The Coronary Flow Reserve Is Transiently Impaired in Tako-Tsubo Cardiomyopathy: A Prospective Study Using Serial Doppler Transthoracic Echocardiography

Methods 

Between September 2005 and January 2007, in a series of 850 patients undergoing coronary angiography for suspected acute coronary syndrome at our institution, 12 consecutive patients (11 female, mean age 68 ± 10 years) satisfying the criteria for LABS2, 3, 4, 5, 6, 7 and without contraindication to adenosine were prospectively included in the study. These 12 patients underwent standard echocardiographic examination and TTE-CFR in the distal part of the LAD during the acute phase of the syndrome (within 48 hours of symptom onset) and after 25 ± 3 days (range: 21-28 days), with commercially available machines (Sequoia 256 ultrasound system, Acuson, Mountain View, CA, and IE33 system, Philips, Amsterdam, The Netherlands). LABS was diagnosed according to the following classic criteria: acute chest pain or dyspnea, a stressful event, transient LV wall-motion abnormalities (WMA) at the apex of the ventricle, S-T segment elevation or T-wave inversion in the electrocardiogram (ECG), and no significant epicardial coronary stenosis on angiography (no luminal narrowing > 50% in all coronary arteries).8, 9 None of the patients in this series had a history of myocardial infarction, significant valvular heart disease, febrile, or acute neurologic disorders. Each patient gave informed consent to participate in the protocol.

Results 

Each case is described in Table 1. Nine patients (75%) were hypertensive, 4 (33%) were treated for dyslipidemia, 3 (25%) were diabetic, and all but one were nonsmokers. The mean body mass index was 27 ± 6 kg/m2 (range: 18-38). A stressful event was identified in each case (emotional n = 10 and physical n = 2). Acute chest pain was the most common presenting symptom. Admission ECGs showed S-T elevation in precordial leads (patients 2, 4, 7, and 8), and in lateral leads (patient 12), slow R progression in precordial leads (patients 1, 10, and 12). There was T-wave inversion in precordial leads (patients 3, 5, 6, 8, 10, and 11) and left bundle branch block (patient 9). The mean QTc was 440 ± 34 milliseconds. Subsequent ECGs showed T-wave inversion at least in precordial leads in all patients. Subacute pulmonary edema that resolved rapidly (within a few hours), was observed in 3 cases on admission (patients 4, 7, and 8). Mechanical and hemodynamic support with intraaortic balloon conterpulsation was required for one patient for 36 hours (patient 12). No other complication occurred during the hospital stay. The mean peak troponin T was 0.6 ± 0.6 μg/L (normal < 0.05). The medications prescribed at discharge were left to the discretion of the physicians. A representative case is described in Figure 1.

Table 1. Patient characteristics

		Case 1	Case 2	Case 3	Case 4	Case 5	Case 6	Case 7	Case 8	Case 9	Case 10	Case 11	Case 12
	Age, y	73	71	54	75	69	74	51	78	65	54	76 (male)	81
	Presenting symptoms	Chest pain	Chest pain	Chest pain	Chest pain, dyspnea	Chest pain	Chest pain	Chest pain, dyspnea	Chest pain, dyspnea	Chest pain	Chest pain	Chest pain, dyspnea	Chest pain, dyspnea
	Stressful event	Emergency operation for her husband	Fear of planned operation	Quarrel with her boss, lost her job	Diagnosis of metastatic cancer for her daughter	Diagnosis of cancer for her daughter	Trauma while home alone	Exacerbated chronic stress (preparation of reception)	Quarrel with a relative	Fear of a threatening dog	Quarrel with her daughter-in-law	Exacerbated lung disease	Repeated annoyance and the strain of a long bus trip
	WMS 1	31	35	37	17	33	36	24	33	26	30	35	36
	WMS 2	17	16	18	16	17	16	16	18	16	16	16	16
	CFR 1	2.15	2.1	2.4	2.7	2.2	2.6	2.8	2.1	1.8	2.4	1.85	1.3
	CFR 2	2.65	2.9	3.05	3	2.7	3.7	3.2	2.7	2.4	2.9	3.1	2.9
	Comorbidity	Hypertension, dyslipidemia	Hypertension, history of atrial flutter	Hypertension, diabetes, dyslipidemia	Hypertension	History of occasional palpitations	Depression	Hypertension	Dyslipidemia, hypertension	Hypertension	Hypertension, diabetes, smoking	Hypertension, lung disease	Hypertension, diabetes, dyslipidemia

