Coronary artery disease detected by coronary computed tomography angiography in adult survivors of childhood Hodgkin lymphoma

Authors

  • Daniel A. Mulrooney MD, MS,

    Corresponding author
    1. Department of Oncology, St. Jude Children's Research Hospital, Memphis, Tennessee
    2. Department of Epidemiology and Cancer Control, St. Jude Children's Research Hospital, Memphis, Tennessee
    3. Department of Pediatrics, University of Tennessee Health Science Center, Memphis, Tennessee
    4. Department of Medicine, University of Tennessee Health Science Center, Memphis, Tennessee
    • Corresponding author: Daniel A. Mulrooney, MD, MS, Division of Cancer Survivorship, Department of Oncology, St. Jude Children's Research Hospital, MS 735, 262 Danny Thomas Pl, Memphis, TN 38105; Fax: (901) 595-5845; daniel.mulrooney@stjude.org.

    Search for more papers by this author
  • Sara E. Nunnery BA,

    1. Department of Pediatrics, University of Tennessee Health Science Center, Memphis, Tennessee
    2. Department of Medicine, University of Tennessee Health Science Center, Memphis, Tennessee
    Search for more papers by this author
  • Gregory T. Armstrong MD, MSCE,

    1. Department of Epidemiology and Cancer Control, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • Kirsten K. Ness PhD, PT,

    1. Department of Epidemiology and Cancer Control, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • Deokumar Srivastava PhD,

    1. Department of Biostatistics, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • F. Daniel Donovan MD, PhD,

    1. Department of Radiology, Methodist Hospitals of Memphis, Memphis, Tennessee
    2. Department of Radiology, University of Tennessee Health Science Center, Memphis, Tennessee
    Search for more papers by this author
  • Beth A. Kurt MD,

    1. Department of Pediatrics, Helen DeVos Children's Hospital, Grand Rapids, Michigan
    Search for more papers by this author
  • Monika L. Metzger MD,

    1. Department of Oncology, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • Matthew J. Krasin MD,

    1. Department of Radiological Sciences, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • Vijaya Joshi MD,

    1. Department of Pediatrics, University of Tennessee Health Science Center, Memphis, Tennessee
    2. Department of Medicine, University of Tennessee Health Science Center, Memphis, Tennessee
    Search for more papers by this author
  • Jean-Bernard Durand MD,

    1. Department of Cardiology, University of Texas MD Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Leslie L. Robison PhD,

    1. Department of Epidemiology and Cancer Control, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • Melissa M. Hudson MD,

    1. Department of Oncology, St. Jude Children's Research Hospital, Memphis, Tennessee
    2. Department of Epidemiology and Cancer Control, St. Jude Children's Research Hospital, Memphis, Tennessee
    Search for more papers by this author
  • Scott D. Flamm MD, MBA

    1. Departments of Diagnostic Imaging and Cardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio
    Search for more papers by this author

Abstract

BACKGROUND

Survivors of Hodgkin lymphoma (HL) have significant cardiovascular risk and require long-term surveillance. The current study assessed the prevalence of coronary artery disease (CAD) by coronary computed tomography angiography (CCTA) in adult survivors of childhood HL.

METHODS

Thirty-one survivors of HL, 13 of whom (42%) were treated with radiotherapy (RT) only and 18 of whom (58%) were treated with multimodal therapy, underwent CCTA, echocardiography, electrocardiography (ECG), and treadmill stress testing. Obstructive CAD was defined as ≥50% occlusion of the left main or ≥70% occlusion of the left anterior descending, left circumflex, or right coronary arteries on CCTA. Echocardiograms with resting wall motion abnormalities or an ejection fraction <50%; ECGs with Q waves, ST abnormalities without Q waves, or T-wave abnormalities without Q waves; and a J-point depression of ≥1 mm with a horizontal or downsloping ST segment on stress testing were considered abnormal.

