Background
During the development of heart failure (HF), the changes of contraction timing pattern and temporal heterogeneity of segmental contraction happen early and may precede both symptomatic HF and the decrease in left ventricular ejection fraction (LVEF). In patients treated with anthracyclines, both symptomatic HF and the decrease of LVEF are detected once significant myocardial injury has occurred. The aim of the current study was to investigate whether changes in the timing of contraction can be detected early after anthracyclines therapy.
Methods
Forty-one women (50 ± 11 years old) with newly diagnosed breast cancer were prospectively enrolled in two centers and underwent an echocardiogram before and after anthracyclines. Peak longitudinal myocardial systolic strain was measured on the apical four- and two-chamber views. The time to peak systolic longitudinal strain (TP), ejection time (ET), isovolumic contraction time (IVCT), systolic time, and diastolic time were measured using strain curves and Doppler tracings and compared before and after anthracyclines. The heterogeneity of contraction (dyssynchrony) was measured by the SD of the TP of all segments.
Results
Anthracyclines treatment was associated with an increase in heart rate (HR) and a decrease in TP. TP was correlated with HR. TP/ET was independent of HR and inversely correlated to peak strain both at baseline and after anthracyclines. TP/ET increased after anthracyclines (1.26 ± 0.19 to 1.31 ± 0.22; P < .001), and this increase was correlated with the decrease in strain. The increase in TP/ET was due to an increase in IVCT/ET. A similar degree of dyssynchrony was found at baseline and after anthracyclines.
Conclusions
Anthracyclines treatment induces an increase in the duration of contraction, mainly by increasing the IVCT. This increase is correlated to the decrease in strain and may therefore have additional prognostic value.
Highlights
- •
The value of changes in the time pattern and temporal homogeneity of contraction in patients treated with anthracyclines is unknown.
- •
The TP/ET, a heart rate–independent parameter, was increased after anthracyclines treatment.
- •
The increase of IVCT/ET as well as the increase of TP/ET are correlated with the change in strain and might indicate cardiac dysfunction.
- •
The heterogeneity of contraction was not increased after anthracyclines (240 mg/m 2 ) in women with breast cancer.
Due to the early diagnosis and the improvement of therapies in breast cancer, more than 2.2 million women are breast cancer survivors in the United States. However, as the length of survival increases, more side effects of anticancer therapies can develop over time, in particular cardiotoxicity. The cardiotoxicity of anthracyclines is well recognized and may lead to symptomatic heart failure. Once symptoms are noted, the cardiotoxicity is rarely reversible and the prognosis is poor. Therefore, research interest is high to find early markers of cardiac injury or dysfunction that occur before the symptoms and are predictive of prognosis. The peak amplitude of the longitudinal myocardial deformation (or strain) measured after anthracyclines therapy has shown promise in the prediction of subsequent LV dysfunction.
It is well recognized that the time pattern of contraction is an early sign of left ventricular (LV) dysfunction. In ischemia, for example, contraction is delayed during early-to-midsystole. A period of mid-to-late systolic lengthening is then noted, followed by late systolic (or postsystolic) contraction. This pattern can be studied using the timing of strain. The time to peak strain in particular is a sensitive index of dysfunction, and its duration is correlated with the degree of injury. Lengthening of the time to peak longitudinal systolic strain (TP) has also been described in patients with heart failure. The time to peak strain, however, shortens as heart rate (HR) increases. Ejection time (ET) has been used previously to take into account changes in HR. To normalize for the anthracyclines-induced increase in HR, the ratio of TP/ET was used. Additionally, LV dysfunction is accompanied by spatial heterogeneity of the temporal pattern of contraction (dyssynchrony). Dyssynchrony can be measured in a variety of ways, including M-Mode, tissue Doppler, and more recently strain imaging. The value of dyssynchrony as a prognostic marker is well known. Most recently, in a study of 75 patients after myocardial infarction, a 10-msec increase of the SD of the TP between segments by speckle-tracking imaging was associated with a higher risk of ventricular tachycardia. The value of changes in the time pattern and homogeneity of contraction in patients treated with anthracyclines is unknown.
