Abstract
Background
Acute hospitalized heart failure (AHHF) is associated with 40–50% risk of death or rehospitalization within 6 months post-discharge. Timely (before hospital discharge) risk stratification of AHHF patients is crucial. We hypothesized that mechanical alternans (MA) and T-wave alternans (TWA) is associated with post-discharge outcomes in AHHF patients..
Methods and Results
A prospective cohort study was conducted in the intensive cardiac care unit (ICCU) and enrolled 133 patients (59.6±15.7 y.; 65% men) admitted with AHHF. Surface ECG and peripheral arterial blood pressure waveform via arterial line were recorded continuously during the ICCU stay. MA and TWA were measured by enhanced modified moving average method. All-cause death or heart transplant served as a combined primary endpoint. MA was observed in 28 patients (25%), while TWA was detected in 33 patients (33%). If present, MA was tightly coupled with TWA. Mean TWA amplitude was larger in patients with both TWA and MA, as compared to patients with lone TWA (median 37(IQR 26–61) vs. 22(21–23) μV; P=0.045). After a median of 10 months post-discharge, 42 (38%) patients died and 2 had heart transplants. MA was associated with the primary endpoint in univariable Cox model [HR 1.84(95%CI 1.00–3.40); P=0.05], and after adjustment for left ventricular ejection fraction, NYHA HF class, and implanted ICD/CRT-D device, [HR 2.12 (95%CI 1.13–3.98); P=0.020]. TWA without consideration of simultaneous MA was not significantly associated with primary endpoint (HR 1.42; (95%CI 0.77–2.64); P=0.260).
Conclusion
MA is independently associated with outcomes in AHHF.
Clinical Trial Registration Information
URL:http://www.clinicaltrials.gov. Unique identifier: NCT01557465
Keywords: T-wave alternans, Mechanical alternans, Heart failure, mortality
Heart failure (HF) is one of the most frequent reasons for hospital admissions in the USA. There are about 1 million admissions per year with the primary diagnosis of HF, and an additional 2 million admissions per year with the secondary diagnosis of HF1. While mortality in chronic HF has substantially decreased over the last 15 years, post-discharge mortality in acute hospitalized HF (AHHF) has not changed2. In order to facilitate the development of novel prognostic and therapeutic approaches, and to improve outcomes, AHHF was recognized as a distinct entity3.
Sudden cardiac death (SCD) is the second (after pump failure) most frequent cause of death in AHHF, responsible for about a third of all deaths4. There is an unmet need for timely (before hospital discharge) risk stratification of AHHF patients. T-wave alternans (TWA)5–7 on the surface ECG, as a manifestation of action potential (AP) duration alternans and AP voltage alternans, is directly linked to ventricular arrhythmia. However, results of clinical studies are controversial7–9. Mechanical alternans (MA) [aka pressure alternans, pulsus alternans] is a phenomenon of alternating strong and weak beats, as measured by pulse, or blood pressure. The first observation 10;11 of pulsus alternans in a HF patient was published in 1872 by Ludwig Traube (1818–1876). MA is prevalent in HF 12, and is associated with cardiovascular death and HF hospitalizations 13. Coupling between MA and TWA has been previously shown 14;15. Mechanistically, abnormal intracellular Ca2+ cycling could be responsible for both TWA and MA16;17. On the other hand, the mechanisms of TWA and MA may be different18, and beat-to-beat alternations in stroke volume could be due to variations in preload and contractility. Very few studies have simultaneously measured both MA and TWA. We conducted a prospective observational cohort study of MA and TWA in order to compare the association of these 2 types of alternans with AHHF outcomes. We hypothesized that MA and TWA is associated with post-discharge outcomes in AHHF patients.
Methods
Study population
We conducted a single-center prospective observational cohort Mechanical Alternans Study (MAS, NCT01557465) at the Johns Hopkins Hospital Intensive Cardiac Care Unit (ICCU) between December 1, 2009 and November 30, 2011, and enrolled 133 participants. The study protocol was approved by the Johns Hopkins University Institutional Review Board, and all patients gave written informed consent before enrolling in the study.
Patients were eligible for the study if they were admitted to the ICCU with either acute decompensated HF or new onset acute HF due to one of the following conditions: (1) decompensated ischemic or non-ischemic cardiomyopathy; (2) acute coronary syndrome; (3) electrical storm in ICD patients or resuscitated out-of-hospital sudden cardiac arrest (SCA). Study participants had to have clinical indications for, and undergo, continuous invasive monitoring of peripheral arterial blood pressure via arterial line. Patients with (1) pericardial effusion, (2) congenital heart disease, (3) hemodynamically significant (moderate or severe) valvular heart disease, (4) children, and (5) pregnant women were excluded.
