Temporal changes in regional function, edema and myocardial perfusion in Takotsubo syndrome

Background: Takotsubo syndrome (TSS) is a common differential diagnosis in case of myocardial infarction with non-obstructive coronary arteries (MINOCA). A previous study identified abnormal myocardial perfusion during the acute phase of TSS¹, as hyperemia in functionally unaffected segments. Currently, knowledge about the temporal trends of myocardial perfusion, together with edema and cardiac systolic dysfunction is limited. Therefore, this study aimed to explore the temporal trends of myocardial perfusion, edema and systolic dysfunction in TSS.

Methods: We prospectively included patients who received a final diagnosis of TSS following evaluation for MINOCA. Patients were scanned at 1.5T (MAGNETOM Sola; Siemens Healthineers, Forchheim, Germany) within 4 days of hospital admission and after 6 months, using the same imaging protocol. We used ECG-gated modified Look-Locker inversion recovery (MOLLI) 5s(3s)3s research sequence² for native T1 and T2-prepared application (MyoMaps, Siemens Healthineers, Forchheim, Germany) sequence for native T2 edema imaging. Extracellular volume maps were calculated by using native and postcontrast T1 maps, calibrated by hematocrit. Quantitative, first-pass perfusion imaging was performed during adenosine stress (140 µg/kg/min or in inadequate stress response, increased to 210 µg/kg/min) and then at rest. Global and segmental values were acquired by delineating endo- and epicardial borders in Segment (Medviso AB, Lund, Sweden). Segments were defined as functionally affected if baseline left ventricular peak circumferential strain (PCS) values were worse than –16.0%.

Results: In total, 15 patients (65±8 years, 100% women) at baseline were included, and 11 underwent a 6-months follow-up, Table 1. At baseline, the affected segments had significantly higher native T1 (1081 [1045-1193] vs 1052 [1008 - 1119] ms; p< 0.001) and native T2 (53 [49-62] vs 50 [47 - 55] ms; p=0.003) values compared to the unaffected segments. Although resting perfusion did not differ between affected and unaffected segments at baseline, stress perfusion was lower in the affected segments (2.13 [1.72 - 2.52] vs 2.33 [1.86 - 2.84] ml/min/g; p=0.025). Table 2, shows the temporal changes during baseline and follow-up where rest perfusion was reduced at baseline in the affected segments, compared to follow-up values (0.85 [0.64 - 0.99] vs 0.86 [0.70 - 1.03]; p=0.012). Multivariable linear regression analysis on the baseline CMR segments (adjusted R²=0.154, p< 0.001) revealed that T2 (β=0.287, p=0.023) and stress perfusion (β= -0.183, p=0.018) were independently associated with worse PCS.

Conclusion: Resting myocardial perfusion did not differ between affected and unaffected segments in contrast to previous results. Both edema and stress perfusion were independently associated with systolic dysfunction. Functionally affected segments showed reduced stress and rest perfusion at baseline compared to follow-up, shedding light on the temporal trends of functional recovery in TTS.

	Baseline (n=15)	Follow-up (n = 11)	p
BSA (m2)	1.8± 0.2	1.8± 0.2	NS
Creatinine (mmol/l)	72±25	77±12	NS
Hematocrit (%)	40±3	42±3	NS

Baseline vs. Follow-Up Global CMR Values in patients with Takotsubo Syndrome

	Baseline (n=15)	Follow-up   (n = 11)	p
BSA (m2)	1.8± 0.2	1.8± 0.2	NS
Creatinine (mml/l)	72±25	77±12	NS
Hematocrit (%)	40±3	42±3	NS
LVEDV (ml)	141±25	141±30	NS
LVEDVi (ml/m2)	79±14	79±16	NS
LVESV (ml)	75±22	65±24	0.04†
LVESVi (ml/m2)	42±12	37±14	0.038†
LVSV (ml)	66±8	76±21	NS
LVSVi (ml/m2)	37±5	42±10	NS
LVM (g)	81±16	69±16	< 0.001†
LVMi (g/m2)	45±8	39±5	< 0.001†
LVEF (%)	47±7	54±10	NS
patients with <55% LVEF	80% (12)	45% (5)	-
CO (l/min)	4.5±0.7	4.6±1.2	NS
CI (l/min/m2)	2.5±0.4	2.6±0.6	NS
LVGLS (%)	-14.1±4.6	-20.2±3.8	< 0.001†
LV global longitudinal systolic strain rate (1/s)	-0.71±0.19	-0.87±0.18	0.004†
LV global longitudinal diastolic strain rate (1/s)	0.61±0.29	0.85±0.30	0.003†
LVGCS (%)	-17.8±3.8	-21.3±4.6	0.012†
LV global circumferential systolic strain rate (1/s)	-0.83±0.17	-0.89±0.17	NS
LV global circumferential diastolic strain rate (1/s)	0.75±0.26	0.89±0.37	0.052
LVGRS (%)	43.9±11.0	53.3±12.3	0.012†
LV global radial systolic strain rate (1/s)	1.89±0.63	1.94±0.45	NS†
LV global radial diastolic strain rate (1/s)	-2.18±0.53	-3.09±0.93	0.003†

CMR findings are presented as mean ± standard deviation. P-values mark paired t-test; ^† marks Wilcoxon signed-rank test. Abbreviations: BSA = body surface area; LVEDV = left ventricular end-diastolic volume; LVEDVi = left ventricular end-diastolic volume index; LVESV = left ventricular end-systolic volume; LVESVi = left ventricular end-systolic volume index; LVSV = left ventricular stroke volume; LVSVi = left ventricular stroke volume index; LVM = Left ventricular mass; LVMi = Left ventricular mass index; LVEF = left ventricular ejection fraction; CO = cardiac output; CI = cardiac index; LVGLS = left ventricular global longitudinal strain; LVGCS = left ventricular global circumferential strain