Cardiac late effects are responsible for a significant burden of mortality and morbidity among pediatric Hodgkin’s lymphoma (HL) survivors (HLS). The aim of our study was to assess clinical and subclinical cardiac sequelae in a cohort of childhood HLS treated in the 1980s with doxorubicin, bleomycin, vinblastine, and dacarbazine (the ABVD regimen) and limited-field radiotherapy (RT).
We retrospectively examined a series of HLS treated from 1979 to 1989. We searched for subtle cardiac abnormalities in a subgroup of asymptomatic individuals, who underwent rest and exercise echocardiography at least 20 years after completing their therapies. Their cardiac assessment included physical examination, electrocardiogram (ECG), and resting and postexercise echocardiograms.
On thorough cardiac assessment a mean of 21 years after their diagnosis, none of the 53 unselected asymptomatic HLS showed physical signs or significant ECG abnormalities during or after the stress echo test. Twenty-two (41%) of the 53 patients revealed valvular abnormalities, with mitral regurgitation in 28%, aortic regurgitation in 9%, and both in 4%. No significant myocardial dysfunction as a result of previous combined doxorubicin treatment and chest RT was identified. Only 2 individuals had mild pericardial alterations.
The present study shows that long-term cardiac effects are common in HLS treated with the ABVD regimen and RT. The most frequent complications observed in this sample were essentially coronary artery disease and valvular abnormalities. None of the survivors in this sample showed overt congestive heart failure, a finding in contrast with larger studies.
Tumori 2017; 103(6): 566 - 571
Article Type: ORIGINAL RESEARCH ARTICLE
AuthorsCarlo Materazzo, Maura Massimino, Elisabetta Schiavello, Marta Podda, Lorenza Gandola, Graziella Cefalo, Serena Catania, Cristina Meazza, Ivan Moschetti, Monica Terenziani
- • Accepted on 12/06/2017
- • Available online on 08/07/2017
- • Published in print on 23/11/2017
This article is available as full text PDF.
Cardiac late effects are responsible for a significant burden of mortality and morbidity among childhood cancer survivors (CCS). Various studies report a 6- to 8-fold higher mortality attributable to cardiovascular (CV) disease in CCS than in the general population (1-2-3). At the same time, CCS are 5 to 15 times more likely than their siblings to experience congestive heart failure (CHF), coronary artery disease, clinically significant pericardial disease, or valvular abnormalities. The cumulative incidence of these events continues to rise over time after their diagnosis (4, 5). Subclinical cardiac abnormalities are even more common, with a frequency that varies widely depending on the definition of the outcome, the group under study, and the length of follow-up (6).
Survivors of pediatric Hodgkin’s lymphoma (HLS) form the group at highest risk of CV complications (5). Chemotherapy and radiotherapy (RT) are both known to be cardiotoxic, and cumulative anthracycline doses, female sex, concomitant therapies, and the dose of radiation delivered to the heart are the main risk factors involved (7-8-9). In recent years, considerable efforts have been made to reduce the incidence of cardiotoxicity by containing the cumulative doses of both anthracycline and RT. Many currently adopted therapeutic schemes consequently contain cumulative doses of anthracyclines of no more than 250-300 mg/m2, and new RT techniques are available that minimize cardiac involvement.
Studies on CCS have focused on anthracycline-induced myocardial dysfunction as a leading long-term cardiac complication of anticancer treatment (10, 11). With longer follow-up, there is some evidence to indicate that valvular and coronary lesions secondary to RT are increasingly being recognized (12).
The aim of our study was to assess the whole spectrum of clinical and subclinical adverse cardiac effects in a HLS cohort treated in the 1980s with doxorubicin (A), bleomycin (B), vinblastine (V), and dacarbazine (D) (the ABVD regimen) plus limited-field RT at the Fondazione IRCCS Istituto Nazionale dei Tumori of Milan.
