Aims: To evaluate whether repetitive assessment of systolic and diastolic cardiac function by dobutamine stress echocardiography (DSE) can predict anthracycline cardiotoxicity.
Methods and results: Thirty-one patients (age, 57±13 years, 22 male) were studied before chemotherapy, with follow-ups during, at the end, and 6 months after chemotherapy. Left ventricular (LV) function was assessed by two-dimensional (2D) echocardiographic wall motion score index (WMSI) and by Doppler echocardiography of mitral valve inflow at rest and during DSE. Radionuclide ventriculography was used as an independent reference for ejection fraction (EF). A reduction of EF ≥5% occurred in 17 patients (group A) at the last follow-up. Patients without decreased EF comprised group B. Early/late diastolic velocity of mitral inflow (E/A ratio) at rest was lower in group A (0.91±0.2 vs 1.28±0.3, P<0.001), and it was an independent predictor of cardiotoxicity (adjusted for baseline patient characteristics and parameters of systolic and diastolic function). At follow-up, WMSI at rest paralleled radionuclide EF. Contractile reserve at low-dose DSE was preserved in group A.
Conclusions: WMSI measured by 2D echocardiography parallels radionuclide EF at follow-up. Assessment of contractile reserve has no incremental value for the early detection of cardiotoxicity. A baseline abnormal E/A ratio is an independent predictor of anthracycline cardiotoxicity.
Anthracyclines are among the most widely used and effective antineoplastic agents. A growing number of patients treated with anthracyclines may have the potential for substantial morbidity and mortality owing to anthracycline cardiotoxicity. Patients younger than 75 years and without heart failure or pulmonary disease are more likely to receive chemotherapy. The main manifestations of acute cardiotoxicity are cardiac rhythm disturbances and the pericarditis/myocarditis syndrome[2,3], while early (several days to months following therapy) and late (years to decades after treatment) cardiotoxicity is mainly characterized by deterioration of myocardial function. Subclinical cardiomyopathy is quite more prevalent than symptomatic heart failure. Various predisposing factors have been proposed, such as total dose of anthracyclines <550 mg/m2, high rate of administration, previous chest irradiation, young or advanced age, female sex, and coexistent heart disease and/or arterial hypertension[6–9]. The early detection of cardiotoxicity may lead to the modification of chemotherapeutic regimen[10,11] and to the timely administration of medications for the treatment of cardiomyopathy, such as beta-blockers and ACE inhibitors[12,13]. Echocardiography during low-dose dobutamine infusion (10 μg/kg/min) has the potential to reveal abnormalities of myocardial contractile reserve, while Doppler echocardiography of the mitral valve inflow during diastole has been used for the assessment of left ventricular (LV) diastolic function. This study examines whether the combination of repetitive dobutamine stress echocardiography (DSE) with evaluation of Doppler mitral inflow pattern can be used to predict the development of anthracycline cardiomyopathy.
Patient Population, Study Protocol
We studied 31 patients (71% male) with a mean±SD age of 57±13 years. All patients had a normal baseline ejection fraction (EF), and none of them had received anthracyclines in the past. Their baseline hematological disorders and clinical characteristics are presented in Tables 1 and 2, respectively. The doses of doxorubicin ranged from 150 to 400 mg/m2 of body surface area (mean dose, 323±88 mg/m2). Six patients received idarubicin (mean dose, 36±0 mg/m2) and two patients received also mitoxantrone, a cardiotoxic compound that belongs to anthracenediones (mean dose, 40±14 mg/m2). In addition, five patients underwent mediastinal irradiation following chemotherapy. Systolic and diastolic parameters of LV function were evaluated repetitively: before chemotherapy, at the midchemotherapy period, at the end of chemotherapy, and after 6 months. The local medical ethics committee approved the study protocol. Patients provided informed consent to take part in the study.
Clinical characteristics of the two groups of patients.
All patients (n=31)
Group A (n=17)
Group B (n=14)
Data are presented as number (%) of patients or mean±SD. CAD, coronary artery disease.
↵a Patients receiving oral antidiabetics or insulin.
↵b Defined as blood pressure 140/90 mmHg, or treatment with antihypertensive medication.
↵c Defined as one or more of the following: history of myocardial infarction, critical stenosis in coronary angiography, or typical angina.
A Hewlett Packard Sonos-5500 imaging system (Andover, MA) equipped with a 1.8 MHz transducer using second harmonic imaging to digitize endocardial border visualization was used to record two-dimensional (2D) echocardiograms. Four standard views were recorded, and two experienced reviewers blinded to the clinical data visually scored the digitized echocardiograms. Regional wall motion and systolic wall thickening were scored using a 16-segment model and a five-point grading scale: 1, normal; 2, mildly hypokinetic; 3, severely hypokinetic; 4, akinetic; 5, dyskinetic. For each patient, a wall motion score index (WMSI, total score divided by the number of segments scored) was calculated.
