HL indicates Hodgkin lymphoma.
CHD indicates coronary heart disease; HF, heart failure; and VHD, valvular heart disease.
eAppendix. Grading Criteria Based on CTCAEv4.0
eTable 1. Description and Grading of CVD
eTable 2. Standardized Incidence Ratios for First Cardiac Events
eTable 3. Risk of Any Cardiac Event Following Hodgkin Lymphoma Treatment
eTable 4. Risk of First Cardiac Event Following Hodgkin Lymphoma Treatment by Follow-up Period
eFigure 1. Standardized Incidence Ratios and Absolute Excess Risks for Any CHD and HF by Follow-up Period
eFigure 2. Standardized Incidence Ratios and Absolute Excess Risks of First CHD and HF by Follow-up Period
eFigure 3. Cumulative Incidence of CVD by Treatment Period
van Nimwegen FA, Schaapveld M, Janus CPM, Krol ADG, Petersen EJ, Raemaekers JMM, Kok WEM, Aleman BMP, van Leeuwen FE. Cardiovascular Disease After Hodgkin Lymphoma Treatment40-Year Disease Risk. JAMA Intern Med. 2015;175(6):1007-1017. doi:10.1001/jamainternmed.2015.1180
Hodgkin lymphoma (HL) survivors are at increased risk of cardiovascular diseases. It is unclear, however, how long the increased risk persists and what the risk factors are for various cardiovascular diseases.
To examine relative and absolute excess risk up to 40 years since HL treatment compared with cardiovascular disease incidence in the general population and to study treatment-related risk factors for different cardiovascular diseases.
Design, Setting, and Participants
This retrospective cohort study included 2524 Dutch patients diagnosed as having HL at younger than 51 years (median age, 27.3 years) who had been treated from January 1, 1965, through December 31, 1995, and had survived for 5 years since their diagnosis.
Treatment for HL, including prescribed mediastinal radiotherapy dose and anthracycline dose.
Main Outcomes and Measures
Data were collected from medical records and general practitioners. Cardiovascular events, including coronary heart disease (CHD), valvular heart disease (VHD), and cardiomyopathy and congestive heart failure (HF), were graded according to the Common Terminology Criteria for Adverse Events, version 4.0.
After a median follow-up of 20 years, we identified 1713 cardiovascular events in 797 patients. After 35 years or more, patients still had a 4- to 6-fold increased standardized incidence ratio of CHD or HF compared with the general population, corresponding to 857 excess events per 10 000 person-years. Highest relative risks were seen in patients treated before 25 years of age, but substantial absolute excess risks were also observed for patients treated at older ages. Within the cohort, the 40-year cumulative incidence of cardiovascular diseases was 50% (95% CI, 47%-52%). Fifty-one percent of patients with a cardiovascular disease developed multiple events. For patients treated before 25 years of age, cumulative incidences at 60 years or older were 20%, 31%, and 11% for CHD, VHD, and HF as first events, respectively. Mediastinal radiotherapy increased the risks of CHD (hazard ratio [HR], 2.7; 95% CI, 2.0-3.7), VHD (HR, 6.6; 95% CI, 4.0-10.8), and HF (HR, 2.7; 95% CI, 1.6-4.8), and anthracycline-containing chemotherapy increased the risks of VHD (HR, 1.5; 95% CI, 1.1-2.1) and HF (HR, 3.0; 95% CI, 1.9-4.7) as first events compared with patients not treated with mediastinal radiotherapy or anthracyclines, respectively. Joint effects of mediastinal radiotherapy, anthracyclines, and smoking appeared to be additive.
Conclusions and Relevance
Throughout their lives, HL survivors treated at adolescence or adulthood are at high risk for various cardiovascular diseases. Physicians and patients should be aware of this persistently increased risk.