RESULTS

The prevalence of disease in participants (median age, 40 years [range, 26 years-55 years]; median time from cancer diagnosis, 24 years [range, 17 years-39 years]) was 39%, with 39 plaques detected among 12 survivors. Three participants (10%) treated with RT only had 4 obstructive lesions; 9 patients (29%; 5 of whom were treated with RT only and 4 of whom were treated with multimodal therapy) had nonobstructive lesions. Approximately 15% of lesions involved the left main, 21% involved the proximal left anterior descending, 18% involved the proximal right coronary, and 13% involved the proximal left circumflex arteries. Of the 12 participants found to have CAD by CCTA, 7 had a positive ECG, 1 had a positive echocardiogram, and 1 had a positive stress test.

CONCLUSIONS

CCTA identified CAD in a substantial percentage of survivors of HL and may be an effective screening modality for this population. Cancer 2014;120:3536–3544. © 2014 American Cancer Society.

INTRODUCTION

The successful treatment of pediatric Hodgkin lymphoma (HL) has evolved over the last 5 decades from treatment with high-dose radiotherapy (RT) alone to risk-adapted and response-adapted regimens combining chemotherapy with lower radiation doses and smaller lymph node fields. Recent protocols are now testing the elimination of radiation altogether.[1] During the 1960s and 1970s, RT doses ≥35 gray (Gy) directed at wide lymph node fields were frequently curative for patients with localized disease but ineffective for those with advanced systemic disease.[2] The introduction of multimodal therapy and refinements in staging, imaging, and response assessments have increased 5-year survival rates for pediatric HL to nearly 97%.[3]

However, survival rates over time are lower compared with expected rates in the US population, with only 74.1% (95% confidence interval [95% CI], 71.8%-76.6%) of survivors of HL alive at 30 years from diagnosis.[4] Cardiovascular disease, the leading noncancer cause of death among survivors of HL, has been associated with exposure to anthracyclines and mediastinal RT and can manifest as coronary artery disease (CAD), congestive heart failure, and/or cerebrovascular accidents.[5, 6] The highest risk has been reported in patients aged <25 years at the time of treatment, among whom the 30-year cumulative incidence rates for any cardiovascular disorder and myocardial infarction are 34.5% and 12.9%, respectively.[7] These risks increase with longer follow-up.

Monitoring for cardiovascular disease, proposed by several groups,[8-10] has focused on cardiomyopathy screening with recommendations for the measurement of cardiac function in survivors exposed to cardiac-directed RT and/or anthracyclines. Additional cardiac imaging recommendations are directed by exposure factors and results of the initial screening. Cardiology consultation is suggested at 5 to 10 years after radiation exposure to evaluate for CAD among those treated with ≥40 Gy alone or ≥30 Gy plus an anthracyline.[8]

To the best of our knowledge, data describing coronary computed tomography angiography (CCTA) as a tool to screen for CAD in asymptomatic, high-risk survivors of HL are limited. Rademaker et al[11] reported CAD in 8 of 9 survivors of HL who had been treated with 34 to 45 Gy of mediastinal radiation. None of these patients had electrocardiogram (ECG) findings suggestive of acute or prior ischemia. Kupeli et al[12] used CCTA to evaluate 119 asymptomatic survivors of HL (mean age, 20 years; range, 6 years-43 years) who had been treated with RT and/or cardiotoxic chemotherapy ≥2 years previously; coronary artery abnormalities were reported in 19 patients (16%). The risk was found to be significantly higher among those treated with mediastinal radiation of >20 Gy (relative risk, 6.8; 95% CI, 1.6-28.8) compared with those exposed to ≤20 Gy. To our knowledge, subclinical CAD in a high-risk population of older survivors (mean age, 40 years) who are further from diagnosis has not yet been investigated with CCTA.

The objective of the current study was to determine the prevalence of CAD detected by CCTA in asymptomatic long-term survivors of HL exposed to chest RT alone or multimodal therapy (RT plus chemotherapy), and to compare these results with standard noninvasive screening tests.