Dobutamine Stress Echocardiography
DSE was performed as previously described. The baseline, low dose, peak stress, and recovery images were displayed as a cineloop format. In addition, we assessed the change in myocardial contractile reserve by subtracting WMSI at low-dose dobutamine infusion from WMSI at rest (delta WMSI). Reduction of wall thickening and new wall motion abnormalities during the stress test were considered to be hallmarks of ischemia, with the exception of the transition of akinesia to dyskinesia, which was considered a mechanically induced phenomenon. The inter- and intra-observer concordances of resting WMSI were 94 and 97%, respectively, while the inter- and intra-observer agreements for the response of WMSI during dobutamine infusion were 92 and 94%, respectively.
Pulsed-Wave Doppler of Mitral Valve Inflow
Pulsed-wave Doppler echocardiography was used to evaluate diastolic LV function. Doppler studies were recorded from the apical four-chamber view, with the Doppler sampler positioned within the inflow portion of the left ventricle, midway between the annular margins of the mitral valve. Mitral velocity profiles were digitized from the modal velocity of the Doppler tracings. The peak E (early rapid ventricular filling) and peak A (atrial-assisted filling) wave velocities were computed to calculate the E/A velocity ratio. The isovolumic relaxation period (IRP) was assessed with the transducer angulated into the apical five-chamber view and the sample volume placed within the LV outflow truck, but in proximity to the anterior mitral valve leaflet, to record both inflow and outflow signals. Measurements of the deceleration time (DT) were made by the computer software, by placing the caliper to the peak of the E wave, and by following the deceleration slope of the E wave down to the intersection with the A wave. The mean value of five consecutive normal beats was calculated. Pulsed-wave Doppler signals were measured at rest and at low-dose (10 μg/kg/min) dobutamine infusion.
Measurements of LV EF were performed by radionuclide ventriculography (RNV), which is the established non-invasive standard for monitoring anthracycline cardiomyopathy. We used this method as an independent reference for the assessment of EF before the performance of every echocardiographic examination. A small field-of-view gamma camera system (Orbiter, Siemens, Erlangen, Germany), oriented in a 45° left anterior oblique position with a 5° to 10° caudal tilt, was used. After injection of Tc-99m (740 MBq), RNV was performed at rest with the patient in the supine position. The LV EF was calculated by standard methods (Odyssey VP, Picker, Cleveland, OH).
Data are presented as mean ± standard deviation. Logistic regression including all clinical, radionuclide, and echocardiographic parameters was used to investigate the existence of independent predictors for cardiotoxicity. Analysis of variance (ANOVA) for repeated measurements was used to compare means of parameters of systolic and diastolic function over time. Continuous variables between groups were compared with the Student's unpaired t-test. A P value ≤0.05 was considered statistically significant.
Six months after the end of treatment, 17 patients (group A) had a worsening of the EF ≥5% (P<0.001). Fourteen patients (group B) preserved their initial EF at the end of the study. No significant differences existed between the two groups in the doses of chemotherapeutic agents and previous irradiation received. During the study, there was a gradual worsening of WMSI at rest in group A that paralleled the decline of radionuclide EF (Fig. 1). In the same group, two patients (6%) manifested signs and symptoms of congestive heart failure 2 months and 1 year after the first dose of anthracyclines, respectively. No ischemia was detected in both groups during the serial testing. Early/late diastolic velocity of mitral inflow (E/A ratio) at rest was lower in group A (0.91±0.2 vs 1.27±0.3, P<0.001). Logistic regression analysis of baseline patient characteristics (age, hypertension, diabetes, coronary artery disease) and baseline parameters of systolic function (radionuclide EF, WMSI) and diastolic function (E/A ratio, DT, IRP) revealed that E/A ratio was an independent predictor of cardiotoxicity.
Changes in radionuclide EF (%) and WMSI in groups A and B during follow-up.
At midchemotherapy testing, delta WMSI did not indicate any decrease in myocardial contractile reserve in group A, since there was an improvement of segmental contractility from rest to low-dose dobutamine infusion in this group. Midchemotherapy E/A ratio at rest and during low-dose dobutamine infusion was also significantly lower in group A compared with group B (P<0.001 and P=0.017, respectively), but it did not change afterwards. DT and IRP increased significantly at follow-up. Table 3 summarizes the changes that occurred in systolic and diastolic parameters over time in groups A and B.
Changes in parameters of systolic and diastolic function in groups A and B.
WMSI low dose
E/A ratio rest
E/A ratio low dose
DT rest (ms)
DT low dose (ms)
IRP rest (ms)
IRP low dose (ms)
Data are expressed as mean±SD. *P value was calculated by ANOVA for repeated measurements. E/A ratio, early diastolic filling velocity/atrial-assisted diastolic velocity ratio; DT, deceleration time; IRP, isovolumic relaxation period; WMSI, wall motion score index.