Hodgkin lymphoma (HL) is the prototype of a curable malignant neoplasm, with 10-year survival rates currently exceeding 80%.1 Consequently, the number of long-term survivors is increasing every year. Treatment for HL has been associated with adverse late effects, such as increased risks of secondary malignant neoplasms and cardiovascular diseases.2- 5 Late cardiovascular complications may arise as a consequence of radiotherapy and chemotherapy and cause substantial excess morbidity and mortality in long-term HL survivors.2,6- 8
Several studies2,9- 12 observed high risks of cardiac mortality in HL survivors after 15 to 25 years of follow-up. Survivors treated before 21 years of age appeared to have the highest relative risk of death from cardiovascular diseases.2,9 We and others also found that HL survivors have an increased risk of cardiac morbidity.7,8,13 Few studies8,13 have compared cardiovascular disease incidence in HL survivors and the general population.
In studies published so far, follow-up rarely exceeded 20 to 25 years. Because most patients with HL are young at diagnosis and cure rates are high, patients in previous reports had often not yet reached the ages at which cardiovascular diseases become more common in the general population. Therefore, studies with prolonged follow-up are needed to assess long-term effects of treatment (eg, when patients with HL enter their sixth or seventh decade of life). Furthermore, the interaction between mediastinal radiotherapy and anthracycline-containing chemotherapy in patients with HL across a broad range of ages at treatment has hardly been studied because follow-up of anthracycline-treated patients was short in most published studies.
This study examined relative and absolute excess cardiovascular disease risk up to 40 years after treatment for HL compared with cardiovascular disease incidence in the general population. In addition, this study examined treatment-related cardiovascular disease risk factors in a cohort of patients with HL treated from January 1, 1965, through December 31, 1995.
Cardiovascular disease risk was studied in a multicenter, hospital-based cohort (N = 2604) of patients treated for HL in 5 Dutch university hospitals or cancer centers. Patient selection has been described in detail previously.2,4,13- 15 Briefly, patients received their first treatment for HL from January 1, 1965, through December 31, 1995, before the age of 51 years and had survived 5 years or more after HL diagnosis.
From the medical records, data on dates of birth and HL diagnosis, histologic findings, stage, and primary and recurrence treatment (radiation fields, prescribed radiotherapy dose, chemotherapy regimens, and number of cycles) were collected. In addition, follow-up information on cardiovascular events and smoking was retrieved by contacting general practitioners and cardiologists in 2004 (complete follow-up for 94%)13,14 and in 2013 (complete follow-up for 83%). Vital status and dates of death were obtained up to July 2013 by linkage with the Dutch Central Bureau of Genealogy. This study was exempt from institutional review board approval according to Dutch law because existing data from medical files were used.
The following cardiovascular events were included: coronary heart disease (CHD), including myocardial infarction and angina pectoris; valvular heart disease (VHD); and heart failure (HF), including congestive heart failure and cardiomyopathy. Fatal incident cardiovascular events are registered by general practitioners in the Netherlands and were also included as events in the analyses. All events were defined and graded according to a slightly adapted version of the Common Terminology Criteria for Adverse Events, version 4.0 (eAppendix in the Supplement).16 The data abstraction forms and coding instructions were developed in collaboration with physicians, and the Common Terminology Criteria for Adverse Events can be used to properly grade cardiovascular events from medical records.17 The any cardiovascular event analyses were used to assess the burden of disease, whereas the first cardiovascular event analysis was performed to study risk factors for the event concerned.
Over time, a wide variety of treatment regimens was used in the study population. Most patients were treated according to European Organisation for Research and Treatment of Cancer Lymphoma Group protocols.18- 22 Frequently used treatments and changes over time have been described in detail previously.13,23 In the 1960s, patients were treated with orthovoltage therapy or cobalt 60; from the 1970s onward, linear accelerators were used. Individual blocks were used to shield normal tissues as much as possible. Patients usually received 40 Gy in fractions of 1.5 to 2.0 Gy when they were treated with radiotherapy only and 30 to 36 Gy when they also received chemotherapy. Mantle field irradiation (including mediastinal, axillary, and neck nodes) was the most commonly applied radiation from the early 1970s to the late 1980s. Since the late 1980s, an increasing number of patients received more limited radiation fields (involved fields). Three-dimensional planning was not used in the current study population. Prescribed radiation dose to the mediastinum was used as a proxy for radiation dose to the heart.
