Main

Late cardiac toxicity of anthracyclines has been of increasing concern to pediatric oncologists for the last decades(1–11). Whereas cumulative doses lower than 550 mg/m2 have long been considered safe in adult patients, it now seems that young children may be more prone to develop anthracycline-related cardiomyopathy and may have subclinical or clinically overt impairment of the cardiac function after doses lower than 550 mg/m2. This impairment may be discovered up to several years post treatment(12–15). Standard cardiac function tests (i.e. echocardiography or isotopic FS or EF) do not predict the future development of clinical overt cardiac failure. More sophisticated echocardiographic studies at rest reveal an important percentage of patients with cardiac impairment. However, the cardiac dysfunction revealed by these tests does not correlate with symptoms(16–19).

We performed stress echocardiography with a supine bicycle in asymptomatic patients with normal rest FS and found it a sensitive and clinically relevant, but cumbersome, tool in the evaluation of cardiac dysfunction(20). In adults dobutamine stress echocardiography has become a valid alternative to exercise echocardiography for the assessment of regional wall motion disturbances(21). In pediatric patients one interesting report discussed the echocardiographic evaluation of functional variables other than wall motion disturbances at different rates of dobutamine infusion(16). Significant differences were found for systolic function variables between asymptomatic doxorubicin-treated long-term survivors of childhood cancer and normal subjects. The data of this preliminary study together with our findings during exercise(20), convinced us of the feasibility and yield of dobutamine stress echocardiography in children and survivors of childhood cancer.

Dobutamine stress echocardiography should allow the assessment of cardiac“reserve” and should make earlier and more accurate detection of cardiac dysfunction possible than by echocardiographic studies at rest and during exercise. The purpose of this study was to evaluate dobutamine stress echocardiography in patients who had received moderate dose anthracycline chemotherapy years before.

POPULATION

Twenty-three patients (14 male, 9 female; 7-25 y) who completed chemotherapy with moderate cumulative doses of anthracyclines (180-380 mg/m2) more than 2 y previously (3-14 y) underwent a dobutamine stress echocardiography. Table 1 displays the patients' data. Only three patients were symptomatic (i.e. dyspnea climbing stairs). All of the patients were compared with a control group of 26 healthy young people (15 male, 11 female; 6-26 y). All of the subjects and their parents were previously informed about the test. There was no significant difference for age or weight between the two groups (13.5 ± 5 y versus 15.5 ± 6 y and 48.1 ± 19 kg versus 47 ± 16 kg), and the daily physical activity of the two groups was comparable.

Table 1 Clinical data*
Full size table

METHODS

To obtain echocardiographic images of high quality, the subjects were examined in the left decubitus position for the apical and parasternal windows and in supine position with hyperextension of the neck for the suprasternal window. The echocardiographic equipment consisted of a Toshiba SSH-140A(Tokyo, Japan) with a 5, 3.75, and 2.5 MHz transducer with pulsed wave and color Doppler. The measurements were recorded in accordance to standard recommendations(22). Dobutamine (Dobutrex, Eli Lilly) was administered by continuous infusion in a large vein of the arm.

The administration was subdivided in three periods: a first period of 15 min with an infusion rate of 0.5 μg/kg/min, a second of 10 min at 2.5μg/kg/min, and a third period of 10 min at 5 μg/kg/min. ECG (peripheral leads) and BP (Dinamap automated vital signs monitor-Criticon) were continuously monitored during the entire procedure.

Echocardiographic measurements were performed at rest and after 10 min in every period. The infusion was stopped after the last measurements. Serious arrhythmias, increase of BP of more than 50%, decrease of BP of more than 20%, increase of heart rate of more than 50%, or significant subjective discomfort of the patient were reasons for stopping the test. All the echocardiographic studies and measurements as well as calculations were performed by the same investigator, who was blinded to the patients' clinical data.

ECHOCARDIOGRAPHIC MEASUREMENTS

Monitoring of systolic function included M-mode fractional shortening(normal FS, mean ± 2 SD: 34 ± 7%) and ejection fraction (normal EF, 63 ± 9%), calculated according to standard formulas(22). Doppler values of aortic flow were measured as a supplementary estimation of systolic function, i.e. the maximal velocity of the aortic Doppler signal (Aomax, mean ± 2 SD: 1.27± 0.32 m/s), the mean velocity of the aortic Doppler signal (Aoi, 0.83± 0.6 m/s), the acceleration time of the aortic Doppler signal (AT), the LV ejection time (ET), the ratio AT/ET (0.32 ± 0.1). AT was measured between the onset of the aortic Doppler signal and maximal velocity. Aomax and Aoi estimate systolic function and AT/ET is an estimation of systemic vascular resistance.