Our eligibility criteria included age under 18 years at the time of diagnosis with HL stages I-III according to the Ann Arbor staging system, treatment from October 1979 to February 1989, and survival for at least 5 years. The prevalence of clinical outcomes such as CHF, coronary and arrhythmic events, or pericardial and valvular abnormalities was retrospectively assessed after completing a follow-up of at least 20 years, and a subgroup of asymptomatic survivors underwent rest and exercise echocardiography. All patients were followed up for at least 20 years after completing their HL therapy.
Informed consent was obtained from individuals attending the late effects follow-up clinic. The protocol was approved by the institutional review board.
All the participants enrolled had been treated according to a combined chemo-RT program. They had received 3 cycles of the ABVD regimen followed by limited-field RT (involved sites plus contiguous areas): the radiation treatment had started approximately 1 month after the third cycle of ABVD. Children with HL stage IB, IIB, or III had received 3 additional cycles of ABVD starting 4 weeks after completing their RT. The radiation doses to the involved sites were 35 and 30 Gy, respectively, when partial or complete remissions had been achieved after 3 cycles of ABVD. The radiation dose to contiguous sites was 25 Gy in all cases. The RT had been delivered with a cobalt unit or high-energy linear accelerator at a daily dose of 1.8 Gy, in 5 fractions per week.
All patients were routinely followed up at our pediatric clinic every 3 months for the first 2 years after therapy, then every 6 months for another 2 years, and annually thereafter. A specialist cardiac assessment was conducted at the discretion of the pediatric oncologist.
For patients living far away from our center, and those who could not undergo or failed any routine tests, detailed clinical information was collected by contacting them directly by phone or through their primary care physician.
Major cardiac events were judged as ≥3rd grade according to the Common Terminology Criteria for Adverse Events (CTCAE), version 3.0 (
Late cardiotoxicity assessment
The cardiac assessment included physical examination, 12-lead electrocardiogram (ECG), and resting and postexercise echocardiograms. The ultrasound scans were obtained using commercially available equipment (Philips Sonos 5500), with a multifrequency transducer, in accordance with the ASE recommendations for measuring left ventricle parameters by 2D echocardiography (13). Echocardiography was performed by 2 clinicians with experience of more than 10,000 ultrasound tests.
The following left ventricular (LV) parameters were measured at rest: dimensions at end of diastole and at end of systole, IV septum, and posterior wall diastolic thickness, shortening fraction (SF), and ejection fraction (EF). The LVEF was measured, using the biplane modified Simpson rule (wherever possible) or the monoplane method, as the mean of 3 cardiac cycles. Shortening fraction and/or EF values below 28% and 55%, respectively, were used as cutoffs for global systolic dysfunction.
Left ventricular diastolic function was assessed from the pulsed-wave Doppler mitral flow according to standard criteria, measuring the peak E wave velocity (E), and peak A wave velocity (A), and calculating the E/A ratio (14). These measurements were compared with those derived from the literature for an age-matched healthy population (15).
Valvular abnormalities were graded according to the ESC guidelines (16). Trivial valvular regurgitations were disregarded.
Postexercise ultrasound consisted of repeating the same echocardiographic measurements within 90 seconds of completing a symptom-limited effort on a cycle ergometer according to the modified Bruce protocol. Electrocardiogram and blood pressure were measured at each stage of the protocol. A postexercise decline in EF of more than 5% points was considered abnormal.
The statistical analysis was conducted using SAS version 9.1 software (SAS Institute, Inc.). Continuous variables such as age, EF, SF, and all echocardiographic parameters were described using mean and standard deviation, while relative and absolute frequencies were used for categorical variables.
Groups (150 vs 300 mg doxorubicin doses) were assessed on differences in their echocardiographic parameters using a t-test, while a paired t-test was used to compare the ultrasound results at rest and postexercise. A level of p≤0.05 for a bilateral test was considered statistically significant.
A test for independent samples was used to compare our population’s Doppler-derived diastolic measurements with the normal values for healthy people.
Eighty-three consecutive HLS entered this study (
Demographic characteristics of survivors of pediatric Hodgkin’s lymphoma
|Whole sample, n (%)||Stress echo subsample, n (%)|
|Total||83 (100)||53 (100)|
|Male||58 (70)||41 (77)|
|Female||25 (30)||12 (23)|
Flow diagram of survivors of pediatric Hodgkin’s lymphoma assessed for late cardiac adverse events.