From the 1960s to the 1980s, chemotherapy consisted mainly of mechlorethamine hydrochloride, vincristine sulfate [Oncovin], procarbazine, and prednisone (MOPP). In the 1980s, anthracycline-containing regimens, such as MOPP and doxorubicin hydrochloride [Adriamycin], bleomycin sulfate, and vinblastine sulfate (ABV) and ABV and dacarbazine (ABVD), were introduced as part of primary treatment. Anthracycline dose was estimated in milligrams of anthracycline per square meter of body surface based on number of cycles received times the standard anthracycline dose in the corresponding chemotherapy regimen during that period. Standard doses of anthracycline per regimen per cycle were 25 mg/m2 at days 1 and 15 for ABVD and hybrid MOPP-ABV and 35 mg/m2 at day 8 for alternating MOPP-ABVD.
The incidence of CHD and HF in the cohort was compared with age-, sex-, and calendar period–specific cardiovascular disease incidence rates for the Dutch population, accounting for person-years of observation. No reference incidence data were present for VHD. Cardiovascular disease incidence data from the Continuous Morbidity Registration Nijmegen of general practices were used as a reference for the calendar period of 1971 to 2001.24,25 For 2002 to 2011, we used reference data from the Netherlands Institute for Health Services Research Primary Care Database.26 Standardized incidence ratios (SIRs) were calculated as the ratios of the observed and expected numbers of cardiovascular events in our cohort. Because expected numbers were based on the registration of multiple events per patient, observed numbers were calculated accordingly. Absolute excess risk was calculated as the observed number of events minus the number expected divided by the number of person-years at risk multiplied by 10 000.
Time at risk of cardiovascular diseases began 5 years after the start of first treatment for HL and ended at the date of diagnosis of cardiovascular disease, the date of most recent medical information, the date of migration, or the date of death, whichever came first. Patients diagnosed as having cardiovascular diseases before HL diagnosis (n = 18) or less than 5 years after HL diagnosis (n = 62) were excluded. The 95% CIs of the SIRs and tests for heterogeneity were performed according to standard methods.27 Generalized linear models with Poisson distributions were used to test for heterogeneity and trends.
The cumulative incidence of cardiovascular diseases was estimated in the presence of death from any cause as a competing risk.28 When specific cardiovascular diseases occurring as a first event were analyzed separately, all other cardiovascular events were considered competing risks as well. Multivariable Cox proportional hazards regression analyses were performed to quantify the effects of different treatments and smoking on cardiovascular disease risk within the patient group, adjusting for confounders. Model fit and model assumptions were assessed using graphic and residual-based methods. All Cox proportional hazards regression models were stratified for age at HL diagnosis because hazards were nonproportional by age. Interactions between mediastinal radiotherapy and anthracycline dose and between mediastinal radiotherapy and smoking at HL diagnosis were tested using standard methods. All analyses were performed using STATA statistical software (Stata Corp). P < .05 was considered statistically significant.
Patient characteristics are listed in Table 1. In total, 2524 patients were included in the analyses. Median age at HL diagnosis was 27.3 years. A total of 2052 patients (81.3%) had received mediastinal radiotherapy, and 773 (30.6%) had received anthracycline-containing chemotherapy. After a median follow-up of 20.3 years (range, 5-47 years), we identified 1713 cardiovascular events in 797 patients; 410 patients (51.4%) developed 2 events or more. The most frequently occurring cardiovascular disease was CHD, with 401 patients developing CHD as their first event (eTable 1 in the Supplement), followed by VHD (374 events) and HF (140 events). A total of 1321 cardiovascular events (77.1%) could be graded. Of the VHDs, 429 (67.0%) were grade 2 or higher, and 243 (38.0%) were grade 3 or higher. Of HFs, 140 (77.0%) were grade 3 or higher (eTable 1 in the Supplement). Median intervals between HL treatment and first events were 18, 24, and 19 years for CHD, VHD and HF, respectively.