MATERIALS AND METHODS

Cases for this study were participants in the St. Jude Lifetime Cohort (SJLIFE) Study, an ongoing study of patients previously treated at St. Jude Children's Research Hospital that was designed to facilitate longitudinal evaluation of health outcomes among adults previously treated for a pediatric malignancy.[13] Medical histories and detailed treatment information are abstracted from the medical records, and participants undergo a core assessment battery and risk-based medical screening based on the Children's Oncology Group Long-Term Follow-Up Guidelines for Survivors of Childhood, Adolescent, and Young Adult Cancers.[8] Informed consent was obtained from all participants, and the protocol was approved by the St. Jude Children's Research Hospital Institutional Review Board.

A convenience sample of survivors of HL participating in the cohort was recruited for this pilot protocol. Eligible patients were ≥15 years past their HL diagnosis, aged ≤55 years, and had received treatment with chest RT alone or multimodal therapy. Participants were excluded if they had an implanted medical device, irregular cardiac rhythm, or were allergic to computed tomography (CT) contrast; were unable to hold their breath for CT imaging or walk on a treadmill; or were pregnant. In addition, participants with a history of congenital heart disease, congestive heart failure, myocardial infarction, or coronary artery revascularization (percutaneous or surgical) were not included.

Survivors with a body mass index (BMI) ≥25 kg/m2 and <30 kg/m2 were considered overweight; those with a BMI ≥30 kg/m2 were considered obese. Hypertension was defined as blood pressure ≥140/90 mm Hg and/or treatment with an antihypertensive. Dyslipidemia was defined as any abnormality on a fasting lipid panel (total cholesterol >200 mg/dL, low-density lipoprotein cholesterol >130 mg/dL, high-density lipoprotein cholesterol <40 mg/dL, and triglycerides >150 mg/dL) and/or treatment with a lipid-lowering agent.

Coronary Artery Evaluation With CCTA

CCTA was performed using a 64-detector CT scanner (Light Speed VCT; GE Healthcare, Milwaukee, Wis). A power injector (EZEM Empower; Bracco Diagnostics Inc, Monroe Township, NJ) was used to deliver contrast media through an 18-gauge needle in an antecubital vein. Depending on patient weight, 80 to 120 mL of iohexol at 350 mg iodine/mL (Omnipaque 350; GE Healthcare) was administered.

A test bolus was used to determine the timing of scan delay and image acquisition. Primary scanning parameters were as follows: 64 detectors; individual detector width, 0.625 mm; gantry rotation time, 350 ms; pitch, 0.24; tube voltage, 120 kilovolts; and tube current, 550 milliamperes during 40% to 80% R wave-to-R wave (RR) interval, decreased to as low as 200 milliamperes during the remainder of the RR interval for retrospective technique.

Prospective low-dose scanning was performed if the patient's heart rate was regular and <65 beats per minute. Patients with a prescan heart rate of ≥65 beats per minute were given 50 mg of metoprolol orally 1 hour before scanning. Retrospective technique with tube current modulation was used if the patient's heart rate was >65 beats per minute and ≤85 beats per minute. CCTA was not performed if the heart rate was >85 beats per minute. On 1 occasion, an initial prospective scan was unsatisfactory, and a second scan was performed with retrospective technique. Thirteen scans used prospective technique and 18 scans used retrospective technique.

CCTA Image Analysis

The coronary tree was segmented according to the modified American Heart Association classification.[14] Original source images, selected maximum-intensity projections, and curved multiplanar reconstructions were individually evaluated by 1 examiner (S.D.F.). Obstructive CAD was defined as ≥50% occlusion of the left main (LM) coronary artery or ≥70% occlusion of the left anterior descending (LAD) artery, left circumflex (LCx) artery, or right coronary artery (RCA).

Other Screening Methods

Two-dimensional Doppler ultrasound echocardiography (VIVID-7; GE Healthcare) was performed as per the American Society of Echocardiography guidelines.[15] Resting wall motion abnormalities or an ejection fraction <50% was considered abnormal.