The objective of the present study was to investigate whether anthracyclines therapy induced a detectable change in the timing pattern of myocardial contraction (measured with strain imaging and pulsed Doppler) and in the synchrony of contraction between segments (measured with strain imaging).
Methods
Patients of at least 18 years old newly diagnosed with breast cancer with positive human epidermal growth factor receptor 2 and scheduled to receive adjuvant therapy including anthracyclines were enrolled at two institutions (Massachusetts General Hospital, Boston, Massachusetts; and the Jewish Hospital, Montreal, Canada). Patients who had a baseline left ventricular ejection fraction (LVEF) < 50% were excluded. The study was approved by the Internal Review Board of the two participating institutions. After a signed informed consent, patients underwent an echocardiogram before chemotherapy and at the completion of anthracyclines therapy (19 ± 9 days after the last anthracyclines cycle). The timing parameters were measured at baseline and at the completion of the anthracyclines therapy.
Echocardiograms
Transthoracic echocardiograms were acquired using the Vivid 7 or E9 (GE Healthcare, Milwaukee, WI). The same ultrasound machine was used to acquire all echocardiograms in each patient. All echocardiograms were analyzed by two readers (M.S.-C. for the LVEF and K.-H.C. for all other measurements). The readers were blinded to each other’s measurements and to the patient’s visit number. Left atrial anteroposterior diameter was measured in the parasternal long-axis view. LV end-diastolic volume and LVEF were calculated from the apical four- and two-chamber views using a modified Simpson biplane method. LV peak longitudinal myocardial systolic strain was measured using a speckle-tracking algorithm (Echopac; GE Medical) on the apical four- and two-chamber views (total of 12 segments). Because of the difficulty in reliably tracking the endocardial border of the three-chamber view in nine patients (22%), the peak systolic longitudinal strain (LS) was defined as the average of the strain and time to peak strain values obtained in the four- and two- chamber views (12 segments).
The TP of one segment was used to characterize the duration of contraction and was defined as the time interval between the null value of strain to the peak of systolic LS in that individual segment ( Figure 1 ). The TP of the 12 segments was also averaged to obtain a global TP. The segments were also stratified into basal, midventricular, and apical regions of the LV and between four-chamber and two-chamber views. The TP has been reported to be positively associated with HR. As HR was increased after anthracyclines, TP was adjusted for HR by normalizing it by ET.
Cardiac events were also measured using pulsed wave Doppler with the sample volume placed in the LV outflow tract just below the aortic valve leaflets. The isovolumic contraction time (IVCT) was defined as the duration between the onset of the QRS complex and the start of the ejection period. The ET was defined as the time interval from the aortic valve opening (onset of flow) to the aortic valve closure (end of the flow) by pulsed Doppler of the aortic valve. The systolic time (ST) was defined as the interval between the onset of the QRS complex on the surface electrocardiogram and the closure of the aortic valve. The diastolic time (DT) was defined as the period between the aortic valve closure and the onset of the QRS ( Figure 1 ). Synchronization of two-dimensional echo images with pulsed Doppler was based on aligning the onset of the QRS from the electrocardiogram in each acquisition from cardiac cycles with matched HRs.
LV dyssynchrony was measured using the SD of the TP values. In addition, the SDs of ET among these 12 segments were also calculated.