Survivors of HL had a 3.2-fold increased SIR (95% CI, 3.0-3.5) of developing a CHD and a 6.8-fold increased SIR (95% CI, 5.9-7.6) of developing HF compared with the general population, corresponding to 70 and 58 excess cases of CHD and HF per 10 000 person-years, respectively (Table 2). When only considering first CHD or HF events, SIRs were substantially elevated as well (eTable 2 in the Supplement).
Although SIRs were significantly increased for all age categories, patients treated at a younger age had higher SIRs than patients treated for HL above the age of 35 years. Patients treated before 25 years of age experienced a 4.6- to 7.5-fold increased risk of CHD and a 10.9- to 40.5-fold increased SIR of HF depending on their attained age (P for trend across attained age categories <.001 and <.001, respectively). Patients treated at 35 to 50 years of age experienced a 2.0- to 2.3-fold increased risk of CHD and 3.1- to 5.2-fold increased risk of HF depending on their attained age (P for trend across attained age categories = .17 and .15, respectively) (Table 2). The risks of CHD and HF remained significantly increased beyond 35 years after HL treatment (SIR, 3.9; 95% CI, 2.6-5.6; and SIR, 5.8; 95% CI, 3.8-8.5; respectively), resulting in high absolute excess risks of 475 and 382 per 10 000 person-years, respectively (Table 2 and eFigure 1 in the Supplement). When only considering first CHD and HF events, the SIR for CHD remained significantly increased up to 30 years or more after HL treatment compared with the general population (eTable 2 and eFigure 2 in the Supplement). The SIR for HF remained elevated as well although not statistically significantly so from more than 30 years after HL treatment.
At 40 years after HL, the cumulative incidence of any cardiovascular disease amounted to 49.5% (95% CI, 46.6%-52.4%). Patients treated with mediastinal radiotherapy had a 40-year cumulative incidence of any cardiovascular disease of 54.6% (95% CI, 51.2%-57.9%) compared with 24.7% (95% CI, 17.2%-32.9%) in patients not treated with mediastinal radiotherapy or anthracyclines (Figure 1A). Cumulative incidence curves appeared similar for patients treated in 1965 through 1974, 1975 through 1984, and 1985 through 1995 (eFigure 3 in the Supplement). Forty-year cumulative incidences for CHD and VHD as first events were 22.9% (95% CI, 20.7%-25.1%) and 25.9% (95% CI, 23.4%-28.5%), respectively (Figure 1B). Heart failure most frequently occurred as a subsequent event; the cumulative incidence of HF as a first event was only 8.1% (95% CI, 6.7%-9.7%) after 40 years, whereas the 40-year cumulative incidence for any HF was 24.8% (95% CI, 21.4%-28.4%) 40 years after HL treatment.
Patients treated before 25 years of age had a higher cumulative incidence of cardiovascular diseases at any attained age (Figure 2). Patients treated before 25 years of age reached a given cumulative incidence 10 to 20 years earlier than patients treated at an older age. For example, a 50-year-old survivor treated before 25 years of age experienced the same absolute risk as a 61-year-old survivor treated at 35 to 50 years of age (Figure 2A).