The median follow-up was 25 years (range 21.6-31.2 years); 43% of the HLS had been seen within the last 2 years, and 82% within the last 3 years.
The major CV events occurring during the follow-up are listed in
Characteristics of survivors of pediatric Hodgkin’s lymphoma developing late cardiovascular (CV) complications
|No.||Sex||Type of event||Time since therapy, years||Age at first CV event, years||Cumulative doxorubicin dose, mg/m2a|
|a All these patients received mediastinal irradiation with 25-35 Gy.|
|1||M||Acute myocardial infarction||20||34||300|
|2||M||Acute myocardial infarction||22||32||300|
|3||M||Acute myocardial infarction||23||37||300|
|4||M||Acute myocardial infarction||23||36||300|
|5||M||Stable angina with left ventricular impairment||22||34||150|
|6||M||Third-degree atrioventricular block||24||36||300|
|7||F||Mitral valve replacement||20||35||150|
|8||M||Aortic valve replacement||9||14||300|
|9||M||Aortic and mitral valve replacement||21||36||300|
None of the 53 unselected asymptomatic HLS (64% of the sample) who underwent extensive cardiac assessment a mean 21 years after their HL was diagnosed showed cardiac symptoms or significant ECG abnormalities during or after the stress echo test. All stress echo tests were stopped due to muscle fatigue.
The ultrasound findings at rest and after exercise are listed in
Echocardiographic parameters at rest and after exercise
|No. of patients||Dose (cycles)||Sex, F/M||Age, y (SD)||SF rest (SD)||SF stress (SD)||ΔSF (SD)||EF% rest (SD)||EF% stress (SD)||ΔEF% (SD)|
|EF = ejection fraction; SF = shortening fraction.|
|53||12/41||32 (4.1)||35.5 (3.2)||43.1 (5.2)||7.7 (4.9)||65.7 (4.7)||73.3 (5.9)||7.7 (6.2)|
|22||300 mg/m2 (6)||5/17||33 (3.7)||34.9 (3.4)||43.1 (5.37)||8.2 (4.7)||65.8 (5.5)||74.6 (7.0)||9.2 (7.7)|
|31||150 mg/m2 (3)||7/24||32 (4.4)||35.8 (3.0)||43.5 (4.9)||6.3 (9.0)||65.7 (4.3)||72.5 (5.1)||6.8 (5.1)|
Diastolic function (measured by E-wave and A-wave peak velocities, and E/A ratio) showed no statistically significant difference between our HLS and normal values for age-matched healthy people obtained from the literature (15). The E-wave and A-wave peak velocities did not differ statistically between the 2 subgroups given a different dose of doxorubicin. There was a trend towards a lower E/A ratio in CCS given 300 mg/m2 than in those treated with 150 mg/m2 of doxorubicin (0.98 [SD 0.21] vs 1.19 [SD 0.3], respectively; p = 0.054) (
Diastolic function: E-wave and A-wave peak velocity, E/A ratio
|No. of patients||p|
|Dose, mg/m2 Cycles||150 3||300 6|
|E, cm/s||89.2 (16.2)||89.6 (17.0)||88.7 (16.0)||NS|
|A, cm/s||85.1 (22.8)||78.5 (20.0)||93.9 (23.0)||NS|
|E/A||1.1 (0.27)||1.19 (0.3)||0.98 (0.21)||0.054|
Twenty-two (41%) of the 53 HLS had valvular abnormalities: there were 15 (28%) cases of mitral regurgitation (11 mild and 4 moderate), 5 (9%) of aortic regurgitation (4 mild and 1 moderate), and 2 (4%) of both mitral and aortic regurgitation. Only 2 HLS had mild pericardial alterations.