All study participants underwent clinical echocardiographic examinations during the ICCU stay. Left ventricular ejection fraction (LVEF) was measured by biplane Simpson’s method.
Continuous recording and analysis of ECG and blood pressure
Simultaneous recording of ECG (leads V1 and II) and peripheral arterial blood pressure waveform via arterial line was performed continuously during the ICCU stay (as long as invasive blood pressure monitoring continued). The initiation and discontinuation of intra-aortic balloon pump or percutaneous cardiopulmonary support were carefully documented.
Recordings were excluded from this analysis if at least one the following was present: (1) use of intra-aortic balloon pump or percutaneous cardiopulmonary support system; (2) atrial fibrillation (AF); (3) frequent premature ventricular contractions (PVCs) as bigeminy, trigeminy, or quadrigeminy. Heart rate, QT interval, systolic and diastolic arterial pressure were averaged for each 15-sec period.
Alternans Detection and Analysis
Customized MATLAB (MathWorks, Inc, Natick, MA) software was developed to detect MA and microvolt TWA. Details of alternans method are provided in the Supplemental Method online. The correlation method was applied first to determine the presence of alternans 19. Then a modified moving average (MMA) method was applied to quantify alternans20. Rapid update factor of 1/8 was used in our study to provide high sensitivity7. A subset of data (n=12) was re-analyzed using an incremental updating factor of 1 in 32. A sustained TWA event was detected when TWA lasted for at least 20 beats (Figure 1A). Similarly, a sustained MA event was defined as arterial pressure alternans that lasted for at least 20 beats (Figure 1B). Actual duration (number of beats) of each alternans events was measured as the number of consecutive beats whose alternans correlation index values alternated around the value of one. No amplitude threshold was defined a priori to detect sustained alternans. The resolution of MA evaluation was 1 mmHg, and the resolution of TWA evaluation was 10 μV.
Figure 1:
Illustration of the method of (A) TWA, and (B) MA measurements by modified moving average. Lower panel shows averaged Odd and Even beats. Alternans are marked by arrows. STT=ST-T ECG segment. MMA=modified moving average method. CM=correlation method. MA=mechanical alternans.
Follow-up and study primary end-point
After ICCU discharge, patients were followed prospectively via the electronic medical record review and via a phone call once a year. A combined endpoint, defined as all-cause death or heart transplant, served as the primary end-point in this study.
Statistical Analysis
Statistical analysis was performed using STATA 13.0 software (StataCorp LP, College Station, TX). All normally distributed continuous variables are presented as mean ± standard deviation (SD), while median and interquartile range (IQR) is reported if distribution was not normal. Two-group mean comparison t-test was used to compare normally distributed characteristics of patients with and without MA. Wilcoxon rank-sum test was used for two-group comparison of skewed continuous variables in patients with lone TWA, vs. those with both MA and TWA. Categorical variables were compared by Pearson’s chi-square test. A p-value < 0.05 was considered statistically significant. Kaplan-Meier survival curves were constructed for subjects with and without MA and TWA. The log-rank (Mantel-Cox) statistic was computed to test the equality of survival distributions. Univariable and multivariable Cox proportional hazards regression analyses were performed. MA and TWA were entered in the models as categorical variables (yes/no). First, the association of each variable [MA, TWA, demographics, and potential confounders (NYHA class III-IV, LVEF≤35%, implanted ICD/CDT-D, mean heart rate >90bpm, type of cardiomyopathy)] with outcome was tested in univariable Cox models. Then, variables that were associated with outcome at P value <0.1 were included in the multivariable model. To evaluate proportionality of hazards, we performed tests of nonzero slope in a generalized linear regression of the scaled Schoenfeld residuals.
Results
Study population
Out of the 133 ICCU patients enrolled in the MAS study, 22 patients were excluded due to the use of the intra-aortic balloon pump or percutaneous cardiopulmonary support system, AF, or frequent PVCs. Data from the remaining 111 patients were analyzed. Clinical characteristics of the study population are presented in Table 1. The mean age of participants was 58.9±15.9 years, and most were white (60%) men (71%). Most of the patients (82.3%) had NYHA HF class III-IV. Mean EF was 29.0±18.1 %. About half of the study population had ischemic cardiomyopathy, and had an ICD or CRT-D implanted.
Table 1.
