Acute stress-induced (Tako-tsubo) cardiomyopathy is an increasingly recognized but insufficiently characterized syndrome. Here, we investigate the pathophysiology of right ventricular (RV) involvement in Tako-tsubo and its recovery time course. We prospectively recruited 31 patients with Tako-tsubo with predominantly ST-elevation electrocardiogram and 18 controls of similar gender, age, and co-morbidity distribution. Patients underwent echocardiography and cardiac magnetic resonance (CMR) imaging on a 3T Philips scanner in the acute phase (day 0 to 3 after presentation) and at 4-months follow-up. Visually, echocardiography was able to identify only 52% of patients who showed RV wall motion abnormalities on CMR. Only CMR-derived RV ejection fraction (p = 0.01) and echocardiography-estimated pulmonary artery pressure (p = 0.01) identify RV functional involvement in the acute phase. Although RV ejection fraction normalizes in most patients by 4 months, acutely there is RV myocardial edema in both functioning and malfunctioning segments, as measured by prolonged native T1 mapping (p = 0.02 for both vs controls), and this persists at 4 months in the acutely malfunctioning segments (p = 0.002 vs controls). The extracellular volume fraction was significantly increased acutely in all RV segments and remained increased at follow-up compared with controls (p = 0.004 for all). In conclusion, in a Tako-tsubo population presenting predominantly with ST-elevation electrocardiogram, we demonstrate that although RV functional involvement is seen in only half of the patients, RV myocardial edema is present acutely throughout the RV myocardium in all patients and results in microscopic fibrosis at 4-month follow-up.
Acute stress-induced (Tako-tsubo) cardiomyopathy is an increasingly recognized clinical syndrome. Although classically viewed as an acute disorder of the left ventricular (LV) systolic function (often triggered by intense emotional of physical trauma), right ventricular (RV) involvement has been increasingly described. Although it is not uniformly clear if RV involvement portends a worse outcome, involvement of the RV has been associated with greater LV dysfunction, more frequent complications, or prolonged hospital stay. We have previously demonstrated that the initial LV dysfunction in Tako-tsubo is accompanied by a significant degree of LV myocardial edema and profound LV energetic impairment with incomplete resolution of both after 4 months. Although RV dysfunction appears to characterize those with a worse spectrum of LV dysfunction, the nature of the pathophysiology underlying the RV involvement is unknown and the recovery time course of the RV has not been described. In the present study, we investigated in a prospectively recruited, predominantly ST-elevation Tako-tsubo population: (1) the prevalence and extent of RV functional involvement, (2) the RV myocardial tissue characteristics, and (3) the recovery time course of the RV at 4-month follow-up.
Methods
Thirty-one consecutive consenting patients (28 women, median age 64 years [range 41 to 84 years]) who met the Mayo Clinic criteria for Tako-tsubo were recruited from Aberdeen Royal Infirmary from 2013 to 2015. All patients were initially suspected of myocardial infarction and most were brought in by Scottish Ambulance Service through the primary percutaneous coronary intervention service. Patients were studied in the acute phase (day 0 to 3 after presentation) and restudied after 4 months. Eighteen controls of similar age (mean 63 years, range 47 to 80 years), gender (16 women), and co-morbidity distribution were also recruited. The study was approved by the North of Scotland Research Ethics Committee, and all subjects provided informed consent.
Echocardiography was performed using a Vingmed E9 system (GE Healthcare, Norway) with a 2.5-MHz probe. Standard images were acquired from parasternal, apical, and subcostal views. Tissue Doppler images of the RV (at lateral tricuspid annular level) were used to derive E’, A’, and S’ indices.
A 3T Philips Achieva (Best, The Netherlands) was used for cardiac magnetic resonance (CMR) imaging. After localizers, 2-, 3-, and 4-chamber views and a full short-axis cine stack were acquired, as well as a full short-axis stack of native and postcontrast 3-3-5 modified Look-Locker imaging T1 mapping.
