Segmental Strain for Myocardial Scar Detection in Acute Infarcts and Follow-Up CMR Using Non-Contrast Cine Images

Scar tissue from myocardial infarction is best visualized with cardiac magnetic resonance 28 (CMR) late gadolinium enhancement (LGE). Gadolinium-free alternatives for detection of 29 myocardial scars are limited. This study investigated the feasibility of myocardial scar detection 30 in acute infarcts and follow-up CMR using non-contrast cine images. SPCS values in infarcted areas were similar in acute infarcts and in follow-up exams (p=0.8). 54 In acute infarcts 74.6 % of all in LGE infarcted segments (141/189) were correctly localized in 55 polar plot strain maps compared to 44.4% (84/189) of infarcted segments detected by visual 56 wall motion assessment only (p < 0.05). In follow-up exams, 81.5 % of all in LGE infarcted 57 segments (93/114 segments) were correctly localized in polar plot strain maps compared to 58 51.8 % (59/114) of infarcted segments detected by visual wall motion assessment (p < 0.05). be be study analyzed global and segmental strain derived from cine images in acute infarcts and follow-up exams and the feasibility of using segmental circumferential strain detection of myocardial scars.


48
While global strain values were impaired in patients with acute infarcts compared to controls

128
Patients with concomitant primary cardiomyopathies (n= 2) or non-diagnostic LGE images 129 (n= 3) were not enrolled. Twenty-eight individuals (2 female, mean age 48 ± 10 years) with 130 normal cardiac MRI examinations during the same time period were also retrospectively 131 included. CMR referrals in the control group were exclusion of structural heart disease (n= 4) 132 or exclusion of coronary artery disease (n=24). This study was conducted in accordance to 133 the Declaration of Helsinki and its later amendments and the institutional review board 134 approved this retrospective study. All participants gave written informed consent.

135
Characteristics of patient groups and controls are shown in Table 1.

253
Since segmental circumferential strain appeared to be suitable for identifying segments with 254 ischemic scars, we performed ROC analysis to detect the optimal cut-off values for SPCS for

293
In our patient group, mean SPCS in infarcted tissue was impaired compared to SPCS of 294 remote myocardium and this was observed in both acute imaging as well as in follow-up CMRs; 295 in ROC analyses cut-off value was -5.9%, below which segments were considered infarcted.

313
Analyzing values for remote myocardium in acute infarcts, segmental circumferential strain 314 was slightly more impaired compared to remote myocardium in follow-up exams and further 315 analyses revealed, that edematous segments were mainly responsible for strain impairment, 316 suggesting influence of myocardial edema on segmental circumferential strain.

317
In summary, segmental circumferential strain based on non-contrast cine images detects most             Figure 1 1a -48 year old patient with RIVA infarction (2 days after acute infarction) Left column: LGE in segment 8, 13,14 (red arrows); middle column: concomitant edema extends additionally into segments 2,7,16 (white arrows). Right colum: Endo-and epicardially contoured basal, midventricular and apical cine short axis slices prepared for circumferential strain calculations with polar plot strain map. 1b -48 year old patient with RIVA infarction (35 days after acute infarction) Left column: LGE in segment 8, 13,14 (red arrows); middle column: no concomitant edema; right colum: Endo-and epicardially contoured basal, midventricular and apical cine short axis slices prepared for circumferential strain calculations with polar plot strain map.

Figure 2
Global strain values in patients and healthy controls While GPCS, GPLS and GPRS values were very similar comparing both imaging time points, they were signi cantly impaired compared to healthy controls. GPCS = global peak circumferential strain, GPLS = global peak longitudinal strain, GPRS = global peak radial strain ROC curve for distinguishing infarcted and remote myocardium based on strain parameters Below a SPCS value of -5,9 % (sensitivity of 86,2 %, speci city of 83,5%) segments are considered infarcted. ROC= Receiver operating characteristic, SPCS= segmental peak circumferential strain Figure 5 Localization of infarcted segments showed in segmental circumferential strain 548 calculations Segmental strain calculations showed signi cantly more infarcted segments than visual assessment of wall motion abnormalities in cine images and this was signi cant in both imaging time points. In followup exams more infarcted segments were found in visual assessment of wall motion compared to acute infarcts (52% vs. 44.4%). LGE = late gadolinium enhancement, SPCS = segmental peak circumferential strain, VWMA = visual wall motion assessment