Clinical characteristics of study participants
| Characteristic | TW(−) MA(−) (n=78) | TWA(+) MA(+) (n=28) | TWA(+) MA(−) (n=5) | P 1 | P 2 |
|---|---|---|---|---|---|
| Age(SD), y | 57.6(15.4) | 62.3(16.7) | 63(61–67) | 0.232 | 0.763 |
| Male sex, n(%) | 55(70.5) | 21(75.0) | 3(60.0) | 0.804 | 0.771 |
| Whites, n(%) | 45(57.7) | 18(64.3) | 4(80.0) | 0.623 | 0.544 |
| History of Ischemic CM, n(%) | 32(41.6) | 13(46.4) | 3(60.0) | 0.155 | 0.169 |
| Acute HF indication, n(%) | 58(74.4) | 19(67.9) | 2(40.0) | 0.654 | 0.234 |
| ACS indication, n(%) | 17(21.8) | 8(28.6) | 1(20.0) | 0.457 | 0.755 |
| SCA hospitalization, n(%) | 9(11.5) | 5(17.9) | 5(100.0) | 0.904 | 0.0001 |
| NYHA class III-IV, n(%) | 62(66.7) | 14(50.0) | 2(40.0) | 0.157 | 0.182 |
| ICD/CRT-D, n(%) | 40(51.3) | 11(39.3) | 2(40.0) | 0.300 | 0.518 |
| Ventricular Pacing, n(%) | 12(15.4) | 6(21.4) | 0 | 0.457 | 0.314 |
| Beta-blockers, n(%) | 49(68.1) | 19(73.1) | 3(60.0) | 0.597 | 0.810 |
| Amiodarone, n(%) | 22(30.6) | 8(30.8) | 0(0) | 0.831 | 0.340 |
| Mechanical ventilation, n(%) | 3(3.9) | 1(3.6) | 0(0) | 0.948 | 0.668 |
| Inotropic medications, n(%) | 31(39.7) | 9(32.1) | 2(40.0) | 0.477 | 0.731 |
| Baseline LVEF(SD), % | 26.6(17.1) | 33.1(18.7) | 60(20–60) | 0.434 | 0.375 |
| Heart rate±SD, bpm | 93.1(21.7) | 81.0(20.9) | 92(82–100) | 0.013 | 0.269 |
| QTc (SD), ms | 456.5(52.1) | 466.4(52.4) | 456(442–468 | 0.415 | 0.650 |
P1 = t-test comparison of patients with and without MA; P2 = Wilcoxon rank-sum test comparison of TWA patients with and without MA
Mechanical alternans and TWA
Mean heart rate of analyzed ECG and blood pressure epochs in all study participants exceeded 80 bpm. Sustained MA events were observed in 28 patients (25.2%). Sustained TWA events were observed in about 30% (33 patients). TWA was observed in all 28 patients with MA, and thus isolated TWA without accompanying MA was observed in 5 patients. A representative example of an observed sustained MA event is presented in Figure 2A. Clusters of sustained MA and TWA events frequently occurred at the same time (Figure 2B). Simultaneous MA and TWA events were observed in 22 out of 28 patients who had both MA and TWA events. The number of simultaneous MA and TWA events in a patient ranged 0 – 11 per hour, and the average duration of the overlapping periods was 25.2± 29.0 sec. Supplementary movie 1 shows relationships and coupling between TWA and MA in 10 study participants. Use of incremental updating factor of 1 in 32 resulted in a negligible decrease in the number of detected alternans (0.5% MA and 3% TWA events).
Figure 2:
A. Representative example of MA in study participant MA074, African-American man, 41y, non-ischemic familial dilated cardiomyopathy, LVEF 15%, NYHA class III. Admitted after VF cardiac arrest. TWA 29 μV. MA 1mmHg. Alive. B. Representative example of detected TWA and MA events during 1h 45 min in study participant MA016, African-American woman, 88y, ischemic cardiomyopathy, LVEF 40%. Admitted due to acute decompensated HF.TWA 11 μV, MA 4.5 mmHg. Died. An upper tracing shows heart rate averaged during every 15-sec epoch.
To explore temporal association of TWA and MA, we computed amplitude and duration of sustained TWA events that occurred within one hour before and after each sustained MA event. There were 624 sustained TWA events (31.4±7.0 beats) that occurred within one hour before sustained MA events, and 1402 sustained TWA events (65.2±17.7 beats) that occurred within one hour after MA events. The durations of TWA events that followed MA events were significantly longer than those of TWA events that occurred before MA events (P < 0.0001). Thus, the number of TWA beats was significantly larger within one hour after each MA event than within one hour before each MA event (Supplemental Figure 1). In other words, on average, longer TWAs were detected within one hour after each MA event. On the other hand, the average duration of MAs detected within one hour before each TWA event (32.7±7.5 beats) was almost identical to the average duration within one hour after each TWA event (32.5±7.5beats). This suggests that MA events were more likely to precede TWA events and that far more TWA events than MA events were detected. MA event did not affect TWA amplitude (27.0±7.5 before MA vs. 24.1±5.6 μV after MA, NS). Similarly, TWA event did not affect MA amplitude (14.5±1.7 before TWA vs. 12.0±3.2 mmHg after TWA event). However, MA amplitude within 1 hour of a TWA event was higher than average MA amplitude (13.3±2.5 vs. 5.0±4.7 mmHg; P=0.046).
