phase[11]. While cardiac dysfunction in the acute stage is well-documented, data on long-term myocardial health, especially when assessed using CMRI, remain limited[6].

In this study, we present one of the longest follow-up assessments of myocardial health in children with prior MIS-C. A total of 12 patients were evaluated using both 3T CMRI and 2D-ECHO at a mean follow-up of 23.58 ± 5.65 months. The mean age at diagnosis was 9.50 ± 0.90 years, and the mean age at enrolment was 11.33 ± 1.37 years. During the acute phase, 7/12 (58.3%) patients presented with shock and reduced EF, requiring vasoactive support; the mean EF in these patients was 47%. Pro-BNP levels were elevated in all patients at presentation and demonstrated a significant negative correlation with EF at presentation measured by 2D-ECHO (ρ = -0.72, P = 0.008).

CAAs were observed in 3 (25%) patients during the acute phase, involving the RCA, LMCA, and LAD. All patients were treated with intravenous immunoglobulin (2 g/kg) and high-dose intravenous methylprednisolone (30 mg/kg/day for 3-5 days), followed by tapering oral steroids. At follow-up, both 2D-ECHO and CMRI demonstrated complete resolution of the CAAs. No coronary artery dilation, aneurysm, thrombosis, or stenosis was observed in any patient on either of the imaging modalities.

Follow-up 2D-ECHO demonstrated normal EF (> 55%) in all patients (mean EF 62.33% ± 5.28%), with no RWMA. Mild valvular abnormalities were observed in 3 patients (25%), including trivial MR, TR, or both. Notably, EF at presentation did not correlate with EF at long-term follow-up.

In contrast, CMRI at the same follow-up interval revealed a reduced EF (< 55%) in 4/12 (33.3%) patients (mean CMRI EF 58.34%; range 47%-70%). Notably, all 4 patients had normal EF on 2D-ECHO, suggesting that CMRI may be more sensitive in detecting subtle residual myocardial dysfunction. No RWMA or LGE was identified; however, elevated native T1 values (> 1240 milliseconds) were observed in 2/12 (16.7%) patients, consistent with diffuse myocardial changes. Reduced stroke volume and cardiac output were also noted in a subset of patients on CMRI. Importantly, all children were clinically stable and asymptomatic at follow-up, indicating that these imaging abnormalities likely represent subclinical myocardial alterations rather than overt cardiac dysfunction.

The differences between 2D-ECHO and CMRI-derived EF may be attributed to methodological differences between the two modalities. CMRI provides an accurate volumetric assessment of ventricular function without geometric assumptions and is considered the reference standard for quantifying LV volumes and EF, whereas 2D-ECHO EF is operator-dependent and relies on geometric modelling[12]. These findings support the potential role of CMRI in detecting subclinical myocardial changes that may not be captured by 2D-ECHO alone.

Galectin-3, a biomarker linked to myocardial fibrosis and inflammation, was also evaluated. The mean (SD) serum galectin-3 levels in patients were 13.49 ± 7.63 ng/mL compared to 7.48 ± 3.13 ng/mL in controls, and the median (IQR) levels were 11 (9.96-13.05) vs 6.54 (5.03-11.09), respectively. Although levels were higher in the study group, the difference did not reach statistical significance (P = 0.063). No significant correlation was observed between galectin-3 levels and ventricular function parameters on either modality. Despite the lack of correlation with EF on imaging, the persistent elevation in galectin-3 levels, nearly 2 years post-illness, may indicate ongoing myocardial remodelling or low-grade inflammation. This biochemical signal, when interpreted in conjunction with subtle CMRI findings such as elevated native T1, provides preliminary evidence of potential ongoing myocardial processes, even in clinically asymptomatic children[13].

The clinical features observed in our cohort during the acute phase were comparable to those reported in previous studies[14]. Fever was present in all patients (100%), with gastrointestinal symptoms in 75%, hypotension in 58.3%, rash in 58.3%, and respiratory symptoms in 50%. These findings are consistent with those of Pouletty et al[14], who reported similar patterns of multisystem involvement in MIS-C. During the acute phase, patients showed elevated Pro-BNP values (median 4472.5 pg/mL), reflecting significant myocardial strain.