CFR, Coronary flow reserve; WMS, wall-motion score. WMS 1 and CFR 1 calculated at the acute phase during the same examination (48 h after admission). WMS 2 and CFR 2 calculated 25 ± 3 days after WMS 1 and CFR 1.

View Large Image Download to PowerPoint

Figure 1. (A) Doppler transthoracic echocardiography (TTE): time course of left ventricular (LV) 4-chamber view (top) and TTE-coronary flow reserve (CFR) in distal part of left anterior descending coronary artery (bottom), at acute phase (left) and 4 weeks later (right). Top, Frames frozen in end systole show large akinetic apical LV chamber at acute phase and normalization of LV systolic function after recovery. Bottom, Coronary flow velocity profile at baseline (left) and during adenosine infusion (right) are depicted at each step and show parallel improvement of CFR: from 2.2 at acute phase (left) to 2.7 after recovery (right). (B) Coronary angiography (top) and left ventriculography (bottom) at acute phase: showing normal coronary anatomy and end-diastolic (left) and end-systolic (right) ventriculograms, with typical balloonlike LV.

In each patient, the initial TTE (performed 6 ± 4 hours within admission) revealed balloonlike LV WMA involving the apical and midsegments with hypercontraction of the basal segments of the LV, and normal LV cavity dimensions. No LV obstruction was observed except in one patient. At the acute phase, patient 5 presented systolic anterior motion of the mitral valve, and mild LV outflow tract obstruction unmasked after the Valsalva maneuver with a peak gradient at 44 mm Hg. Her mitral leaflets were thickened and she presented signs of mitral billowing and mild mitral regurgitation. After recovery, she did not exhibit any LV obstruction.

Adenosine infusion was well tolerated and no serious adverse effect was observed during the two tests. TTE-CFR was performed with the patients in sinus rhythm. The first TTE-CFR was performed while LV WMA was still present in all but one patient who recovered very early (patient 4). Except for heart rate, no significant change of baseline hemodynamic variables was observed between the two examinations (Table 2). The results of coronary flow velocities and CFR in the whole population are described in Table 3 and the course of CFR and WMS for each patient (calculated during the same examination) is shown in Table 1.

Table 2. Time course of hemodynamic variables at the acute phase and after recovery

		Acute phase		Recovery
	N = 12	Baseline	Hyperemia	Baseline	Hyperemia
	Heart rate	84±14	91±16⁎	70±11†	77±16⁎
	SBP, mm Hg	117±15	114±13	132±17	125±17⁎
	DBP, mm Hg	66±12	67±10	64±8	63±13
	RPP, beat/min, mm Hg	9855±2719	10,412±2783	9165±1939	9724±2782

DBP, Diastolic blood pressure; hyperemia, using intravenous adenosine infusion; RPP, rate-pressure product; SBP, systolic blood pressure.

⁎ P < .05 vs baseline. † P < .05 vs acute phase.

Table 3. Time course of coronary flow velocities and coronary flow reserve at the acute phase and after recovery

		Acute phase		Chronic phase
	N = 12	Peak V	Mean V	Peak V	Mean V
	Basal	30±7	23±4	30±7	23±6
	Hyperemia	69±17⁎	51±12⁎	89±20⁎†	67±17⁎†
	CFR	2.3±0.4	2.2±0.4	3±0.4†	2.9±0.3†

CFR, Coronary flow reserve; V, velocity (cm/s) in the left anterior descending coronary artery.

⁎ P < .01 vs baseline. † P < .01 vs acute phase.