No significant difference in LVEF, medications and indications for ICCU admission between patients with and without MA was observed (Table 1). Remarkably, patients with lone TWA had less impaired systolic function. All 5 patients with lone TWA were admitted to the ICCU due to either resuscitated out-of-hospital SCA, or electrical storm in ICD patients. Mean TWA amplitude and TWA amplitude at maximum heart rate were significantly larger in patients with both TWA and MA, as compared to patients with lone TWA (Table 2). Alternans was measured during ventricular pacing (VP) in 18 (16%) patients (Table 1). Amplitude of alternans in VP-patients did not differ from the amplitude of alternans, measured in patients in sinus rhythm.
Table 2.
Comparison of TWA characteristics
| Characteristic | TWA without MA (n=5) | TWA with MA (n=28) | P |
|---|---|---|---|
| Mean amplitude of TWA, median(IQR), μV | 22(21–23) | 37(26–61) | 0.045 |
| TWA amplitude at max HR, median(IQR), μV | 36(23–42) | 58(44–137) | 0.021 |
| Number of TWA beats, median(IQR),n | 22(22–26) | 24(22–30) | 0.480 |
| Mean Heart rate at TWA, median (IQR), bpm | 79.8(70.6–92.6) | 72.4(57.8–86.2) | 0.292 |
| Mean QTc at TWA, median (IQR), ms | 431(385–441) | 410(359–481) | 0.920 |
Prospective follow-up
During a median 10 months of follow-up, 44 patients (39.6%) died or had successful heart transplantation. Out of these 44 patients, 38 patients (86.4%) were hospitalized in the ICCU with decompensated end-stage HF, 24 patients (59%) had NYHA class IV symptoms, and 27 patients (61.4%) had ICD or CRT-D implanted. Review of the medical records showed that heart transplantation (n=2), or progressive pump failure with pulseless electrical activity, or asystole was an underlying cause of the primary endpoint in 35 patients (80%). In the remaining 9 patients (20%), the immediate cause of death was not determined. Patients with the primary endpoint were likely to have ischemic cardiomyopathy with advanced systolic dysfunction (Table 3).
Table 3.
Comparison of patients with and without primary endpoint
| Characteristic | Free from primary endpoint (n=67) | With primary endpoint (n=44) | P |
|---|---|---|---|
| Age(SD), y | 59.8(16.1) | 59.0(15.0) | 0.795 |
| Male sex, n(%) | 44(65.2) | 35(79.6) | 0.103 |
| Whites, n(%) | 42(62.1) | 25(56.8) | 0.578 |
| History of Ischemic CM, n(%) | 22(32.3) | 28(59.1) | 0.005 |
| Acute HF hospitalization, n(%) | 41(60.6) | 38(86.4) | 0.004 |
| ACS hospitalization, n(%) | 18(25.8) | 9(20.5) | 0.521 |
| SCA hospitalization, n(%) | 15(21.2) | 5(11.4) | 0.181 |
| NYHA class III-IV, n(%) | 35(51.5) | 33(75.0) | 0.013 |
| ICD/CRT-D, n(%) | 26(37.9) | 27(61.4) | 0.016 |
| Beta-blockers, n(%) | 45(71.0) | 26(65.0) | 0.526 |
| Amiodarone, n(%) | 15(22.6) | 15(37.5) | 0.103 |
| Mechanical ventilation, n(%) | 2(3.0) | 2(4.6) | 0.677 |
| Inotropic medications, n(%) | 21(31.8) | 20(45.5) | 0.147 |
| Baseline LVEF(SD), % | 33.4(18.7) | 22.7(15.2) | 0.0027 |
| Heart rate±SD, bpm | 90.3(22.9) | 89.0(20.4) | 0.750 |
| QTc (SD), ms | 455.1(33.5) | 442.0(52.2) | 0.665 |
| Mechanical alternans, n(%) | 12(18.2) | 16(36.4) | 0.032 |
| TWA, n(%) | 17(25.8) | 16(36.4) | 0.234 |
There was no difference in the percentage of patients with ICD/CRT-D implanted amongst alternans groups (Table 1). Importantly, study participants died regardless of having ICD/CRT-D implanted. Moreover, significantly larger percentage of patients with the primary endpoint had an implanted ICD/CRT-D (Table 3), as compared to end-point-free group.