Myocardial contractile reserve assessed by low-dose DSE does not predict the systolic LV function after chemotherapy, since patients with a reduction in EF at rest preserve their ability to improve LV contractility during low-dose dobutamine infusion. Also, an abnormal E/A velocity ratio at rest, measured prior to chemotherapy, independently predicts patients prone to develop anthracycline cardiomyopathy.
Previous Studies of Systolic LV Function
The study of Gottdiener et al. in 1981 showed that 20 of the 32 patients demonstrated asymptomatic LV dysfunction 1–9 months after treatment with anthracyclines. This is consistent with the present findings, since 17 of the 31 patients in our study had a decrease in LV EF, while only two of them developed symptomatic heart failure during follow-up.
Gottsauner-Wolf et al. compared RNV and 2D echocardiography for the assessment of EF in 339 consecutive patients and found that echocardiography was of limited value compared to RNV, especially in patients with slight or moderate reduction in EF. The current findings suggest that the 16-segment evaluation of LV function at rest by WMSI may be superior to global 2D echocardiographic measurement of EF. Autopsy studies have shown that the cardiac injury caused by anthracyclines is patchy, and at times is limited to one or more walls of a ventricle. Therefore, segmental abnormalities can be detected before any global systolic LV dysfunction is apparent.
In 1996, De Wolf et al. performed DSE in 23 childhood cancer survivors who had undergone chemotherapy more than 2 years previously. They found that 85% of these patients showed an abnormal response, both systolic and diastolic, to dobutamine infusion. In contrast, in the present study, all patient with deterioration of EF 6 months after the end of study showed a substantial improvement of systolic LV function at low-dose dobutamine infusion. The relatively short follow-up may explain our findings. It is known that patients with mild and subclinical deterioration of systolic LV function can compensate for a decreased cardiac output with a number of adaptive mechanisms, i.e. increasing preload, heart rate, and contractility during stress. After this period of compensation, LV starts thinning and working against high systolic stress. The latter is a finding of late cardiotoxicity, usually observed several years after the completion of chemotherapy. Our results are also in agreement with that of Bae et al. who, in 1988, studied 18 patients under therapy with anthracyclines. These patients underwent RNV at rest and during exercise. The study concluded that regional wall motion abnormalities were more readily detectable at rest, while exercise did not increase the diagnostic accuracy for the detection of cardiotoxicity.
Previous Studies of Diastolic LV Function
In 1998, Cittadini et al. studied 21 patients before and after treatment with anthracyclines and reported impaired diastolic LV filling compared with the control group. Furthermore, Marchandise et al. in 1989, included 45 patients under chemotherapy and showed a prolongation of the IRP (32%) and a reduction of the E/A ratio (23%). These changes preceded systolic LV dysfunction.
To the best of our knowledge, this is the first study to report that a parameter used for the assessment of diastolic function, the E/A velocity ratio, might have predictive value prior to chemotherapy. A baseline E/A velocity ratio less than unity indicates increased LV stiffness and abnormal relaxation, resulting in impaired LV filling. Patients with this baseline diastolic abnormality seem to be more sensitive to the deleterious effects of anthracyclines on myocytes by a mechanism that still remains unclear. An issue that has to be addressed in the future is the optimal length of follow-up, since it is unclear whether these patients will also show a higher incidence of late cardiotoxicity at longer term follow-up.
This study has several limitations. First, the number of patients studied was relatively small. Second, RNV and DSE were not always performed on the same day. Although these tests were performed in the same week, minor alterations in the hemodynamic status of these patients might have occurred. Finally, quantitative information on wall motion was not available in this study.
At early stages of anthracycline administration, assessment of myocardial contractile reserve during low-dose DSE does not give additional information for the early detection of anthracycline cardiomyopathy. Patients who are candidates for treatment with anthracycline agents and found to have an abnormal baseline E/A velocity ratio require an even more careful monitoring during chemotherapy and for at least 6 months after the last dose of anthracyclines. Repetitive measurements of radionuclide EF and echocardiographic evaluation of systolic and diastolic function at rest seem to be equally effective in detecting cardiotoxicity. If any further worsening of LV function occurs, then modification of chemotherapy regimen and/or the use of cardioprotective agents (i.e. dexrazoxane) and agents indicated for the treatment of heart failure, can be life-saving.
. Stress echocardiography: recommendations for performance and interpretation of stress echocardiography. Stress Echocardiography Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 1998;11:97-104.
Bae JH, Schwaiger M, Mandelkern M, Lin A, Schelbert HR. Doxorubicin cardiotoxicity: response of left ventricular ejection fraction to exercise and incidence of regional wall motion abnormalities. Int J Card Imaging 1988–89; 3: 193–201.