Survivors of childhood HL form a group with a very high risk of therapy-related CV late effects (5). Both chemotherapeutic agents (especially anthracyclines) and RT are known to harm the heart (17). There is a long list of factors associated with the likelihood of cardiotoxicity, including female sex, black race, latency period, and dosing schedule, while there is some controversy concerning the influence of age at the time of exposure (18-19-20). The most relevant, however, by far are the cumulative doses of anthracycline and cardiac irradiation (18). Patients treated with a cumulative doxorubicin dose of ≥250 mg/m2 are reportedly at higher risk of developing cardiomyopathy than those given lower doses, with an estimated risk at 20 years after starting the therapy of nearly 10% (19). Similarly, a cumulative mediastinal radiation exceeding ≥35 Gy is associated with a significant risk of cardiac toxicity, and an increased relative risk of CV mortality (17, 21, 22). The combination of the 2 treatments (anthracycline and mediastinal radiation) exposes cancer survivors to a 44-fold higher risk of CHF than in patients without such exposures.
One study on HLS reported a 3.3-fold higher risk of dying as a result of cardiac disease after receiving RT than in the general population (1). In another large study on CCS, a radiation dose of 15-34 Gy was associated with a hazard ratio for AMI of 2.4 in comparison with individuals not exposed to radiation (5). In a sample of 1,474 survivors of HL under 41 years old at the time of their treatment, Aleman et al (23) reported an overall 30-year incidence of AMI and valvular disorders of 12.9% and 19.7%, respectively.
In a population of similar age, we found a high prevalence (11%) of coronary artery disease, valvular dysfunction, and conduction abnormalities, which are all complications of heart irradiation. Several previous studies showed that these complications may occur a variable amount of time after receiving RT (generally at least 5-10 years) if the heart is totally or partially included in the radiation fields (23-24-25). The most significant abnormalities revealed by our ultrasound screening were valvular dysfunctions, especially involving the left heart valves: the prevalence we identified (41%) is similar to that of some previous reports (12, 22), while other studies found a higher frequency (up to 60%) of some types of valve disease, though the mantle irradiation dose involved (mean 40 Gy, and at least 35 Gy) was also higher than the one administered in our population (26). All of our asymptomatic HLS had regurgitant valve lesions, which were of mild degree and without hemodynamic consequences in most cases. Apart from the patients (
What is surprising is that we found no significant myocardial dysfunction as a result of previous combined anthracycline treatment and chest radiotherapy. None of our HLS showed overt heart failure, and only one had left ventricular systolic function parameters below normal limits. These findings are clearly in contrast with historical data and recent cohort studies. In 1991, Steinherz et al (10) reported an 11% incidence of CHF 4-20 years after completing anthracycline therapy with cumulative doses of less than 400 mg/m2. van Dalen et al (19) followed up 830 children treated with a mean cumulative dose of 288 mg/m2 of anthracycline for a mean 8.5 years, finding a 2.5%. prevalence of CHF. The estimated risk of heart failure 20 years after starting anthracycline therapy increased to 5.5%, and to 9.8% for doses of 300 mg/m2 or more. In the analysis of the Childhood Cancer Survivors Study cohort, the 30-year cumulative incidence of CHF was 4.1%, with a hazard ratio of 6.8 for CCS treated for HL vis-à-vis a group of matched siblings (5).
Subclinical cardiotoxicity has been described even more frequently. In a systematic review conducted on 25 studies, each including more than 50 patients treated with anthracycline, the reported frequency of asymptomatic left ventricular abnormalities was as high as 57%, and ranged from 0% to 57% (6). A clear dose-related response was apparent, with abnormal SF in the range of 15.5%-27.8%, and abnormal afterload in the range of 19%-52% for patients treated with anthracycline doses exceeding 300 mg/m2, as opposed to an abnormal left ventricular function in the range of 0%-15.2% for patients who received lower cumulative doses. More recently, after a mean follow-up of 15.4 years in 514 patients surviving childhood cancer for 5 years or more, van der Pal et al (18) reported a SF of less than 30% in 27% of their sample. Analyzing the doxorubicin dose-related response revealed a slight decline in SF for cumulative doses below 100 mg/m2, and a linear decline for higher doses.