Twelve-lead ECG was performed and centrally coded as per the Novacode criteria for classification. Tracings were considered positive for CAD if coded a high likelihood of Q-wave myocardial infarction (MI) (Q-wave MI with major Q waves or Q-wave MI with moderate Q waves with ST-T abnormalities), a moderate likelihood of Q-wave MI (possible Q-wave MI with moderate Q-waves without ST-T abnormalities or possible Q-wave MI with minor Q-waves with ST-T abnormalities), or isolated ischemic abnormalities (ST abnormalities without Q-waves or T-wave abnormalities without Q-waves).[16]

Treadmill stress testing was performed as per the Balke Treadmill Protocol, in which the walking speed was held constant at 3 km/hour while the grade was increased by 2.5% every 2 minutes. Observation of a J-point depression ≥1 mm with a horizontal or downsloping ST segment was considered to be positive for CAD.

Statistical Analysis

The usefulness of CCTA in detecting CAD was evaluated by comparing the prevalence of abnormalities detected by CCTA and other screening tests such as echocardiography, ECG, and treadmill stress testing.

Radiation exposure from the CCTA scan was only available for 19 study participants. The 2 primary factors affecting exposure were the type of gating used and the patient's body habitus. The Fisher exact test was used to assess whether the frequency distribution of prospective and retrospective ECG-gated coronary artery acquisitions differed significantly between the 2 groups, and a 2-sample Student t test, assuming unequal variances, was used to assess whether the BMI distribution significantly differed between the 2 groups (19 patients with an available CCTA radiation dose vs 12 patients for whom the dose was unavailable).

RESULTS

Forty-five SJLIFE participants agreed to additional participation in this study. Thirteen participants could not complete the study because of CT contrast allergy (2 patients), tachycardia (5 patients), renal failure (1 patient), failed intravenous access (2 patients), or refusal (3 patients). Of the 31 survivors (69%) enrolled, 13 were treated with RT alone and 18 were treated with multimodal therapy. The median age of the patients was 40 years (range, 26 years-55 years) at the time of evaluation, and the median time from the initial diagnosis was 24 years (range, 17 years-39 years) (Table 1). Those patients treated with RT only were slightly older (median, 42 years; range, 37 years-55 years) than those treated with multimodal therapy (median, 38 years; range, 26 years-46 years). Patients treated with RT only were also further from therapy (median, 33 years; range, 21 years-39 years) than those treated with multimodal therapy (median, 22 years; range, 18 years-32 years). Cardiovascular risk factors are detailed in Table 1. The majority of patients were considered to be at low risk based on National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) risk scoring for asymptomatic adults,[17] and only 1 patient would have met recommendations for coronary artery calcium screening.[18]

Table 1. Demographic and Treatment Characteristics of Survivors of HL
Patient CharacteristicTotal (n = 31)Radiotherapy Only (n = 13)Multimodal Therapy (n = 18)
No. (%)No. (%)No. (%)
  1. Abbreviations: BMI, body mass index; Gy, gray; HL, Hodgkin lymphoma; NA, not applicable.

Sex
Male12 (39)5 (39)7 (39)
Female19 (61)8 (62)11 (61)
Age, y
25-341 (3)0 (0)1 (6)
35-4425 (81)9 (69)16 (89)
45-555 (16)4 (31)1 (6)
Age at cancer diagnosis, y
Birth-96 (19.5)4 (31)2 (11)
10-146 (19.5)4 (31)2 (11)
15-1919 (61)5 (38)14 (78)
Y since diagnosis
15-2418 (58)3 (23)15 (83)
25-348 (26)5 (38.5)3 (17)
35-395 (16)5 (38.5)0 (0)
Race/ethnicity
Non-Hispanic white26 (84)12 (92)14 (78)
Non-Hispanic black5 (16)1 (8)4 (22)
Treatment
Anthracycline exposure
<150 mg/m26 (19)NA6 (33)
150-249 mg/m29 (29)NA9 (50)
≥250 mg/m23 (10)NA3 (17)
Chest radiotherapy
<20 Gy3 (10)0 (0)3 (17)
20-29 Gy13 (42)0 (0)13 (72)
≥30 Gy15 (48)13 (100)2 (11)
Cardiovascular risk factors
Overweight (BMI 25-29)18 (58)6 (46)12 (67)
Obesity (BMI ≥30)5 (16)1 (8)4 (22)
Diabetes mellitus1 (3)0 (0)1 (6)
Hypertension7 (23)2 (15)5 (28)
Dyslipidemia15 (48)6 (46)9 (50)
Smoker
Current6 (19)3 (23)3 (17)
Past7 (23)3 (23)4 (22)
Never18 (58)7 (54)11 (61)

