Time Trends of Left Ventricular Ejection Fraction and Myocardial Deformation Indices in a Cohort of Women with Breast Cancer Treated with Anthracyclines, Taxanes, and Trastuzumab


Trastuzumab, a HER2 monoclonal antibody, has transformed the prognosis of patients with the aggressive HER2-positive breast cancer type. Trastuzumab augments the cardiotoxic effects of anthracyclines, but its effect is thought to be at least partially reversible. The objective of this study was to examine the time trends of left ventricular (LV) size and function in a cohort of women treated with anthracyclines and trastuzumab.


Twenty-nine patients >18 years of age with first-time breast cancer treated with anthracyclines and trastuzumab were monitored using echocardiography before, at the completion of, and at a median follow-up of 24.7 months (interquartile range, 15.9–34 months) after the end of their cancer treatment. LV volume, LV ejection fraction, and global peak systolic longitudinal strain and strain rate were measured in the apical four- and two-chamber views. Left ventricular ejection fraction was measured using a modified Simpson’s biplane method.


LV end-diastolic and end-systolic volumes increased at the end of treatment compared with baseline and did not recover during follow-up. Left ventricular ejection fraction, strain, and strain rate decreased at the end of treatment compared with baseline (from 64 ± 6% to 59 ± 8%, from −20.0 ± 2.5% to −17.6 ± 2.6%, and from −1.26 ± 0.23 to −1.13 ± 0.16 sec −1 , respectively; P < .05 for all parameters) and remained decreased at follow-up.


LV dilation and subclinical impairment in cardiac function persists >2 years after the end of anthracycline and trastuzumab treatment, without significant recovery after trastuzumab cessation, suggestive of long-term underlying cardiac damage and remodeling.

New developments in cancer therapies and altered treatment approaches have increased the survival of patients with cancer. Unfortunately, this increase in survival length, associated with the aging of the patients, is paralleled by a concurrent increase in the rate of cardiovascular complications, particularly if there is a known history of heart disease. Anthracyclines have long been recognized as cardiotoxic, with a prevalence of 1% of all cases of cardiomyopathy. Up to 50% of all patients treated with anthracyclines will show some degree of cardiac dysfunction 10 to 20 years after chemotherapy, and 2% to 5% of them will go on to develop overt heart failure. However, anthracyclines continue to be used widely, as they are effective in the treatment of aggressive cancers, with 60,000 to 80,000 patients treated every year in the United States with anthracyclines.

Another class of chemotherapeutic agents that has an important role in the treatment of breast cancer is taxanes, which are antimicrotubule agents that prevent cell division. Taxanes, which include paclitaxel and docetaxel, have now become a standard component of the initial chemotherapy for women with breast cancer, because they are among the most active chemotherapeutic agents, particularly for patients with metastatic disease. Some of the cardiotoxicity in these patients may also be related to the taxanes, although this effect appears more modest than that of anthracyclines or trastuzumab.

HER2 (Her2/ neu ) is a transmembrane receptor tyrosine kinase that plays a role in mediating the growth, differentiation, and survival of cells. HER2 protein overexpression or the amplification of the HER2 oncogene occurs in the tumor of 20% to 30% of patients with breast cancer, resulting in an aggressive form of breast cancer previously associated with decreased overall survival. Trastuzumab has transformed the prognosis of patients with HER2-positive breast cancer, but clinical trials have revealed significant cardiac side effects from trastuzumab, with a synergistic effect of trastuzumab and anthracyclines. Although trastuzumab-related cardiotoxicity is most often manifested by an asymptomatic decrease in left ventricular (LV) ejection fraction (LVEF) and less often by clinical heart failure, a recent meta-analysis of 11,991 women with HER2-positive early breast cancer showed that patients exposed to both anthracyclines and trastuzumab were at greater risk for severe heart failure and reduction in LVEF than those treated with non-trastuzumab-based adjuvant or neoadjuvant chemotherapy.