Adjusted for year of HL diagnosis, sex, and ever smoking, mediastinal radiotherapy and anthracycline-containing chemotherapy were associated with increased risk of any cardiovascular disease (hazard ratio [HR] for mediastinal radiotherapy, 3.6; 95% CI, 2.8-4.6; HR for anthracycline chemotherapy, 1.5; 95% CI, 1.2-1.8) (Table 3). Radiation below the diaphragm or vincristine-containing chemotherapy did not influence the risks of any cardiovascular disease. The association between mediastinal radiotherapy and cardiovascular disease risk appeared to be strongest for VHD when evaluating risk for a first event (HR, 6.6; 95% CI, 4.0-10.8) (Table 3) and risk of any VHD event (HR, 5.2; 95% CI, 3.6-7.6) (eTable 3 in the Supplement). Mediastinal radiotherapy was also associated with an increased risk of CHD (HR, 2.7; 95% CI, 2.0-3.7) and HF (HR, 2.7; 95% CI, 1.6-4.8) as a first event. Risk of all cardiovascular diseases combined and of VHD and HF as first events increased with a higher prescribed mediastinal radiation dose (P for trend = .003, .02, and .03, respectively). Anthracycline-containing chemotherapy was associated with increased risks of VHD (HR, 1.5; 95% CI, 1.1-2.1) and HF (HR, 3.0; 95% CI, 1.9-4.7) as first events but not with CHD (Table 3). Risk of VHD increased with higher anthracycline dose (P = .04). No clear dose-response relationship could be identified for HF as a first event; risk of HF was already strongly increased for low anthracycline doses (Table 3).
No interactions on a multiplicative scale were found between mediastinal radiotherapy and anthracycline dose or mediastinal radiotherapy and smoking; the joint effects of these outcomes appeared to be additive rather than multiplicative (Table 3).
When examining whether treatment-related cardiovascular disease risk changed in the course of follow-up, we observed that the risks for any cardiovascular disease and first CHD after mediastinal radiotherapy were statistically significantly higher after 20 to 47 years of follow-up than in the 5 to 20 years of follow-up interval (P for interaction = .002 and .01, respectively) (eTable 4 in the Supplement). Anthracycline-associated risk did not seem to alter with increasing follow-up time (Pforinteraction >.05). However, after 20 years of follow-up, anthracycline-associated risk was still significantly elevated for any cardiovascular disease and VHD and HF as first events (eTable 4 in the Supplement).
To the best of our knowledge, our study is the first to report cardiovascular disease incidence for such a long period after HL treatment, to examine combined effects of age at treatment and attained age, and to evaluate interactions between anthracycline dose and mediastinal radiotherapy in a large cohort treated across a broad age range. We observed that HL survivors experience increased risks of cardiovascular diseases up to at least 40 years after initial HL treatment. Compared with the general population, 4- to 7-fold increased risks of CHD or HF are observed 35 years or more after HL treatment, resulting in 857 excess cardiovascular events per 10 000 person-years. The cumulative incidence of any type of cardiovascular disease amounted to 50% at 40 years after HL diagnosis. Patients who received mediastinal radiotherapy experienced increased risk of not only CHD but also VHD and HF. Anthracycline-containing chemotherapy was associated with increased HF and VHD risk. Half of the patients with a cardiovascular disease developed multiple cardiovascular events. Although patients treated at younger ages had the highest relative risk of cardiovascular disease compared with age-matched peers in the general population, substantial absolute excess risks were also seen for patients treated at 40 years or older.
Only a few studies13,29 observed an association between anthracyclines and VHD. The pathophysiologic mechanism behind this association is still unclear. Because anthracycline-containing chemotherapy has been associated with HF13,30,31 and HF can result in valvular dysfunction by dilation of the ventricles,32,33 asymptomatic HF may underlie this association. Another possibility is that anthracyclines damage the papillary muscles of the valves, causing valvular regurgitation. Allen et al33 suggest that anthracycline-related papillary muscle dysfunction, which is too subtle to be detected by changes in fractional shortening, may result in mitral regurgitation.
We did not find evidence of multiplicative interaction between mediastinal radiotherapy and anthracycline-containing chemotherapy for any cardiovascular disease. Our analysis examining risks for various combinations of anthracycline dose, mediastinal radiotherapy, and smoking suggests additive effects of these risk factors. This finding is in line with the results of Myrehaug et al,11 who found a nonsignificant supra-additive interaction between mediastinal radiotherapy and anthracyclines for cardiac morbidity in a cohort of 615 HL survivors. Studies29,34- 37 in childhood cancer survivors and survivors of other cancers, such as breast cancer, did not address this issue in depth or had insufficient power to study interactions between treatment modalities.