All survivors were exposed to RT. Those treated with RT alone received ≥30 Gy, whereas only 2 patients in the multimodal group (11%) received doses that high; the majority of patients (16 patients; 89%) received 20 to 29 Gy. Survivors in the multimodal group had a median anthracycline exposure of 191 mg/m2 (range, 96 mg/m2−316 mg/m2).

A total of 39 coronary artery lesions were identified in 12 patients: 59% of the lesions were calcified, 26% were noncalcified, and 15% were mixed (Figs. 1 and 2). Four lesions were obstructive (10%), whereas 35 lesions (90%) were nonobstructive. Approximately 67% of lesions (obstructive and nonobstructive) were located proximally in the coronary artery tree, and 74% involved the left-sided circulation (Table 2) (Fig. 2). Six lesions (15%) were identified in the LM coronary artery; 5 resulted in a <25% reduction in diameter, but 1 had 50% to 70% stenosis. Eight lesions (21%) were identified in the proximal LAD artery, 5 (13%) in the proximal LCx artery, and 7 (18%) in the proximal RCA.

Figure 1.

Coronary computed tomography angiography multiplanar reconstruction in the 2-chamber orientation is shown demonstrating an obstructive lesion of the proximal left anterior descending artery (arrow) secondary to mixed (calcified and noncalcified) atherosclerotic plaque.

Figure 2.

Coronary computed tomography angiography (Left) 3-dimensional reconstruction and (Right) curved multiplanar reconstruction from the same patient are shown revealing a nonobstructive lesion (50%-70% stenosis) of the middle left anterior descending artery (arrows) secondary to noncalcified atherosclerotic plaque.

Table 2. Coronary Artery Disease Detected by CCTA (N = 39)
Location of PlaqueNo. (%)
  1. Abbreviation: CCTA, coronary computed tomography angiography.

Left main artery6 (15)
Left anterior descending artery
Proximal8 (21)
Middle6 (15)
Distal1 (3)
Diagonals2 (5)
Left circumflex artery
Proximal5 (13)
Distal0 (0)
Right coronary artery
Proximal7 (18)
Middle2 (5)
Distal2 (5)

The overall prevalence of obstructive and nonobstructive disease detected by CCTA was 39%. Four obstructive lesions were identified in 3 participants (10%), all of whom were treated with RT only. These lesions were proximally located (Table 3) (Fig. 3). One individual had stenosis of the LAD and LCx arteries, 1 had RCA disease, and 1 had a 50% to 70% narrowing of the LM coronary artery. All these patients subsequently underwent conventional angiography, with confirmation of disease in all 3. Only 1 patient reported a history of angina. Two patients underwent surgical revascularization, and 2 subsequently died of cardiovascular disease (1 with and 1 without revascularization).

Table 3. Obstructive Lesions Detected by CCTA
Age, YearsSexRadiation, cGyAnthracycline, mg/m2CCTAEchocardiogramStress TestElectrocardiogram10-Year CV RiskStatus
  1. Abbreviations: CABG, coronary artery bypass graft; CAD, coronary artery disease; CCTA, coronary computed tomography angiography; cGy, centigray; CV, cardiovascular; D1, first acute diagonal branch of the left anterior descending artery; EF, ejection fraction; LAD, left anterior descending artery; LCx, left circumflex artery; LM, left main artery; RCA, right coronary artery.