Statistics
The study was designed as a prospective multicenter study evaluating the predictive value of echocardiography in patients with breast cancer treated with anthracyclines. The present study analyzed the timing events in a subset of patients (two institutions). All data are presented as mean ± SD or as proportions. To examine the relationships between TP and HR, LVEF, and the peak LS, the generalized estimating equation approach was used to take into account the within-subject correlation due to the analysis of multiple segments in a single patient. The relationships between TP/ET, HR, LVEF, and peak LS were examined separately. The significance of the relationships was corrected for multiple analyses (Bonferonni correction). One-way analysis of variance (ANOVA) for repeated measures before and after treatment was used to detect the differences of TP, TP normalized by ET (TP/ET) over time. Cases were grouped by patient or by patient and view (or region) in separate analyses. Standard paired t -tests were used to compare pulsed Doppler parameters before and after anthracyclines. An ANOVA was used to compare the baseline or postanthracyclines TP and TP/ET in all segments. The SD and coefficient of variation were calculated to assess the LV dyssynchrony. Interobserver and intraobserver variabilities were calculated as the difference between two observations by the same reader (K.-H.C.) or two separate readers (K.-H.C. and B.M.B.L.A) in 10 patients divided by the means of the observations and were expressed as percentages. The intraclass coefficients (ICCs)were also reported. The intraobserver variability of the TP was 0.4% ± 8.1% with an ICC of 0.952 ( P < .001), and the interobserver variability was 0.8% ± 9.5% (ICC of 0.925, P < .001).
Results
Patient Baseline Clinical Characteristics
Forty-one patients were studied. All patients received 240 mg/m 2 of doxorubicin (60 mg/m 2 every 3 weeks). None of the patients received radiotherapy. Baseline characteristics are presented in Table 1 .
| Clinical data | |
| Age (years) | 50 ± 11 |
| Body mass index (kg/m 2 ) | 24 ± 4 |
| Systolic blood pressure (mmHg) | 120 ± 18 |
| Diastolic blood pressure (mmHg) | 73 ± 10 |
| HR (beats per minute) | 69 ± 9 |
| LVEF (%) | 64 ± 4 |
| Side of breast cancer | |
| Right (%) | 21 (51) |
| Left (%) | 17 (42) |
| Both (%) | 3 (7) |
| Cardiovascular risk factors | |
| Diabetes mellitus (%) | 0 (0) |
| Hypertension (%) | 11 (27) |
| Hyperlipidemia (%) | 8 (20) |
| Smoking (%) | 5 (11) |
| Cardiovascular treatment | |
| Angiotensin-converting enzyme inhibitors (%) | 3 (7) |
| β-blockers (%) | 2 (5) |
Left Atrial Dimension and LV Volume
There were no significant differences in the left atrial anteroposterior diameter between baseline and after anthracyclines (32 ± 12 mm before and 32 ± 9 mm after anthracyclines; P = .99) or in the LV end-diastolic volume (74 ± 13 mL before and 75 ± 16 mL after anthracyclines; P = .75).
Variation of Global Time to Peak Strain with Anthracyclines Treatment
The global TP was 399 ± 68 msec before chemotherapy ( Table 2 ). Baseline HR, peak LS, and LVEF were all independently associated with TP ( Table 3 ). The higher the HR, LVEF, or the absolute value of strain at baseline, the shorter the TP. For example, each increase of 1 beat per minute in the HR was associated with a shortening of 2.5 msec in the TP. The global TP trended to decrease after treatment (393 ± 69 msec; P = .05 in Table 2 ). Both LVEF and the peak LS were significantly reduced after anthracyclines (LVEF, 64% ± 6% to 62% ± 5%, P = .043; LS, 22.3% ± 6.4% to 20.8% ± 5.3%, P < .001), whereas HR increased and blood pressure decreased ( Tables 4 and 5 ). Postanthracyclines HR and peak LS were both independently associated with postanthracyclines TP. Similarly, the change in TP was associated with the change in HR ( P < .001). By multivariate analysis, the change in HR was the only variable independently associated with the change in TP ( Table 3 ).