Association between alternans and mortality
There were no statistically significant differences in the presence and characteristics of MA and TWA in patients with and without the primary endpoint (Table 4). In univariable Kaplan-Meier survival analysis, MA (Figure 3A) was associated with the primary outcome. Global test (P=0.484), as well as the test for MA (P=0.141), confirmed that proportional hazards assumption has been met. In univariable Cox regression, association of MA with primary outcome was borderline [HR1.84; 95%CI 1.00–3.41); P=0.050). The presence of sustained TWA (without consideration of simultaneous sustained MA, Figure 3B) was not significantly associated with the primary endpoint (HR 1.42; (95%CI 0.77–2.64); P=0.260). An interaction between MA and TWA was observed. A univariable Kaplan-Meier survival analysis showed differences in cumulative survival in patients with and without alternans (Figure 3). The worst survival was observed in patients with both MA and TWA at baseline, whereas patients without alternans had an intermediate probability of survival. Surprisingly, all patients with lone TWA remained free from primary endpoint at the end of the follow-up period (Figure 3C), although due to the very small group size (N=5) observed differences cannot be considered conclusive and should be validated in another prospective study.
Table 4.
Comparison of alternans in patients with and without primary endpoint
| Characteristic | Mechanical alternans at baseline |
P | |
|---|---|---|---|
| Alive (n=12) | Dead (n=16) | ||
| Amplitude of MA, median(IQR), mmHg | 4.1(2.7–9.3) | 4.7(1.5–21.1) | 0.816 |
| Number of MA beats, median(IQR),n | 31(25–64) | 39(29–56) | 0.546 |
| Mean Heart rate at MA(SD), bpm | 85.7(21.4) | 83.3(17.7) | 0.757 |
| MA at max HR, median(IQR), mmHg | 17.3(9.6–32.9) | 15.7(7.8–49.2) | 0.963 |
| TWA at baseline |
|||
| Alive (n=17) |
Dead (n= 16) |
||
| Amplitude of TWA, median(IQR), μV | 29(20–49) | 37(26–64) | 0.235 |
| Number of TWA beats, median(IQR),n | 25(22–29) | 24(21–29) | 0.515 |
| Mean Heart rate at TWA(SD), bpm | 76.9(15.3) | 68.6(19.2) | 0.184 |
| Mean QTc at TWA(SD), ms | 395(142) | 443(154) | 0.362 |
| TWA at max HR, median(IQR), μV | 50(35–59) | 98(44–221) | 0.119 |
Figure 3:
Kaplan‐Meier curves for the probabilities of the primary endpoint (all‐cause death or heart transplant) in patients with and without (A) MA, (B) TWA, (C) MA and TWA.
In the Cox regression analysis after adjustment for LVEF and NYHA class, the presence of MA was associated with a nearly tripled risk of death or transplantation (hazard ratio [HR] 2.81 (95%CI 1.43–5.50); P=0.003). In the Cox model that included MA and LVEF, dichotomized at 35%, low EF was a stronger predictor than MA: MA HR 2.00 (95%CI 1.07 – 3.71), P=0.029; LVEF≤35% HR 2.69 (95%CI 1.28–5.63), P=0.009. In the final multivariable Cox model (Table 5), MA remained a significant predictor of the primary outcome [HR 2.12 (95%CI 1.13–3.98); P=0.020].
Table 5.
Univariable and multivariable Cox regression hazard ratios
| Univarible |
Multivariable |
|||
|---|---|---|---|---|
| Predictor | Hazard ratio (95% CI) | P value | Hazard ratio (95% CI) | P value |
| MA presence | 1.84(1.00–3.41) | 0.050 | 2.12(1.13–3.98) | 0.020 |
| Age, y | 1.00(0.98–1.02) | 0.868 | - | |
| Female sex | 0.63(0.34–1.17) | 0.146 | - | |
| Black race | 0.97(0.57–1.65) | 0.912 | - | |
| NYHA III-IV class | 2.12(1.14–3.97) | 0.018 | 1.91(0.87–4.15) | 0.104 |
| LVEF ≤35% | 2.36(1.24–4.51) | 0.009 | 1.72(0.73–4.04) | 0.217 |
| ICD/CRT-D device implanted | 1.68(0.98–2.87) | 0.058 | 1.40(0.72–2.74) | 0.322 |
| Non-ischemic cardiomyopathy | 0.66(0.35–1.25) | 0.207 | - | |
| Heart rate≥90 bpm | 1.54(0.90–2.61) | 0.114 | - | |
Discussion
This prospective cohort study of patients with acute HF admitted to the ICCU showed that MA was independently associated with a 2–3-fold increased risk of all-cause death or heart transplant short-term post-discharge. Importantly, MA was associated with the primary outcome after adjustment for LVEF, NYHA class, and implanted ICD/CRT-D. If validated in another prospective study, MA could become a valuable tool for risk stratification in AHHF patients.