43Female3500050%-70% LMEF 54%; no wall motion abnormalitiesNormalMajor Q-wave abnormality1%3-vessel CABG; died of CAD
25%-50% proximal LAD
25%-50% middle LAD
25%-50% proximal LCx
50%-70% proximal RCA
40Male36000>70% proximal LADEF 51%; no wall motion abnormalitiesAbnormalMajor Q-wave abnormality2%6-vessel CABG; alive
<25% middle LAD
25%-50% D1
>70% proximal LCx
25%-50% proximal RCA
<25% middle RCA
53Female39000>70% proximal RCAEF 51%; no wall motion abnormalitiesNormalMajor Q-wave abnormality1%Died of CAD
<25% LM
<25% proximal LAD
25%-50% middle LAD
<25% distal LAD
<25% proximal LCx
Figure 3.

Schematic plot of the distribution of lesions identified by coronary computed tomography angiography along coronary territories is shown, demonstrating clustering of lesions in the proximal coronary arteries. Red indicates survivors of Hodgkin lymphoma who were treated with radiotherapy only; blue, survivors of Hodgkin lymphoma who were treated with multimodal therapy; red squares, obstructive plaque; red and blue circles, nonobstructive plaque. RCA indicates right coronary artery; LAD, left anterior descending artery; LCx, left circumflex artery; D1, first acute diagonal branch of the left anterior descending artery; D2, second acute diagonal branch of the left anterior descending artery.

Nine survivors (29%), 5 of whom were treated with RT only and 4 with multimodal therapy, had nonobstructive lesions (Table 4), with most stenoses having a <25% to 50% reduction in diameter. One individual had a 50% to 70% middle LAD stenosis. One individual had a calcified lesion with >70% stenosis of the RCA ostium identified by CCTA. Coronary angiography revealed only nonobstructive disease, although review of the images revealed a small vessel with coronary spasm during the procedure, potentially confounding the findings. None of these patients had clinically evident CAD.

Table 4. Nonobstructive Lesions Detected by CCTA
Age, YearsSexRadiation, cGyAnthracycline, mg/m2CCTAEchocardiogramStress TestElectrocardiogram10-Year CV RiskStatus
  1. Abbreviations: CCTA, coronary computed tomography angiography; cGy, centigray; CV, cardiovascular; D1, first acute diagonal branch of the left anterior descending artery; D2, second acute diagonal branch of the left anterior descending artery; EF, ejection fraction; LAD, left anterior descending artery; LCx, left circumflex artery; LM, left main artery; LVH, left ventricular hypertrophy; RCA, right coronary artery; SMN, second malignant neoplasm.

42Male35110>70% proximal RCAEF 59%; no wall motion abnormalitiesNormalNegative1%Alive
<25% LM
37Male36000<25% proximal LADEF 59%; no wall motion abnormalitiesNormalNegative3%Died of SMN
25%-50% middle LAD
<25% D2
43Male35000<25% LMEF 60%; no wall motion abnormalitiesNormalMajor isolated ST-T abnormalities14%Alive
<25% proximal LAD
25%-50% proximal LCx
<25% proximal RCA
39Female38500<25% LMEF 64%; no wall motion abnormalitiesNormalNegative1%Alive
55Female38400<25% proximal LADEF 62%; no wall motion abnormalitiesNormalMajor Q-wave abnormality1%Alive
41Male2075149<25% LMEF 54%; no wall motion abnormalitiesNormalNegative1%Alive
50%-70% middle LAD
25%-50% proximal D1
<25% proximal RCA
42Female205013625-50% proximal LADEF 66%; no wall motion abnormalitiesInvalidIsolated ST/T abnormalities; LVH5%Alive
<25% proximal LCx
<25% proximal RCA
25%-50% distal RCA
37Female1920147<25% proximal LADEF 68%; no wall motion abnormalitiesNormalNegative1%Alive
46Male2000170<25% middle LADEF 69%; no wall motion abnormalitiesNormalIsolated ST/T abnormalities1%Alive
<25% middle RCA