The risk of cardiovascular diseases did not appear to decrease in the more recent decades: high risk of cardiovascular diseases was also found in patients treated in the 1990s. This finding may be attributable to the large number of patients undergoing mediastinal radiotherapy with cardiac exposure, albeit with somewhat more limited radiation fields and doses, and to combined treatment with cardiotoxic chemotherapy. Modeling studies by Maraldo et al,38,39 however, predict reduced cardiac morbidity and mortality with modern radiotherapy techniques introduced since the late 1990s.
A strength of our study is the use of general practitioner– and cardiologist–reported outcomes instead of patient–reported outcomes, reducing potential misclassification bias. The high response rate from the general practitioners renders selection bias unlikely. Furthermore, we were able to grade most cardiovascular events. Most graded events were moderate to severe (grade ≥2, 83%), confirming that this population was not routinely screened for cardiovascular diseases. There is no cardiovascular screening program for HL survivors in the Netherlands. However, a subpopulation (a part of the females treated with mediastinal radiotherapy) was recalled for breast cancer screening (not for cardiovascular screening). Physical examination probably frequently included auscultation of the heart, possibly leading to identification of asymptomatic cardiovascular disease. However, no striking differences were seen in cardiovascular disease incidence between males and females, and no effect modification was found by sex, rendering it unlikely that breast cancer surveillance led to increased incidence of cardiovascular diseases.
Our results have important implications for long-term HL survivors. Screening guidelines for childhood cancer survivors treated with mediastinal radiotherapy and anthracyclines have been developed40- 42 and should be extended to survivors of adult HL. Our results are also important for patients with HL treated to date because anthracyclines continue to be used as standard treatment for HL and various other cancers, such as breast cancer. Furthermore, mediastinal radiotherapy is still indicated for substantial numbers of patients with HL,43,44 although smaller radiation fields and more modern radiotherapy techniques are used. Currently applied doses (20-30 Gy for 10-15 fractions) are lower than in our study but may still be cardiotoxic. Consistent with other studies, we found a dose-response relationship for prescribed mediastinal radiotherapy dose and cardiovascular disease risk. A limitation of this study is, however, that radiation dose-response relationships could not be studied in more depth because radiation dosimetry could not be performed for the entire study population. Future studies of radiation dose-response relationships using doses to specific cardiac structures can guide modern radiotherapy planning in patients with newly diagnosed HL by providing specific dose constraints.
Survivors of HL remain at substantially increased risk of various cardiovascular diseases for at least 40 years after HL diagnosis. A sizeable proportion of HL survivors develop multiple events over time. Treating physicians and patients should be aware of the persistently increased risk of cardiovascular diseases throughout life, and the results of our study may direct guidelines for follow-up of patients with HL.
Accepted for Publication: March 2, 2015.
Corresponding Author: Flora E. van Leeuwen, PhD, Department of Epidemiology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, the Netherlands (f.v.Leeuwen@nki.nl).
Published Online: April 27, 2015. doi:10.1001/jamainternmed.2015.1180.
Author Contributions: Ms van Nimwegen and Dr Schaapveld had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: van Nimwegen, Schaapveld, Aleman, van Leeuwen.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: van Nimwegen, Schaapveld.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: van Nimwegen, Schaapveld, van Leeuwen.
Obtained funding: van Leeuwen.
Administrative, technical, or material support: Janus, Krol, Petersen, Raemaekers, Kok.
Study supervision: Schaapveld, Aleman, van Leeuwen.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grant NKI 2010-4720 from the Dutch Cancer Society (Dr van Leeuwen).
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Previous Presentation: An abstract from preliminary analyses of this study was presented at the American Society of Clinical Oncology Annual Meeting; June 3, 2014; Chicago, Illinois.
Additional Contributions: K. Kooijman, MSc, S. Fase, MSc, S. de Vries, MSc, and M. van Wouwe, MD, from The Netherlands Cancer Institute collected data for the study. We are indebted to the thousands of physicians who provided follow-up data for the study. No compensation was provided.