• cardiac mri
• heart rupture
• infarcted wall thickness
• infarcted/remote thickness ratio
• intramyocardial hemorrhage
• microvascular obstruction
• no-reflow
• pericardial effusion

BACKGROUND Heart rupture is a rare but devastating complication of acute myocardial infarction. Acute rupture causes cardiac tamponade and death within minutes, subacute rupture results in slower bleeding provoking hypotension until cardiogenic shock, chronic rupture is covered by epicardial thrombus or pericardial adhesions which prevent rupture expansion even for more than 2 weeks, allowing multimodal imaging for correct diagnosis and multidisciplinary decision-making for treatment. The pathophysiological basis of heart rupture is represented by coronary no-reflow, possibly associated with intramyocardial hemorrhage. Coronary no-reflow is defined as the failure to restore myocardial tissue reperfusion by coronary microcirculation despite restoration of epicardic coronary blood flow. Microvascular obstruction is the underlying pathophysiological mechanism of coronary no-reflow and it depends on three main factors: myocardial ischemia, spontaneous or iatrogenic distal embolization and reperfusion-related injury. All these mechanisms, together, weaken the infarcted myocardial tissue and predispose to heart rupture. The most sensitive technique to detect microvascular obrstruction and intramyocardial hemorrhage and diagnose coronary no-reflow is cardiac magnetic resonance imaging (MRI). Microvascular obstruction is identified as a region of persistent hypoenhancement within the core of the infarcted myocardium on late gadolinium enhancement images. Intramyocardial hemorrhage appears as a hypointense region within the infarct on T2-weighted sequences. Furthermore cardiac MRI provides excellent diagnostic capacity in detecting mechanical complications in patients with myocardial infarction such as heart rupture, pericardial complications and intracavitary and epicardial thrombosis. The aim of our retrospective study was to identify the clinical and imaging features associated with the incidence of covered/impending heart rupture assessed by cardiac MRI, in order to identify potential predictors that could guide clinical suspicion and subsequent diagnostic confirmation through cardiac MRI.

METHODS Cardiac MRI examinations performed within the first week from ST-elevation acute myocardial infarction in consecutive patients from january 2018 to november 2024 were retrospectively evaluated. Univariate and multivariable logistic regression analysis were used to explore the impact of each significant variable in univariate analysis to predict the presence of covered\ impending heart rupture. Sensitivity, specificity, area under the curve (AUC), positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy to predict heart rupture were calculated for each significant parameter.

RESULTS The whole population included 205 patients with recent myocardial infarction. Overall, 26 (12%) patients had MRI evidence of covered\impeding heart rupture (Rupture group). There were no significant differences of infarcted territory between patients with and without rupture. Patients of Rupture-group had lower median LV EF (p=0.002) and presented more frequently LV dysfunction (73% vs 46%, p=0.009). Patients of Rupture group had more frequently pericardial effusion (88% vs 15%, p<0.0001). Interestingly, the thickness of infarcted myocardium, measured at the day of CMR, was higher in Rupture group than in others (p<0.0001) as well as the difference and the ratio of thickness between infarcted and remote myocardial segment (p<0.0001). MRI showed also that the LV mass index was greater in patients of Rupture group than in others (p = 0.003). All the patients of Rupture group had microvascular obstruction and haemorrhagic infarction whereas they were found with significantly lower prevalence in other patients (respectively 39% and 26%). Intracavitary LV thrombosis was found in 48% of patients of Rupture group and in 13% of other patients (p<0.0001). Among patients of Rupture group, 6 (24%) had ventricular pseudo-aneurysm.
At the univariate logistic regression analysis for predicting the endpoint of covered\impending heart rupture, LV dysfunction, LV mass index, thickness of the infarcted segment, an infarcted\remote thickness ratio >1, microvascular obstruction, LV thrombus and pericardial effusion were associated with the endpoint. However, the parameter with the greatest OR was the combination of pericardial effusion + thickness of the infarcted segment >12 mm + infarcted\remote thickness ratio >1. These parameters were tested in a multivariable logistic model (table 4). In this model microvascular obstruction (OR 27, 95% CI 1.7-416, p = 0.02) and the combination of pericardial effusion + thickness of the infarcted segments >12 mm + infarcted\remote thickness ratio >1 (OR 263, 95% CI 28-2512, p <0.0001) were independent predictors of heart rupture. Considering separately all the parameters, thickness of the infarcted segment >12 mm (AUC 0.84), pericardial effusion (AUC 0.87) and haemorrhagic infarction (AUC 0.87) were the best predictors of covered\impeding heart rupture. Taken separately all these parameters are characterized by great NPV (ranging from 93 to 100%) but lower PPV (Ranging from 19 to 47%). The combination of parameters of pericardial effusion + thickness of the infarcted segment >12 mm + infarcted\remote thickness ratio >1 had 0.90 AUC and 0.97 diagnostic accuracy to predict covered\impending heart rupture maintaining a great NPV (97%) and also increasing the PPV (91%).

CONCLUSIONS The present study demonstrates that covered/impending heart rupture is associated with the presence of pericardial effusion along with increased or non-thinned infarcted wall thickness due to haemorrhagic infarction. Future studies should clarify whether these parameters could also be used with echocardiography to better select recent myocardial infarction patients for CMR before discharge.