Our data are consistent with the findings of Sorensen et al (27), and Rammeloo et al (28), who found no deterioration in cardiac function in patients treated with low-dose anthracycline (daunorubicin <240 mg/m2), although the median follow-up in these studies was no longer than 15 years.
The exercise test failed to reveal any significant subclinical impairment in cardiac function in our sample. None of our HLS had a pathologic response in EF with exercise, and only a minority showed a flat response (the clinical significance of which is unknown, and probably irrelevant). Two studies similar to ours, which used exercise as a cardiac stressor, reported a lower ventricular systolic reserve, or a lower EF postexercise than at rest in a large proportion (about 40%) of asymptomatic children or young adults previously treated with anthracyclines (29, 30).
We found no relevant abnormalities in diastolic function. Compared with previously published reference values for age-matched normal individuals (15), we found a trend towards a lower E/A ratio in patients treated with higher doses of doxorubicin (300 mg/m2), but it did not reach statistical significance. Altered diastolic indexes (that may sometimes precede a decrease in systolic function) have often been identified, using various methods, during and after the completion of anticancer therapy (31-32-33-34).
The main limitations of the present study concern the small sample considered, which was treated and followed up at the same institution, which could lead to a patient selection bias; our inability to submit all HLS to exercise echocardiography; and the lack of a control group of normal subjects. This last aspect could interfere with the interpretation of some results, particularly regarding the response of left ventricular function to exercise. Normal values for left ventricular systolic function at rest are fairly well-validated, and generally accepted worldwide, but defining a set of reference values for exercise meets with some difficulties. We somewhat arbitrarily considered as clearly abnormal a drop in EF after physical exercise >5% compared with the resting EF, but this approach finds little support in the literature, and may have limited our ability to identify subtle cardiotoxic effects correctly. In addition, the more sophisticated echocardiographic measurements that were investigated in other recent studies (35-36-37-38) were not considered here. We chose to focus instead on parameters with a more established prognostic value, i.e., SF and EF (39, 40).
The present study confirmed that long-term cardiac effects are common in HLS treated with the ABVD regimen and RT.
The most frequent complications observed in this study were essentially coronary artery disease and valvular abnormalities, whereas none of our HLS showed overt CHF. This latter finding (in contrast with other, larger studies) could have important implications for surveillance of the long-term adverse effects of anticancer therapy. The current guidelines recommend periodically assessing cardiac status in CCS, according to a schedule that is based on their exposure to anthracycline and the cumulative dose, the presence/absence of cardiotoxic RT, and relevant medical history (41, 42). Judging from our study, it would be advisable to screen for coronary artery disease and valvular dysfunction several years after completing the anticancer therapy, when the risk of such complications begins to increase.
Möller TR Garwicz S Barlow L et al. Association of the Nordic Cancer Registries; Nordic Society for Pediatric Hematology and Oncology: Decreasing late mortality among five-year survivors of cancer in childhood and adolescence: a population study in Nordic countries. 2001 19 13 3173 81
Quiñones MA Otto CM Stoddard M Waggoner A Zoghbi WA Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. Recommendations for quantification of Doppler echocardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. 2002 15 2 167 84
- Materazzo, Carlo [PubMed] [Google Scholar] 1, * Corresponding Author ()
- Massimino, Maura [PubMed] [Google Scholar] 2
- Schiavello, Elisabetta [PubMed] [Google Scholar] 2
- Podda, Marta [PubMed] [Google Scholar] 2
- Gandola, Lorenza [PubMed] [Google Scholar] 3
- Cefalo, Graziella [PubMed] [Google Scholar] 4
- Catania, Serena [PubMed] [Google Scholar] 2
- Meazza, Cristina [PubMed] [Google Scholar] 2
- Moschetti, Ivan [PubMed] [Google Scholar] 5
- Terenziani, Monica [PubMed] [Google Scholar] 2, * Corresponding Author (email@example.com)
Cardiology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan - Italy
Pediatric Oncology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan - Italy
Pediatric Radiotherapy Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan - Italy
Pediatrics Departments, Ospedali Santi Paolo e Carlo, Milan - Italy
Mario Negri Institute for Pharmacological Research IRCCS, Milan - Italy