Importantly, study showed that during at least 12 hours of recording, sustained MA was periodically accompanied by sustained TWA, but an opposite statement does not hold true. TWA may be present as “lone” TWA, without MA. Further studies of coupled and uncoupled TWA and MA are needed in order to improve accuracy of the prediction of outcomes.
Post-discharge outcomes in ICCU patients
AHHF is characterized by extremely high short-term post-discharge mortality. In our study, all-cause mortality (39.6%) during 10 months post-discharge (4.0% per month) was comparable with high-risk subgroup of patients as reported in OPTIMIZE-HF21. Our finding of the independent association between MA and post-discharge mortality suggests that after validation MA could become an important prognostic tool and help to identify a population of patients that more likely benefit from advanced HF support (ventricular assist devices, VADs), rather than from ICDs. Importantly, even if LVEF was a stronger predictor of outcomes in this study, MA remained predictive after adjustment for LVEF, NYHA class, and implanted ICD/CRT-D. The high frequency of large MA amplitude (>10 mmHg) observed represents another advantage of MA assessment and provides an opportunity to recognize MA phenomenon without any especial equipment.
Mechanical alternans
Our study is the first relatively large prospective observational cohort study of MA in AHHF patients admitted to the ICCU. It was shown that MA is associated with adverse prognosis in patients with idiopathic dilated cardiomyopathy22. Development of MA was linked to deteriorating pump function in HF patients12;23. Correlation has been demonstrated between TWA and systolic pressure alternans detected noninvasively (via finger photoplethysmography)15, and cases of ventricular fibrillation24 in patients with MA have been described.
Yet, the mechanism of MA is complex, and is not completely understood14. Beat-to-beat alternations in stroke volume can result from the variations in preload and ventricular compliance. Reduced stroke volume results in elevated end diastolic volume for the next contraction, which, in turn, results in increased stroke volume and, therefore, increased systolic pressure with the next beat25. The strong beat suggests prolonged systole26, a shortened diastolic filling time and reduced end-diastolic volume of the subsequent weak beat, and alternans persists.
The second mechanism could be explained by the existence of the groups of myocytes that are electrically activated in an alternating fashion (e.g. due to differences in refractory period in action potential duration alternans). Therefore, these cells are synchronized and mechanically activated in the same alternating pattern. Alternations in action potential morphology27;28 can lead to alternation in the contractile strength of each cell. Theoretically, this mechanism could be responsible for the manifestation of both MA and TWA in this study.
The predictive value of TWA remains controversial8. The results of our study suggest that a simultaneous assessment of both MA and TWA may be needed to improve the predictive value of alternans. The very small subgroup of lone TWA (n=5) observed in our study does not allow any conclusive statement. However, possible interaction between TWA and MA indicate that lone TWA versus TWA coupled with MA can carry different risks. We speculate that when coupled with MA, TWA carries a high risk of pump failure progression, while lone TWA is associated with ventricular arrhythmias. We further speculate that AHHF patients with lone TWA would likely benefit from ICD/CRT-D whereas AHHF patients with MA would likely benefit from advanced heart failure support (e.g VAD). Further studies of simultaneously measured TWA and MA are needed to test these hypotheses.
Limitations
Several limitations of this study need to be acknowledged. First, the study population was heterogeneous, and even though this study is the largest prospective study of MA, statistical power was not sufficient for subgroup analysis. However, this study provided important prognostic information for the heterogeneous critical care patients studied. Additional prospective evaluation of ambulatory HF patients is warranted. Importantly, small subgroup of lone TWA patients (n=5) warrants further evaluation. It is possible that our methodology of MA detection was not sensitive enough and therefore we missed MA in 5 patients with lone TWA. It is equally possible that our methodology for TWA detection was overly sensitive, and we detected sub-threshold TWA. Important differences in outcomes warrant simultaneous assessment of both MA and TWA in future studies. Study participants were on beta-blockers and/or amiodarone, and/or inotropes at the time of alternans evaluation, which might affect alternans. However, this observational study represents a real-life scenario, and in spite of these limitations, the study showed predictive value of alternans.
Supplementary Material
Clinical Perspective.
Acute hospitalized heart failure (AHHF) is a global pandemic with high post-discharge mortality and readmission rates. Competing risk of sudden cardiac death due to ventricular arrhythmias, vs. pump failure death brings difficult management choices. We conducted a prospective observational cohort study of AHHF patients admitted to the Intensive Cardiac Care Unit. Study showed that the risk of death or heart transplant during 10 months post-discharge is twice higher in patients with mechanical pulsus alternans than in AHHF patients without alternans, even after adjustment for left ventricular ejection fraction, New York Heart Association class, and implanted ICD/CRT-D. Large amplitude (>10 mmHg) mechanical alternans could be easily recognized without any special equipment. At the same time, T-wave alternans (TWA) did not predict the outcome in this study. In addition, this study showed that mechanical alternans was often accompanied by TWA, with lone TWA events observed in a small group of patients. When coupled with mechanical alternans, TWA carried a high risk of pump failure progression, while lone TWA was associated with ventricular arrhythmias. Further studies of simultaneously measured TWA and mechanical alternans are needed. If validated in another prospective study, mechanical alternans could become a valuable tool for risk stratification in AHHF.