Resting coronary flow velocity did not vary significantly during follow-up, compared with the acute phase. A significant increase of hyperemic flow velocity was observed at follow-up, compared with the acute phase. A significant improvement of CFR was, therefore, observed between the two examinations (Figure 1). TTE-CFR increased from 2.2 ± 0.4 (median = 2.2) at the acute phase to 2.9 ± 0.3 (median = 2.9, P = .0005). All patients exhibited an increase of CFR between the two tests (delta CFR = 0.73 ± 0.4, range: 0.3-1.6) (Figure 2).

View Large Image Download to PowerPoint

Figure 2. Time course of change in mean coronary flow reserve (CFR): chronic (2) versus acute (1) phase. Dot and line diagram shows that CFR increases in each patient whereas left ventricular function improves.

The interval of agreement for repeated measures of CFR (second-first test) obtained in 10 volunteers was from −9% to +12%. The individual variation of CFR was considered significant when outside this interval of agreement. An increase exceeding 12% in the CFR after recovery was detected in all but one patient (patient 4) affected by LABS (mean increase 38%).

LV systolic dysfunction and WMA rapidly reversed, returning to the normal range in all patients (Figure 1), as assessed by follow-up TTE 25 ± 3 days after admission, and as early as 48 hours for patient 4 with a history of a similar episode 2 months previously. Therefore, LV EF, assessed by echocardiography, increased from 40 ± 7% at the acute phase to 68 ± 5% at the chronic phase (P = .0005) and WMS decreased from 31 ± 6 at the acute phase to 16.5 ± 0.8 at the chronic phase (P = .0005) with delta WMS = −14.6 ± 6 (range: −20 to −1). Similarly, WMSI decreased from 1.93 ± 0.36 at the acute phase to 1.02 ± 0.05 (P < .01), delta WMSI = −0.9 ± 0.36. A significant correlation was observed between delta CFR and delta WMS (r = −0.89, P = .003), between delta CFR and delta WMSI (r = −0.88, P = .003), and between delta CFR and WMS and WMSI at the acute phase (r = 0.84, P = .005, and r = 0.92, P = .002, respectively) (Figure 3).

View Large Image Download to PowerPoint

Figure 3. Scatter plots showing relationship between change in coronary flow reserve and change in wall-motion score (WMS) (first panel) or WMS index (second panel). Highly significant correlation was observed between these variables.

Discussion 

These 12 patients presented features of LABS: elderly women (11 of 12) with chest pain and reversible balloonlike LV WMA at the apex extending beyond the perfusion territory of a single coronary artery and sparing the basal segments, ST-T abnormalities on ECG, absence of major epicardial coronary artery narrowing on angiography, and a possible emotional or physical stress as a trigger.2, 3, 4, 5, 6, 7, 8, 9 They also presented no evidence of myocarditis, pheochromocytoma, or neurologic disorders suggestive of subarachnoid hemorrhage. A vasospastic mechanism can also be discussed: for patients 2, 4, 7, 8, and 12, the coronary angiogram performed during acute S-T elevation was normal, and the distribution of WMA at admission encompassed a single vessel territory. For the other patients, the initial ECG result argued against coronary spasm, which would be expected to be associated with immediate S-T elevation.