Anthracyclines and trastuzumab cause cardiac dysfunction through different mechanisms. Treatment with anthracyclines results in cardiomyocyte apoptosis and necrosis. In contrast, treatment with trastuzumab induces loss of myocyte contractility. Hence, trastuzumab is less likely to be associated with myocyte death and is presumed to cause temporary dysfunction, which is thought to be mostly reversible with treatment discontinuation.

The early detection of anthracycline-induced cardiotoxicity is important; once symptomatic, anthracycline-induced cardiomyopathy carries a worse prognosis than many other forms of cardiomyopathies, with a <50% rate of 2-year survival free of transplantation. Current detection of cardiotoxicity in the clinical setting is based primarily on monitoring LVEF. Detecting cardiotoxicity by a decrease in LVEF, however, may already be too late for intervention; 36% of patients with decreases in LVEF detected 1 month after the end of anthracycline treatment did not respond to heart failure therapy. Hence, there is some debate over the adequacy of LVEF as a measure of subclinical cardiotoxicity.

Myocardial deformation (strain) and deformation rate (strain rate) decrease earlier than LVEF in patients treated with anthracyclines. Additionally, their early decrease predicts subsequent decreases in LVEF. The overall objective of the present study was to examine the long-term effect of anthracyclines, taxanes and trastuzumab on LV remodeling and function. LV function was analyzed using LVEF and peak longitudinal strain and strain rate. For this purpose, we analyzed serial echocardiograms in a cohort of women with breast cancer before and after their treatment with anthracyclines, taxanes, and trastuzumab.


Patients were enrolled prospectively from two institutions (Massachusetts General Hospital, Boston, MA, and Jewish General Hospital, Montreal, QC, Canada). Patients >18 years of age diagnosed with HER2-overexpressing breast cancers were eligible if they were scheduled to receive treatment including anthracyclines and trastuzumab or trastuzumab after previous anthracycline treatment. Patients with LVEFs <50% before commencing treatment were excluded.

All patients provided written informed consent, which was approved by the institutional boards of the participating institutions. Patients were studied using echocardiography at baseline, before the initiation of their chemotherapy regimen; at the end of the trastuzumab treatment (15 months); and at a follow-up time point >12 months after the completion of their last dose of chemotherapy. The normal range for longitudinal strain was defined as −15.9% to −22.1%. Cardiotoxicity was defined on the basis of criteria set out by the Cardiac Review and Evaluation Committee of trastuzumab-associated cardiotoxicity, that is, either a reduction in LVEF of ≥5% to <55% with symptoms of heart failure (diagnosed by a cardiologist at the site) or an asymptomatic reduction of LVEF of ≥10% to <55%.