Doppler measurements of mitral flow evaluated diastolic function,i.e. the maximal velocity of the E wave of the mitral inflow Doppler signal (Emax, 1 ± 0.36 m/s), the mean velocity of the E wave(Ei, 0.64 ± 0.6 m/s), the maximal velocity of the A wave (Amax, 0.48 ± 0.26 m/s), the mean velocity of the A wave (Ai, 0.28 ± 0.34 m/s), the ratio E/A (2.22 ± 1.25), the ratio Ei/Ai (2.69 ± 2), the deceleration time of the E wave corrected for heart rate (Dec/214 R-R, 0.15 ± 0.06 s). Emax and Ei measure the early mitral flow and estimate ventricular relaxation. Dec/214 R-R depends on relaxation and compliance. Amax and Ai measure the mitral flow during atrial contraction and estimate compliance. E/A and Ei/Ai are dependent on relaxation and compliance. Isovolumetric relaxation time, another parameter of ventricular relaxation, was not measured because of lack of phonocardiographic equipment and simultaneous recording of aortic and mitral Doppler was not attempted.

End systolic meridional stress (stress) was measured according to a modified formula: stress = (0.334 × BPs × LVIDs)/(LVPws × (1+ LVPws/LVIDs)). End systolic wall stress (stress, 64 ± 22 g/cm2) is inversely related to LV wall thickness (LVPws, 1.2 ± 0.38 cm) and directly related to LV size (LVIDs, 2.9 ± 0.76 cm) and intracavitary pressure at end systole and estimates afterload. End systolic wall stress is theoretically related to end systolic BP, which is indirectly measured by means of carotid pulse or axillary pulse tracing(23, 24). We compared end systolic BP measured by means of carotid pulse, and peak systolic pressure measured by manometer(Dinamap vital signs monitor 1846-Criticon), and found an excellent correlation (r = 0.98, Fig. 1). Although physiologically not entirely correct, but mathematically correct, we replaced the end systolic BP in the end systolic wall stress formula with the peak systolic BP. This simplified method makes measurement of “end systolic” wall stress less cumbersome and more accessible to the average echolab.

Figure 1
figure 1

Relation between peak and end systolic pressure.

Full size image

The mean velocity of circumferential fiber shortening was corrected for heart rate: VcFc =(LVIDd - LVIDs)/(LVIDd × LVET/214 R-R). VcFc (1.12± 0.28 c/s) is a preload-independent measure of cardiac contractility. The relationship between VcFc and wall stress, or stress-velocity index is a load-independent estimation of contractility (Figs. 2 and3). A difference of more than 2 SD compared with the mean value of the control group was considered abnormal.

Figure 2
figure 2

Stress-velocity index at rest for patients post chemotherapy (anthra) compared with normal controls(control). s, symptomatic; S, 2 SD above the mean for stress of the control population; L, lower 95% confidence limit for the stress-velocity index. All figures below L show decreased contractility; all figures to the right of S show elevated wall stress.

Full size image
Figure 3
figure 3

Stress-velocity index at a rate of 5 μg/kg/min dobutamine for patients post chemotherapy (anthra) compared with normal controls (control). s, symptomatic; S, 2 SD above the mean for stress of the control population; L, lower 95% confidence limit for the stress-velocity index. All figures below L show decreased contractility; all figures to the right of S show elevated wall stress.

Full size image

STATISTICAL ANALYSIS

The results of the groups were compared with a nonparametric test(Mann-Whitney). The results of the changes under increasing dobutamine rates were evaluated with repeated analysis of variance. The potential effects of age at completion of therapy, length of follow-up, cumulative dose, gender, and mediastinal irradiation on the incidence of abnormal responses to dobutamine were evaluated with the use of logistic regression models.p < 0,05 was considered significant. LV contractility was considered abnormal for each patient individually if the relation between stress and VcFc was below the 95% confidence limit (seeFigs. 2 and3). Similarly for each subject, increased afterload was defined as a value for stress more than 2 SD above the mean in the normal population. Differences for stress or FS according to age did not influence the results.