The pathophysiology of LABS has not been clearly elucidated. An exaggerated transient catecholamine release giving rise to myocardial stunning in a predisposed myocardium has been postulated.6, 7 Sudden surges in circulating catecholamine levels could induce direct myocyte injury or microcirculatory damage or both, but the exact mechanism has not been identified. In this setting, the apex is thought to be more vulnerable than other regions.2 The earliest studies that have directly assessed the microcirculation in the setting of LABS provided conflicting results.2, 3, 4 In the study by Abe et al,4 3 patients were submitted to an invasive Doppler flow wire to evaluate the CFR at the middle portion of the LAD at the acute phase of the syndrome. The median value of CFR was 2.2 (range: 1.3-3.6). The authors concluded that no significant abnormality in the microcirculation was detected. However, because of its invasive nature, no subsequent CFR was performed during follow-up, particularly after recovery, and no definitive conclusions can be drawn concerning the CFR values obtained at the acute phase. Tsutchihashi et al2 described a patient with transient depression of CFR as assessed by nuclear cardiographic evaluations performed at the acute phase and 5 months later by positron emission computed tomography. Kurisu et al3 suggested that microvascular spasm could be one of the mechanisms of tako-tsubo. Bybee et al10 found that patients with LABS had significantly abnormal thrombolysis in myocardial infarction frame count, an index of coronary blood flow, compared with control subjects, and this abnormality involved all 3 major epicardial vessels in nearly two thirds of their patients (10 of 16). A recent study using a Doppler guidewire11 suggested transient coronary microcirculatory damage at the acute phase of the syndrome. Echocardiography is underused in this setting, represented only by few recent case reports using myocardial contrast echocardiography and TTE-CFR.15, 16, 17 Yet, because of the capacity of TTE to noninvasively analyze both LV wall motion and coronary flow, this imaging modality allows serial evaluation in the setting of transient cardiomyopathy such as LABS. As the patients described in this study did not present any flow-limiting LAD stenosis, CFR in this setting explores the microcirculation.21, 23 Adenosine acts substantially at the microcirculatory level, given that great epicardial arteries are relatively insensitive to this agent.26 Consequently, the variation of flow velocity recorded in the distal part of the LAD with adenosine is a surrogate marker of the variation of regional blood flow. Although the cut-off value of 2 for CFR is well documented to detect significant coronary stenosis or myocardial ischemia,22, 23 no cut-off value has been clearly defined when assessing microcirculation. However, comparison with the patient acting as his or her own control with serial measurements before and after an event can suggest a variation in microcirculatory disorders. Interestingly, the majority of patient (75%) had a CFR greater than or equal to 2 at the acute phase of LABS contrasting with the extent of WMA. However, the mean value of CFR was lower than what would be expected in patients without coronary artery disease.21 A significant increase of CFR was subsequently observed in all but one patient after the event had resolved. The CFR increased whereas LV wall motion improved, suggesting a relationship between WMA and microcirculatory impairment in the tako-tsubo syndrome. Illustrating this link, the correlation between delta WMS and delta WMSI and delta CFR, and between the extent of WMA at the acute phase and the change in CFR over time, despite the small sample size, was very significant. Interestingly, patient 4, who recovered very rapidly (within 48 hours), had one of the highest values of CFR at the acute phase and a less marked increase of CFR at follow-up. On the contrary, patient 12, who had extensive WMA and temporary mechanical support, had the lower CFR at the acute phase and the greater increase of CFR at follow-up, in parallel to total recovery of LV function. To our knowledge, this is the first study to assess the coronary microcirculation in LABS, particularly in a Caucasian population, using serial TTE-CFR measurements. Our results are in agreement with a recent Asian invasive study using a Doppler guidewire in the 3 coronary arteries of 8 patients,11 indicating that these two different populations share at least certain common pathophysiologic features of LABS.

Improvement of hemodynamic factors and metabolic state27 or prescription of drugs with a favorable action on the coronary microcirculation could have influenced the increase of CFR at follow-up. However, although the heart rate decreased slightly, no significant change of the rate pressure product was observed between the two examinations and no differences could be observed concerning the prescription of major drugs such as beta-blockers, angiotensin-converting enzyme inhibitors, and statins before and after the event in the whole population. CFR increased at follow-up as hyperemic flow velocity was higher on the second test whereas baseline diastolic flow velocity remained unchanged, suggesting a blunted vasodilating capacity at the acute phase of the syndrome instead of a higher metabolic state as a cause of the lower initial CFR. The precise mechanism of the transient decrease of hyperemic coronary blood flow remains to be elucidated.

Finally, serial noninvasive CFR measurements performed in these 12 patients with LABS suggested transient microcirculatory impairment during the acute phase of the syndrome. The wall-motion improvement parallel to the dynamic improvement of the CFR suggests a role of coronary microcirculatory damage in the pathogenesis of acute and transient WMA in LABS.