Transthoracic echocardiography was performed using the Vivid 7 or E9 (GE Healthcare, Milwaukee, WI). The same ultrasound machine was used to acquire all echocardiograms in each patient. LVEF was calculated from the apical four- and two-chamber views using a modified Simpson’s biplane method. LV mass was calculated on the basis of the formula outlined by the American Society of Echocardiography guidelines and indexed to body surface area. To measure myocardial strain, two-dimensional grayscale images were obtained in the apical four- and two-chamber views. The timing of aortic valve closure was obtained using pulsed-wave Doppler traces. The same software was used to analyze strain in all patients at all the different time points. All strain measurements and analysis were performed offline at the Cardiac Ultrasound Laboratory in Massachusetts General Hospital using EchoPAC (GE Healthcare). All sonographers involved in the acquisition of images for strain underwent appropriate training for the acquisition of images for strain analysis. Images were acquired according to a predetermined protocol to maintain consistency, and all images for strain measurements were acquired at a frame rate of 60 to 80 frames/sec. Patients who had technically difficult images for LV tracing (two patients) were excluded from the study. Global peak systolic longitudinal strain was calculated by averaging the values of peak systolic strain in the basal, midwall, and apical segments of the four- and two-chamber views. To minimize any inherent bias, LVEF and strain were measured separately by two different readers, each reader blinded to the results of the other parameter that they did not measure. LV volumes and LVEF were measured by S.B., who was blinded to the longitudinal systolic strain measurements. Peak systolic longitudinal strain and strain rate were measured by T.C.T., who was blinded to the results of the LVEF measurements. Both readers were blinded to the time points of the echocardiographic examination and the clinical history and treatment of the subjects. The interobserver and intraobserver variability for LV volumes, LVEF, and longitudinal strain are reported as mean error ± SD of 10 measurements in absolute values and in percentages. Intraobserver variability for LV end-diastolic volume (LVEDV) was −0.2 ± 0.9 mL (−0.2 ± 1.2%), for LV end-systolic volume (LVESV) was −0.3 ± 1.8 mL (−0.7 ± 4.0%), for LVEF was 0.04 ± 0.70% (0.05 ± 1.0%), and for longitudinal strain was 0.2 ± 0.6% (1.4 ± 3.5%). Interobserver variability for LVEDV was 8.1 ± 4.2 mL (9.2 ± 4.6%), for LVESV was 2.7 ± 3.2 mL (8.9 ± 11.7%), for LVEF was 0.07 ± 0.9% (0.1 ± 1.5%), and for longitudinal strain was −1.2 ± 0.9% (6.0 ± 4.9%).

Statistical Analysis

The normality of the distribution of the parameters studied was checked using the Shapiro-Wilk test. Parameters of interest were then compared at the three time points using a one-way analysis of variance for repeated measures. If the time effect was significant, the parameters at follow-up were compared with their baseline values and after trastuzumab treatment using contrast analysis. P values < .05 were defined as indicating statistical significance. Data are expressed as mean ± SD if parametric or as median (<25% quartile, >75% quartile; range) if nonparametric. All other statistical analyses were performed using the JMP statistical package (SAS Institute Inc, Cary, NC).



Thirty-one patients were prospectively enrolled in the study. Two were excluded because of technically difficult images for LV tracing. A final total of 29 patients who were prospectively enrolled in the study (mean age, 50.4 ± 10.4 years) were examined. The patients in this study were treated with either doxorubicin (cumulative dose, 240 mg/m 2 ) or epirubicin (cumulative dose, 300 mg/m 2 ) for 3 months, followed by weekly paclitaxel (80 mg/m 2 ) and trastuzumab (2 mg/kg) for 3 months and trastuzumab only (6 mg/kg) every 3 weeks for 9 more months. The median follow-up period was 24.7 months (15.9 months, 34.0 months; 12.1–46.9 months). The mean heart rate was 69 ± 11 beats/min at baseline and 69 ± 9 beats/min at the end of treatment, not significantly changed ( P = .87). The baseline characteristics of this population of patients are presented in Table 1 .

Table 1

Baseline characteristics of patients treated with anthracyclines, taxanes, and trastuzumab ( n = 29)

Variable Value
Age (yrs) 50.4 ± 10.4
Height (m) 1.68 ± 0 .18
Weight (kg) 68.1 ± 12.4
Body mass index (kg/m 2 ) 24.5 ± 5.4
Dose of anthracyclines
Doxorubicin (240 mg/m 2 ) 28 (97%)
Epirubicin (300 mg/m 2 ) 1 (3%)
Radiotherapy 18 (62%)
Left side 10 (34%)
Right side 6 (21%)
Both sides 2 (7%)
Side of breast cancer
Left side 15 (52%)
Right side 11 (38%)
Both sides 3 (10%)
Cardiac risk factors
Hypertension 7 (24%)
Hyperlipidemia 6 (21%)
Diabetes mellitus 0 (0%)
Smoker 3 (10%)
Cardiovascular therapy
Angiotensin-converting enzyme inhibitors 4 (14%)
β-blockers 1 (3%)

Data are expressed as mean ± SD or as number (percentage).