RESULTS

Table 2 displays the results of the measured variables. No serious adverse effects were recorded. One patient complained of headache after termination of the test. One patient showed a transient second degree atrioventricular block and another patient showed disappearance of a left bundle branch block. No patients had to be excluded from the analysis.

Table 2 Echocardiographic data*
Full size table

At rest significant differences were found for FS, EF, LVPwTh, VcFc, Dec/214 R-R. Fifty percent of the subjects had at least one abnormal value for the variables tested (mean ± 2 SD). Echocardiography at dobutamine infusion rates of 2.5 and 5 μg/kg/min showed significant differences for FS, EF, LVIDs, LVPws, LVPwTh, stress, VcFc, velocity stress index, Aomax, Ai, Emax, E/A, Ei/Ai, Dec/214 R-R. Eighty-five per cent of the subjects showed at least one abnormal value for all the variables tested.

There was no significant difference in BP and heart rate between the two groups, and heart rate did not change with dobutamine. Analysis of variance revealed a significant contribution of dobutamine for E/A, Ai, Ei/Ai, LVIDs, LVPws, LVPwTh, FS, stress, stress-velocity index, Aomax, and Emax.

An abnormal percentage of LV posterior wall thickening at 5 μg of dobutamine correlated with an interval of 10 y or more since completion of anthracycline chemotherapy (or more than 10 y of follow-up). Younger age at the completion of chemotherapy was correlated with a decreased LV posterior wall thickness in systole and diastole and increased systolic wall stress. There was no linear correlation between cumulative dose or gender and any abnormal variable of the dobutamine stress echocardiography. Two of the 3 patients with mediastinal irradiation showed an abnormal response to dobutamine. The age-dependent variation (i.e. FS and stress) did not influence the results. The results of the dobutamine stress test can be divided and evaluated by functional groups.

Systolic function. FS and EF (an index of the stroke volume) already showed a significant difference at rest, but most of the values were within the normal range. At 5 μg of dobutamine, 43% of the subjects showed an abnormal response and almost all subjects scored below the normal mean. The three symptomatic patients showed evidently the worst results(Fig. 5). In contrast to the systolic diameter, the diastolic internal diameter of the LV did not differ between the two groups, confirming that dilatation of the LV was not a dominant feature in the early pathologic stages of the cardiac dysfunction. The inadequate increase of the maximal aortic velocity under dobutamine was in accordance with the abnormal results of FS and EF.

Figure 5
figure 5

Percent fractional shortening at a rate of 5μg/kg/min dobutamine for patients post chemotherapy (anthra) compared with normal controls (control). The anthracycline group has been sorted according to the cumulative dose of anthracyclines (x axis), whereas the control group has been dispersed over the graphic.s, symptomatic; r, mediastinal irradiation. -2 SD, 2 SD below the mean of the control group.

Full size image

Percentage LV posterior wall thickening: LVPwTh (a measure that reduces dependence on body surface area)(25) was very variable at rest, but 60% of post anthracycline patients had an insufficient thickening of the posterior wall of the LV in response to dobutamine, suggesting an abnormal contractile mass (Fig. 4).

Figure 4
figure 4

Percent LV posterior wall thickening at a rate of 5μg/kg/min dobutamine for patients post chemotherapy (anthra) compared with normal controls (control). The anthracycline group has been sorted according to the cumulative dose of anthracyclines (x axis), whereas the control group has been dispersed over the graphic.s, symptomatic; r, mediastinal irradiation.

Full size image

Systolic wall stress. At rest 26% of the subjects already showed an increased wall stress. During dobutamine stress echocardiography this figure rose to 52%, including the three symptomatic patients(Fig. 6). Wall stress is inversely related to LV wall thickness, and directly related to LV size and intracavitary pressure. In these patients pressures were normal and the increased wall stress was exclusively attributable to reduced wall thickness.

Figure 6
figure 6

Systolic wall stress at a rate of 5 μg/kg/min dobutamine for patients post chemotherapy (anthra) compared with normal controls (control). The anthracycline group has been sorted according to the cumulative dose of anthracyclines (x axis), whereas the control group has been dispersed over the graphic. s, symptomatic; r, mediastinal irradiation.

Full size image

Contractility. As assessed by the relation between afterload and the velocity of circumferential fiber shortening, contractility was not impaired in the study group overall (Figs. 2 and3). The stress-velocity index, a load-independent parameter, in contrast to FS and EF, showed that only a minority of subjects suffered from diminished contractility, but the majority suffered from elevated afterload. The three symptomatic patients displayed a severely elevated afterload in response to dobutamine.