Both echo and CMR images were each analyzed by a pair of 2 independent expert observers (CS, AR, CN, and JS) blinded to each other, to the other imaging method and for T1 mapping to the order of the scan. Qualitative interpretation (wall motion score [WMS]) was resolved by adjudication/agreement with a third expert. All quantitative analyses were subjected to interobserver variability. Echo images were analyzed offline using Echopac (GE Healthcare, Norway). Pulmonary artery pressure (PaP) was derived from the maximum tricuspid regurgitation velocity as measured by the formula 4 v 2 +estimated right atrial pressure depending on the inferior vena cava collapse during inspiration. A simple index of RV longitudinal function was also measured as tricuspid annular pansystolic excursion (TAPSE) as previously described. The CMR images were analyzed in CMR tools (Cardiovascular Imaging Solutions, London, United Kingdom) for computation of LV volumes and mass, RV volumes and to derive volumetric ejection fractions (EFs) for the 2 ventricles. RV WMSI was obtained using a 6-segment model in which the basal, mid cavity, and apical slices were each divided into anterior and posterior segments ( Figure 1 ). Each segment was scored for functional status (1 = normal, 2 = hypokinetic, 3 = akinetic, and 4 = dyskinetic). T1 maps were generated using Philips’ RelaxMap, quality controlled with chi-square maps and imported into ImageJ (National Institute of Health, Bethesda, Maryland) where precontrast and postcontrast T1 values for the RV myocardium were generated for each of the 6 segments. To avoid the difficulties posed by the thin RV wall and possible partial volume effect, all T1 map images were coregistered with the exact spatially corresponding cine sequences in the following manner: the RV boundaries were carefully delineated manually on magnified cine images, then inwardly eroded by at least 2 pixels to avoid blood on the endocardial side or fat on the epicardial side and finally the regions of interest thus obtained were imported into the corresponding T1 map images for deriving precontrast and postcontrast T1 values. Extracellular volume fraction (ECV) was calculated according to the formula:
A repeat echocardiogram and CMR were scheduled at 4 months and achieved in 28 patients because of death (2) and device implant (1).
Data are presented as mean ± SD or median (range) if not normally distributed. Comparisons between controls and patients or between patients at baseline and follow-up were performed using independent/paired Student t test or Mann–Whitney rank-sum/Wilcoxon signed-rank tests depending on the data distribution. To account for multiple comparisons, the p value chosen for statistical significance was Bonferroni-adjusted depending on the number of variables compared between groups. Interobserver variabilities were calculated as mean ± SD of the percentage ratios between differences and means of the 2 independently measured variables.
Results
General characteristics are listed in Table 1 . In this study cohort, most patients were middle aged and elderly women, with a preceding emotional trigger, presenting with ST-elevation eelctrocardiograms, modest troponin increase and apical type ballooning. The controls were chosen to be similar in age and gender distribution to the patients with Tako-tsubo and to have a comparable cardiac past medical history/therapy. Typically, patients with Tako-tsubo were given the emergency therapy required for a presumed myocardial infarction until the diagnosis of Tako-tsubo was established at cardiac catheterization; after this, medication was reestablished as before presentation in each case.
| Clinical Characteristics | Tako-tsubo (n=31) | Controls (n=18) |
|---|---|---|
| Past medical history | ||
| Hypertension | 7 (22%) | 4 (22%) |
| Prior coronary artery disease | 1 (3%) | – |
| Diabetes mellitus | 2 (6%) | – |
| Thyroid disease | 8 (26%) | 5 (27%) |
| Mental health disorder | 9 (29%) | – |
| Previous Tako-tsubo | 6 (19%) | – |
| Smoker | 11 (35%) | 6 (33%) |
| Alcohol abuser | 3 (9%) | – |
| Body mass index>25 kg/m 2 | 16 (51%) | 9 (50%) |
| Clinical presentation | ||