LV Dimensions and Volumes

LVEDV and LVESV had increased significantly by 12 ± 19% and 26 ± 34%, respectively, at the end of treatment, and neither returned to baseline levels ( Table 2 ). Interventricular septal thickness, posterior wall thickness, and LV mass were unchanged.

Table 2

Temporal changes in echocardiographic parameters at baseline, after the completion of chemotherapy, and at follow-up

Parameter Baseline End of chemotherapy Follow-up P (ANOVA)
LVEDD (mm) 43 ± 4 45 ± 3 45 ± 4 <.001
IVST (mm) 8 ± 1 8 ± 1 8 ± 1 .94
PWT (mm) 8 ± 1 8 ± 1 8 ± 1 .45
LV mass index (g/m 2 ) 60 ± 11 65 ± 12 69 ± 12 .021
LVEDV index (ml/m 2 ) 53 ± 7 59 ± 8 57 ± 9 .0097
LVESV index ml/m 2 ) 19 ± 4 23 ± 4 22 ± 5 .0003
LVEF (%) 64 ± 6 59 ± 6 59 ± 8 .0007
Peak systolic longitudinal strain (%) −20.0 ± 2.5 −17.6 ± 2.5 −17.6 ± 2.6 <.0001
Peak systolic longitudinal strain rate (sec −1 ) −1.26 ± 0.23 −1.09 ± 0.14 −1.13 ± 0.16 .0040

ANOVA , Analysis of variance; IVST , interventricular septal thickness; LVEDD , left ventricular end-diastolic diameter; LVEDV , left ventricular end-diastolic volume; PWT , posterior wall thickness.

Data are expressed as mean ± SD.

Statistically significant ( P < .05) compared with baseline.


LVEF had decreased by 8 ± 11% at the end of treatment and did not recover after >2 years of treatment completion ( Table 2 ). Nine patients (31%) had decreased LVEFs by >10% at the end of treatment. Two of these nine patients (22%) showed ≥10% recovery in LVEF at follow-up (although LVEF in only one of the two recovered completely). The change in LVEF between the end of treatment and at follow-up did not correlate with the time interval between the end of treatment and follow-up. Four patients (14%) had cardiotoxicity on the basis of Cardiac Review and Evaluation Committee criteria at the end of treatment, whereas seven patients (24%) were found to have cardiotoxicity at follow-up. Of the seven patients with cardiotoxicity at follow-up, only two had cardiotoxicity at the end of treatment. None of the patients reported any symptomatic heart failure at follow-up.

LV Peak Systolic Longitudinal Strain and Strain Rate

Overall peak systolic longitudinal myocardial strain and strain rate had decreased by 10 ± 15% and 11 ± 19%, respectively ( Table 2 ), at the end of treatment compared with baseline. Neither parameter recovered at follow-up. Sixteen of 29 patients (55%) had decreased peak systolic longitudinal strain by ≥10% from baseline to the end of treatment. Of these 16 patients, seven (44%) showed ≥10% (of baseline values) subsequent recovery of strain at follow-up (but only two showed recovery within 5% of baseline values). For strain rate, 20 of 29 patients (69%) had decreases of ≥10% from baseline at the end of treatment, of whom only seven of 20 (35%) recovered ≥10% (of baseline values) at follow-up. Of the seven, only one showed recovery within 5% of baseline values. Similar to LVEF, the changes in strain and strain rate between the end of treatment and follow-up did not correlate with the time interval between the end of treatment and follow-up.

Only gold members can continue reading. Log In or Register to continue

Apr 21, 2018 | Posted by in CARDIOLOGY | Comments Off on Time Trends of Left Ventricular Ejection Fraction and Myocardial Deformation Indices in a Cohort of Women with Breast Cancer Treated with Anthracyclines, Taxanes, and Trastuzumab

Full access? Get Clinical Tree

Get Clinical Tree app for offline access