Diastolic function. Doppler values of the mitral flow at rest showed only a prolongation of the deceleration time of the E wave. At a rate of 5 μg of dobutamine, differences were enhanced with a decrease of E/A, Ei/Ai, Emax, a prolongation of the deceleration time and an increase of Ai. All these changes suggest a disturbance of LV relaxation. For most of these values the standard deviations were relatively high compared with the mean values with important individual variations.

Although for several systolic function variables significant differences already exist between the patient and control group at rest, most of the individual patient data are still within normal limits and thus not very useful to define the onset of cardiac dysfunction in individual observations. The separation becomes more distinct at the maximal rate of dobutamine allowing the identification of cardiac dysfunction in individual subjects.

DISCUSSION

Increasing numbers of children are surviving childhood cancer. Recently reports of late congestive heart failure, ventricular tachycardia, and sudden death late after completion of anthracycline chemotherapy focused on late cardiac toxicity of anthracyclines(12–15). Several studies using different methods of evaluation of cardiac function in survivors of childhood cancer, who were treated with anthracyclines, found important percentages of patients with cardiac dysfunction (from 30 to 70%)(16–20). Long-term studies with follow-up for more than 15 y confirmed the progressive nature of the cardiac damage(15). None of these studies could, however, correlate outcome (heart failure) with a single variable of cardiac function at rest. Most of them even failed to find any correlation between abnormal values and symptoms.

We used bicycle stress echocardiography to elicit subclinical cardiac dysfunction in asymptomatic patients with normal rest values (FS and Aomax). The results indicate that stress echocardiography is more sensitive for assessing cardiac toxicity than echocardiography at rest(20). By evaluating “cardiac reserve,” stress echocardiography could possibly predict which patients will develop overt cardiac failure over time. However, bicycle stress echocardiography is technically difficult, and the amount of variables recorded simultaneously remains very restricted. Echocardiography during pharmacologic-induced stress(i.e. dobutamine) allows more precise and complete evaluation of cardiac function.

Dobutamine is a synthetic catecholamine with α1, β1, and β2 mimetic activity. In the pediatric population dobutamine has different hemodynamic effects on different thresholds of plasma concentrations (related to infusion rates). Stroke volume (and FS and EF) increases at 1 μg/kg/min, systolic BP increases at 2.5 μg/kg/min (as a result of increasing stroke volume, not of vascular resistance), and heart rate increases at 5.5μg/kg/min. Plasma concentrations are at steady state after 10 min, accounting for the 10-min intervals in the test as described. At 5μg/kg/min the threshold for increasing stroke volume was reached for every subject without increasing heart rate(26–28). As a consequence one measures a simple inotropic effect at 5 μg without adding a chronotropic stimulus or an important peripheral vascular effect. It makes interpretation of the results easier than in the standard exercise test where multiple factors other than inotropism act at the same time. Another interesting mechanism is the stimulation of cardiac β1 receptors. Down-regulation of these receptors plays an important role in the pathophysiology of anthracycline cardiomyopathy and increases the threshold for stimulation by catecholamines. Important side effects are associated with infusion rates of 10 μg/kg/min and higher, making those high rates unwarranted for this purpose. Nevertheless the test remains extremely safe, and doses in post infarction patients are considerably higher(21, 29).

Dobutamine stress echocardiography elicited an abnormal response in 85% of the patients studied. Both systolic and diastolic functions were affected. In contrast to the values at rest, almost all patients had values below the mean for FS and EF, and 43% had values lower than 2 SD below the mean. The three symptomatic patients displayed the worst results, suggesting a good relation between FS during dobutamine stress echocardiography and clinical symptoms.

In the majority of the patients (52%), the systolic dysfunction was not a result of decreased contractility (Fig. 3), but of increased systolic wall stress (afterload). The excess afterload was almost exclusively attributable to reduced LV posterior wall thickness. This suggests an inappropriate LV muscular mass. The elevated systolic wall stress probably contributes to the systolic dysfunction seen in these patients.