Acknowledgement
The authors would like to thank Durgesh Das, BS, and Eun Sun Choi, BS, for helping with the data analysis.
Funding Sources
Study was supported by Medtronic, Inc. as an Investigator-initiated Research Project.
Conflict of Interest Disclosures
Study was supported by Medtronic, Inc. as an Investigator-initiated Research Project (PI Tereshchenko).
References
- 1.Gheorghiade M, Vaduganathan M, Fonarow GC, Bonow RO. Rehospitalization for heart failure: problems and perspectives. J Am Coll Cardiol 2013;61:391–403. [DOI] [PubMed] [Google Scholar]
- 2.Gheorghiade M and Peterson ED. Improving postdischarge outcomes in patients hospitalized for acute heart failure syndromes. JAMA 2011;305:2456–2457. [DOI] [PubMed] [Google Scholar]
- 3.Gheorghiade M, Shah AN, Vaduganathan M, Butler J, Bonow RO, Rosano GM, Taylor S, Kupfer S, Misselwitz F, Sharma A, Fonarow GC. Recognizing hospitalized heart failure as an entity and developing new therapies to improve outcomes: academics’, clinicians’, industry’s, regulators’, and payers’ perspectives. Heart Fail. Clin 2013;9:285–290. [DOI] [PubMed] [Google Scholar]
- 4.O’Connor CM, Miller AB, Blair JE, Konstam MA, Wedge P, Bahit MC, Carson P, Haass M, Hauptman PJ, Metra M, Oren RM, Patten R, Pina I, Roth S, Sackner-Bernstein JD, Traver B, Cook T, Gheorghiade M. Causes of death and rehospitalization in patients hospitalized with worsening heart failure and reduced left ventricular ejection fraction: results from Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with Tolvaptan (EVEREST) program. Am Heart J 2010;159:841–849. [DOI] [PubMed] [Google Scholar]
- 5.Smith JM, Clancy EA, Valeri CR, Ruskin JN, Cohen RJ. Electrical alternans and cardiac electrical instability. Circulation 1988;77:110–121. [DOI] [PubMed] [Google Scholar]
- 6.Verrier RL and Nearing BD. Ambulatory ECG monitoring of T-wave alternans for arrhythmia risk assessment. J Electrocardiol. 2003;36 Suppl:193–197. [DOI] [PubMed] [Google Scholar]
- 7.Verrier RL, Klingenheben T, Malik M, El-Sherif N, Exner DV, Hohnloser SH, Ikeda T, Martinez JP, Narayan SM, Nieminen T, Rosenbaum DS. Microvolt T-wave alternans physiological basis, methods of measurement, and clinical utility-consensus guideline by international society for holter and noninvasive electrocardiology. J. Am. Coll. Cardiol 2011;58:1309–1324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Jackson CE, Myles RC, Tsorlalis IK, Dalzell JR, Rocchiccioli JP, Rodgers JR, Spooner RJ, Greenlaw N, Ford I, Gardner RS, Cobbe SM, Petrie MC, McMurray JJ. Spectral microvolt T-wave alternans testing has no prognostic value in patients recently hospitalized with decompensated heart failure. Eur. J. Heart Fail 2013;15:1253–1261. [DOI] [PubMed] [Google Scholar]
- 9.Gupta A, Hoang DD, Karliner L, Tice JA, Heidenreich P, Wang PJ, Turakhia MP. Ability of microvolt T-wave alternans to modify risk assessment of ventricular tachyarrhythmic events: a meta-analysis. Am Heart J 2012;163:354–364. [DOI] [PubMed] [Google Scholar]
- 10.Traube L Ein fall von pulsus bigeminus nebst bemerkungen tiber die lebershwellungen bei klappenfehlern und uber acute leberatrophic. BerKlin Wschr. 1872;9:889. [Google Scholar]
- 11.Adler D, Wong AY, Mahler Y. Model of mechanical alternans in the mammalian myocardium. J. Theor. Biol 1985;117:563–577. [DOI] [PubMed] [Google Scholar]
- 12.Kodama M, Kato K, Hirono S, Okura Y, Hanawa H, Ito M, Fuse K, Shiono T, Watanabe K, Aizawa Y. Mechanical alternans in patients with chronic heart failure. J. Card Fail 2001;7:138–145. [DOI] [PubMed] [Google Scholar]
- 13.Hirashiki A, Izawa H, Somura F, Obata K, Kato T, Nishizawa T, Yamada A, Asano H, Ohshima S, Noda A, Iino S, Nagata K, Okumura K, Murohara T, Yokota M. Prognostic value of pacing-induced mechanical alternans in patients with mild-to-moderate idiopathic dilated cardiomyopathy in sinus rhythm. J. Am. Coll. Cardiol 2006;47:1382–1389. [DOI] [PubMed] [Google Scholar]