| Segment | Baseline | After | P value |
|---|---|---|---|
| Overall global (msec) | 399 ± 68 | 393 ± 69 | .054 |
| Four chamber | |||
| Global (msec) | 401 ± 71 | 396 ± 75 | .118 |
| Basal septal (msec) | 423 ± 75 | 439 ± 72 | |
| Midseptal (msec) | 369 ± 38 | 382 ± 51 | |
| Apical septal (msec) | 393 ± 48 | 385 ± 47 | |
| Apical lateral (msec) | 419 ± 108 | 405 ± 119 | |
| Midlateral (msec) | 394 ± 58 | 381 ± 78 | |
| Basal lateral (msec) | 408 ± 68 | 386 ± 78 | |
| Two chamber | |||
| Global (msec) | 397 ± 66 | 390 ± 61 | .255 |
| Basal inferior (msec) | 425 ± 93 | 421 ± 81 | |
| Midinferior (msec) | 379 ± 50 | 362 ± 39 | |
| Apical inferior (msec) | 387 ± 47 | 381 ± 48 | |
| Apical anterior (msec) | 389 ± 57 | 380 ± 60 | |
| Midanterior (msec) | 384 ± 60 | 378 ± 54 | |
| Basal anterior (msec) | 419 ± 69 | 421 ± 54 | |
| Region | |||
| Basal (msec) | 419 ± 77 | 417 ± 74 | .909 |
| Mid (msec) | 382 ± 52 | 376 ± 48 | .035 ∗ |
| Apical (msec) | 397 ± 70 | 388 ± 74 | .127 |
| Univariate P value | Multivariate P value | |
|---|---|---|
| Baseline TP | ||
| Baseline HR | <.001 ∗ | <.001 ∗ |
| Baseline LS | .003 ∗ | .014 ∗ |
| Baseline LVEF | .011 | .003 ∗ |
| Postanthracyclines TP | ||
| PA HR | <.001 ∗ | <.001 ∗ |
| PA LS | <.001 ∗ | <.001 ∗ |
| PA LVEF | .233 | |
| Δ TP | ||
| Δ HR | <.001 ∗ | <.001 ∗ |
| Δ LS | .022 | .075 |
| LVEF | .353 | .168 |
| Baseline TP/ET | ||
| Baseline HR | .239 | |
| Baseline LS | .006 ∗ | .008 ∗ |
| Baseline LVEF | .002 ∗ | .001 ∗ |
| Postanthracyclines TP/ET | ||
| PA HR | .973 | |
| PA LS | <.001 ∗ | <.001 ∗ |
| PA LVEF | .053 | .169 |
| Δ TP/ET | ||
| Δ HR | .259 | |
| Δ LS | .028 | .028 ∗ |
| Δ LVEF | .458 | |
| Baseline | After anthracyclines | Change | P value | |
|---|---|---|---|---|
| Systolic blood pressure (mmHg) | 121 ± 19 | 110 ± 15 | 10 ± 14 | <.001 ∗ |
| Diastolic blood pressure (mmHg) | 73 ± 10 | 70 ± 9 | 3 ± 7 | .012 ∗ |
| HR (beats per minute) | 66 ± 8 | 71 ± 9 | −5 ± 12 | .013 ∗ |
| ET (msec) | 317 ± 22 | 301 ± 29 | 16 ± 30 | .002 ∗ |
| Baseline | After anthracyclines | Change | P value | |
|---|---|---|---|---|
| ST (msec) | 391 ± 27 | 378 ± 28 | −13 ± 35 | .029 ∗ |
| ST/ET | 1.23 ± 0.05 | 1.26 ± 0.06 | 0.02 ± 0.06 | .019 ∗ |
| DT (msec) | 550 ± 123 | 485 ± 92 | −65 ± 118 | .002 ∗ |
| DT/ET | 1.72 ± 0.34 | 1.60 ± 0.26 | −0.12 ± 0.34 | .042 ∗ |
| IVCT (msec) | 74 ± 14 | 77 ± 15 | 2.8 ± 16.8 | .310 |
| IVCT/ET | 0.23 ± 0.05 | 0.26 ± 0.06 | 0.02 ± 0.06 | .017 ∗ |
The TP/ET before chemotherapy was not associated with HR; however, both peak LS and LVEF were independently associated with TP/ET value at baseline ( Table 3 ). The global TP/ET increased (1.26 ± 0.19 to 1.31 ± 0.16, P < .001) after anthracyclines treatment ( Table 6 ). The absolute value of the TP/ET after anthracyclines was associated with the absolute value of strain value after anthracyclines even after adjustment for HR ( P = .004). Additionally, the decrease in peak LS was associated with the increase in TP/ET ( Table 3 ).