Abnormal relaxation was the predominant diastolic dysfunction. As usual individual data varied considerably due to the impossibility of controlling and measuring all variables that affect mitral inflow in a clinical model(30), questioning the clinical relevancy of these results in the individual subject. Symptomatic patients showed the worst results for FS, EF, and wall stress in contrast to their data at rest. For all the variables dobutamine magnified the differences between patients and controls facilitating the detection of cardiac dysfunction as also demonstrated in a previous study(16).

The abnormal percentage of LV posterior wall thickening in response to dobutamine was related to the length of follow-up suggesting inadequate cardiac growth with time or a progressive cardiac muscle damage. The correlation between younger age at the completion of chemotherapy and decreased wall thickness and increased stress supports this hypothesis, because the inhibition of growth will be accentuated in younger children, whose LV mass is smaller(31). There was no relation to cumulative dose (probably as a consequence of the selection of patients who did not receive high cumulative doses of anthracyclines) or to gender. Two out of three patients who underwent mediastinal irradiation had an abnormal response to dobutamine, but numbers were too small for any conclusions.

These results confirm that dobutamine stress echocardiography detects systolic and diastolic cardiac dysfunction in patients after moderate cumulative doses of anthracycline chemotherapy. Much to our surprise, impaired contractility is not a major contributing factor to this cardiac dysfunction, in contrast to the increase in systolic wall stress and the impairment of cardiac relaxation. This confirms the results of other studies. The increase of systolic wall stress is a consequence of inadequate LV posterior wall thickening during systole, suggesting an inadequate ventricular mass. This is confirmed by autopsy and myocardial biopsy data in patients who presented with cardiac failure or sudden death as a consequence of late anthracycline cardiotoxicity. Fibrosis and hypertrophy of the remaining myocites without signs of acute toxicity are prominent(15, 18). The progressive nature of the cardiac dysfunction is demonstrated by the relationship between length of follow-up and percentage LV posterior wall thickening, as well as data of other studies(15, 18).

The hypothesis that the combination of inappropriate cardiac muscle mass(with progressive damage?) and increasing wall stress eventually lead to overt heart failure makes dobutamine stress echocardiography attractive. By evaluating the reaction on a positive inotropic stimulus one could get an idea of the functional “cardiac reserve.” Our findings that symptomatic patients have the worst response for FS and wall stress add value to the hypothesis that dobutamine stress echocardiography could eventually predict which patients will further deteriorate with time. Although dobutamine stress echocardiography is already more sensitive than echocardiography at rest, the correlation with length of follow-up suggests that sequential studies using each patient as his own control could enhance the predictive value of the dobutamine stress echocardiography by discovering a progressive decline of cardiac function. This is important because the evolution can be very fast once clinically overt cardiac failure has developed(12, 13, 15).

Another important result of this study is the 85% incidence of“cardiac dysfunction” with dobutamine. This questions the idea of a safe dose of anthracycline. As a consequence all patients who completed chemotherapy with anthracycline should have lifelong cardiac follow-up. This follow-up could consist of FS and echocardiographic wall stress measurement at rest each year. A dobutamine stress echocardiography could be performed at the end of chemotherapy and at 5 y intervals. In case of abnormal results at rest or during dobutamine stress echocardiography or in case of a decline of function variables over time one can increase the frequency of the sequential dobutamine stress echocardiographic studies.

Regardless of the future evolution of the cardiac function in our patients and of the possible predictive value of stress echocardiography, dobutamine stress echocardiography may be of utility by assessing the intensity of physical effort a patient may cope with, especially if he or she wants to engage in sport activities. The question can also be raised whether the patients who show important anomalies in response to dobutamine could benefit from a prophylactic treatment with angiotensin converting enzyme-inhibitor drugs. Although these results are impressive, one should not consider lowering anthracycline doses during chemotherapy without a suitable alternative preserving the patient's chances of survival.

CONCLUSION

Dobutamine stress echocardiography is a very sensitive method to detect subclinical and clinical cardiac dysfunction in patients post anthracycline chemotherapy. The incidence and severity of abnormal responses to dobutamine in patients after moderate cumulative doses of anthracyclines questions the concept of a safe dose. This seems only logical according to the linear relationship between cumulative dose and histopathologic damage, which is already present at low doses.

The correlation between length of follow-up and decreasing LV posterior wall thickness underlines the progressive nature of the late anthracycline cardiotoxicity. The good correlation with symptoms suggests that (sequential) dobutamine stress echocardiography could be of value to predict which patients will ultimately develop overt cardiac failure and need more regular follow-up. Only prospective follow-up will answer this question.