- 14.Clancy EA, Smith JM, Cohen RJ. A simple electrical-mechanical model of the heart applied to the study of electrical-mechanical alternans. IEEE Trans. Biomed. Eng 1991;38:551–560. [DOI] [PubMed] [Google Scholar]
- 15.Selvaraj RJ, Suszko A, Subramanian A, Mak S, Wainstein R, Chauhan VS. Microscopic systolic pressure alternans in human cardiomyopathy: noninvasive evaluation of a novel risk marker and correlation with microvolt T-wave alternans. Heart Rhythm 2011;8:236–243. [DOI] [PubMed] [Google Scholar]
- 16.Lab MJ and Lee JA. Changes in intracellular calcium during mechanical alternans in isolated ferret ventricular muscle. Circ Res 1990;66:585–595. [DOI] [PubMed] [Google Scholar]
- 17.Kihara Y and Morgan JP. Abnormal Cai2+ handling is the primary cause of mechanical alternans: study in ferret ventricular muscles. American Journal of Physiology - Heart andCirculatory Physiology 1991;261:H1746–H1755. [DOI] [PubMed] [Google Scholar]
- 18.Surawicz B and Fisch C. Cardiac alternans: diverse mechanisms and clinical manifestations. J. Am. Coll. Cardiol 1992;20:483–499. [DOI] [PubMed] [Google Scholar]
- 19.Burattini L, Bini S, Burattini R. Correlation method versus enhanced modified moving average method for automatic detection of T-wave alternans. Comput. Methods Programs Biomed 2010;98:94–102. [DOI] [PubMed] [Google Scholar]
- 20.Nearing BD and Verrier RL. Modified moving average analysis of T-wave alternans to predict ventricular fibrillation with high accuracy. J Appl. Physiol 2002;92:541–549. [DOI] [PubMed] [Google Scholar]
- 21.O’Connor CM, Abraham WT, Albert NM, Clare R, Gattis SW, Gheorghiade M, Greenberg BH, Yancy CW, Young JB, Fonarow GC. Predictors of mortality after discharge in patients hospitalized with heart failure: an analysis from the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF). Am Heart J 2008;156:662–673. [DOI] [PubMed] [Google Scholar]
- 22.Hirashiki A, Izawa H, Cheng XW, Unno K, Ohshima S, Murohara T. Dobutamine-induced mechanical alternans is a marker of poor prognosis in idiopathic dilated cardiomyopathy. Clin. Exp. Pharmacol. Physiol 2010;37:1004–1009. [DOI] [PubMed] [Google Scholar]
- 23.Kodama M, Kato K, Hirono S, Hanawa H, Okura Y, Ito M, Fuse K, Shiono T, Tachikawa H, Hayashi M, Abe S, Yoshida T, Aizawa Y. Changes in the occurrence of mechanical alternans after long-term beta-blocker therapy in patients with chronic heart failure. Jpn. Circ. J 2001;65:711–716. [DOI] [PubMed] [Google Scholar]
- 24.Kodama M, Ramadan MM, Mitsuma W, Okamura K, Kashimura T, Tachikawa H, Ito M, Fuse K, Hirono S, Okura Y, Kato K, Hanawa H, Watanabe K, Aizawa Y. Ventricular fibrillation in two cases with dilated cardiomyopathy and mechanical alternans. Pacing Clin. Electrophysiol. 2005;28:1347–1349. [DOI] [PubMed] [Google Scholar]
- 25.Lab MJ and Seed WA. Pulsus alternans. Cardiovasc. Res 1993;27:1407–1412. [DOI] [PubMed] [Google Scholar]
- 26.Carlson CJ and Rapaport E. Postextrasystolic pulsus alternans and heart rate. Am J Physiol 1984;246:H245–H249. [DOI] [PubMed] [Google Scholar]
- 27.Narayan SM, Bayer JD, Lalani G, Trayanova NA. Action potential dynamics explain arrhythmic vulnerability in human heart failure: a clinical and modeling study implicating abnormal calcium handling. J. Am. Coll. Cardiol 2008;52:1782–1792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bayer JD, Narayan SM, Lalani GG, Trayanova NA. Rate-dependent action potential alternans in human heart failure implicates abnormal intracellular calcium handling. Heart Rhythm 2010;7:1093–1101. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.