| Chest pain | 25 (81%) | |
| Syncope/pre-syncope | 4 (13%) | |
| Ventricular tachycardia | 2 (6%) | |
| Precipitating emotional stressor | 27 (87%) | |
| No stressor | 4(13%) | |
| Heart rate at presentation [bpm, mean±SD] | 80±14 | |
| Systolic blood pressure at presentation [mm Hg, mean±SD] | 125±23 | |
| Diastolic blood pressure at presentation [mm Hg, mean±SD] | 79±17 | |
| ECG changes at presentation | ||
| ST-elevation | 25 (81%) | |
| T wave inversion | 4 (12%) | |
| Left bundle branch block | 2 (7%) | |
| Troponin I (ng/ml) | ||
| At presentation [median, (range)] | 1.28 (<0.04, 10.93) | |
| 12 hour [median, (range)] | 3.45 (0.22, 11.97) | |
| C-reactive protein [median, (range)] | 21.5 (<4, 75) | |
| Coronary disease at angiography | ||
| Left anterior descending <50% | 6 (19%) | |
| Left circumflex <50% | 1 (3%) | |
| Right <50% | 6 (19%) | |
| Left ventricular angiogram | ||
| Apical ballooning | 26 (84%) | |
| Mid-cavity ballooning | 5 (16%) | |
| Drug history | ||
| Aspirin | 3 (9%) | – |
| Beta-blocker | 1 (3%) | – |
| ACE inhibitor | 6 (19%) | 2 (11%) |
| Calcium channel blocker | 3 (9%) | – |
| Statin | 4 (12%) | – |
| Diuretic | 5 (16%) | 3 (16%) |
RV systolic wall motion was assessed on a segmental basis as presented in Figure 1 . Patients with Tako-tsubo were grouped according to the WMSI on CMR images as those with RV involvement (RV+, mean WMSI 1.5 ± 0.3, n = 16) and without RV involvement (RV−, WMSI = 1, n = 15). There was no difference between the RV+ and RV− groups with regard to age, medical history, heart rate, or systolic/diastolic blood pressure at presentation (p = not significant for all). Figure 1 shows a typical example of RV dyskinesia. In all cases, only the apical and mid cavity RV segments were seen to be involved. Of note, echocardiography was able to identify a wall motion abnormality in only 9 patients in the RV + group (52%). Table 2 shows that in both groups, there was a significant improvement in LV EF at 4 months compared with baseline (p <0.01 in both). The RV + group had decreased RV EF (p = 0.01), decreased TAPSE (p = 0.002), and increased PaP (p = 0.01) acutely compared to follow-up and also compared to controls (all p <0.01). The RV – group also showed a decreased TAPSE acutely compared with controls (p = 0.002), which improved at follow-up (p = 0.001 vs acute).
| Controls n=18 | Acute study (day 0-3) | Follow up (4 months) | |||
|---|---|---|---|---|---|
| RV+ n=16 | RV- n=15 | RV+ n=14 | RV- n=14 | ||
| Cardiac Magnetic Resonance | |||||
| Left ventricular end-diastolic volume (ml) | 133±14 | 126±35 | 131±22 | 122±21 | 127±16 |
| Left ventricular end-diastolic volume index (ml/m 2 ) | 73±6 | 80±20 | 74±13 | 78±12 | 71±10 |
| Left ventricular end-systolic volume (ml) | 45±8 | 67±30 ∗ | 57±20 ∗ | 45±15 † | 45±11 † |
| Left ventricular end-systolic volume index (ml/m 2 ) | 24±4 | 42±18 ∗ | 32±12 ∗ | 28±9 † | 25±7 † |
| Left ventricular mass (g) | 121±13 | 137±16 ∗ | 134±18 ∗ | 125±17 † | 118±18 † |
| Left ventricular mass index (g/m 2 ) | 66±10 | 84±10 ∗ | 77±17 ∗ | 75±11 † | 70±14 † |
| Left ventricular ejection fraction (%) | 67±4 | 48±10 ∗ | 60±8 ∗ | 63±6 † | 64±6 † |
| Right ventricular end-diastolic volume (ml) | 100±20 | 121±23 | 107±22 | 111±20 | 107±26 |
| Right ventricular end-diastolic volume index (ml/m 2 ) | 58±9 | 66±12 | 58±10 | 60±10 | 58±13 |
| Right ventricular end-systolic volume (ml) | 44±7 | 43±15 | 38±7 | 38±15 | 34±9 |
| Right ventricular end-systolic volume index (ml/m 2 ) | 25±4 | 25±10 | 20±4 | 22±10 | 18±4 |
| Right ventricular ejection fraction (%) | 63±5 | 57±11 ∗ | 63±9 | 66±11 † | 67±8 |
| Echocardiography | |||||
| Right ventricular E’ (cm/s) | 0.14±0.01 | 0.09±0.04 | 0.10±0.04 | 0.12±0.02 | 0.11±0.03 |
| Right ventricular A’ (cm/s) | 0.16±0.01 | 0.17±0.05 | 0.14±0.05 | 0.13±0.05 | 0.15±0.04 |
| Right ventricular S’ (cm/s) | 0.13±0.01 | 0.09±0.02 | 0.12±0.03 | 0.12±0.03 † | 0.12±0.03 |
| Tricuspid annular pan-systolic excursion (cm) | 2.6±0.1 | 1.7±0.4 ∗ | 2.0±0.3 ∗ | 2.4±0.5 † | 2.6±0.4 † |
| Pulmonary artery Pressure – estimated (mm Hg) | 24±7 | 43±15 ∗ | 31±6 | 33±8 † | 29±6 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