Our findings are consistent with prior short- to mid-term studies. Belhadjer et al[6] reported persistent LV dysfunction in 5/35 patients with MIS-C, despite early recovery in most cases. Similarly, Kavurt et al[15] documented residual diastolic dysfunction at discharge in patients who had reduced EF at admission. Capone et al[7] reported functional normalisation by 6 months but did not use CMRI, in contrast with Benvenuto et al[16], who identified pericardial effusion and myocardial edema in approximately 5% of patients at 3 months using CMRI[7,16]. Similarly, Varadarajan et al[17] evaluated 270 MIS-C patients and reported early cardiac involvement on 2D-ECHO, including LV dysfunction (29.3%) and coronary aneurysms (28.5%), with near-complete resolution on follow-up up to one year; however, advanced imaging modalities such as CMRI were not utilized. Our findings extend these observations to a longer follow-up period and highlight the need for continued cardiac surveillance beyond the early recovery period in patients with MIS-C.

While 2D-ECHO remains a valuable first-line tool, CMRI offers superior spatial resolution, detailed tissue characterisation, and a radiation-free approach for assessing myocardial fibrosis, edema, and perfusion abnormalities[18]. Native T1 mapping, in particular, has emerged as a sensitive biomarker for diffuse myocardial injury[19], and its elevation in our cohort supports its role in the long-term evaluation of children with prior MIS-C.

Galectin-3 has been proposed as a biomarker of chronic cardiac inflammation and fibrosis, with established utility in adult heart failure and KD[8]. Although its role in MIS-C remains unclear, our findings suggest that galectin-3 levels may remain elevated months after recovery, potentially reflecting ongoing myocardial remodelling. However, given the small sample size, these findings should be considered exploratory. The lack of correlation with ventricular function further underscores the need for additional studies to clarify its clinical utility.

Similar to follow-up strategies used in KD, children recovering from MIS-C with significant acute cardiac involvement may benefit from individualised long-term follow-up[20,21]. Although the observed subclinical reductions in stroke volume, cardiac output, and EF were not associated with clinical symptoms at follow-up, these findings warrant continued surveillance. Longitudinal follow-up will be essential to better understand the long-term myocardial consequences of MIS-C and to inform future monitoring and management strategies.

Key strengths of our study include the extended follow-up duration, comprehensive dual-modality cardiac imaging, and the incorporation of a fibrosis biomarker (galectin-3). However, we also identify our limitations. The small sample size and single-centre design may limit statistical power and the generalizability of the findings. Excluding children who require sedation for MRI may have introduced selection bias. Emerging studies indicate that extracellular volume estimation is more sensitive than LGE for detecting subclinical myocardial fibrosis and correlates well with the degree of fibrosis. extracellular volume mapping was not performed in this study, limiting definitive characterisation of fibrosis. Although native T1 mapping was included, T1 values are influenced by magnetic field strength and scanner-specific acquisition parameters. Accordingly, locally derived reference values for our 3T system (1200 ± 60 milliseconds) were applied.

Larger, multicenter studies incorporating serial CMRI and biomarker evaluation are warranted to validate these findings and guide long-term management strategies.

CONCLUSION

At a mean follow-up of 23.6 months in children with prior MIS-C, CMRI identified persistent subclinical myocardial abnormalities in a notable subset of patients, despite normalisation of LV systolic function and coronary findings on 2D-ECHO. CMRI detected reduced EF (< 55%) in 33.3% of patients, low stroke volume in 41.7%, and elevated native T1 values in 16.7%, suggesting subtle myocardial changes not apparent on 2D-ECHO. No LGE, coronary artery aneurysms, stenosis, or thrombosis were observed on MRI. Serum galectin-3 levels were higher in patients than in controls; however, they did not correlate with imaging parameters, and their clinical significance remains to be established. These findings support the potential role of CMRI as a sensitive modality for long-term cardiac assessment in selected children with prior MIS-C. Larger multicenter studies are required to confirm these observations and determine their clinical relevance.

ACKNOWLEDGEMENTS

The author gratefully acknowledges technical contributions and support from Siemens Healthineers, India.