Radiation exposure from CCTA as measured in dose-length product was 661 ± 350 milligray-cm (mean ± standard deviation) among 19 patients for whom dose was archived. The BMI of this subset of patients was no different from those for whom the dose was not archived (25.6 kg/m2 vs 27.8 kg/m2; P =.25), and the ratio of prospective to retrospective ECG-gated coronary artery acquisitions was not significantly different between the 2 groups (P =1.0), suggesting that the patient subset RT dose is a reasonable estimate for the entire studied population.

Other than ECG abnormalities, few positive findings were identified by the other testing modalities. Among the 3 individuals with obstructive lesions identified by CCTA, all had positive findings on resting ECG, 1 had evidence of ischemia on treadmill stress testing, and no patient had an abnormal resting echocardiogram. None of the patients with nonocclusive lesions had positive stress testing or echocardiography; 4 of the patients with nonocclusive plaques had positive findings on resting ECG. Among those participants with no evidence of CAD on CCTA (19 patients), no abnormalities on stress testing were identified. On echocardiogram, 2 patients had an ejection fraction <50% (46% and 47%, respectively); 2 others had ST-T segment abnormalities on resting ECG.

DISCUSSION

In this population of survivors of HL, we found evidence of obstructive and nonobstructive coronary plaque on CCTA, without significant corresponding changes on echocardiography or treadmill stress testing, although all 3 patients with obstructive CAD had positive ECGs. The majority of the coronary artery plaques were calcified, with approximately two-thirds located in proximal arterial segments and 15% in the LM coronary artery.

Radiation-induced CAD is a known late effect after treatment for HL, with few noninvasive screening options and little consensus regarding the best method for long-term monitoring.[19] Historically, screening has included echocardiography with variable follow-up after the completion of therapy. The median age of patients with CAD in this study was 40 years, which is significantly younger for a diagnosis of CAD compared with a noncancer population,[20-22] suggesting that this patient population requires more vigilant screening. Multidetector CCTA is a potentially useful screening modality for these patients. In an investigation of adults without known CAD who were evaluated for nonemergent chest pain, researchers reported sensitivity, specificity, and positive and negative predictive values of 94%, 83%, 48%, and 99%, respectively, for CCTA in detecting ≥70% stenosis.[23] Ha et al[20] used CCTA to evaluate healthy young adults (aged ≤40 years) with no history of cancer and found that 11% of study participants had evidence of occult CAD, but no cases were obstructive. In the similarly young population in the current study, 10% of patients had obstructive disease; an additional 29% had nonobstructive lesions. Both obstructive and nonobstructive plaques have been associated with future adverse cardiovascular events.[24, 25] Alternatively, a lack of coronary plaque on CCTA is associated with a low probability of a future event (negative likelihood ratio, 0.008; 95% CI, 0.0004-0.17).[25] Approximately 61% of our population had no identifiable plaque. CCTA screening may therefore help clinicians to identify and risk-stratify survivors who would benefit from closer surveillance or earlier medical interventions.

Studies designed to assess ischemic heart disease in this patient population have been limited. Rademaker et al[11] described CAD in 8 of 9 survivors of HL who received RT (5 of whom were diagnosed as children [aged ≤21 years]); the majority had diffuse disease but only 2 patients had obstructive disease as defined in the current study. Unlike the current study, the study by Rademaker et al also assessed coronary calcium scores and found that 6 survivors had scores >90th percentile for age and sex and 4 survivors had values 15 years beyond their current age.[11] In a cohort of irradiated and nonirradiated survivors of childhood HL, Kupeli et al[12] found that 19 survivors (16%) had coronary abnormalities of varying degrees, albeit only 2 with obstructive disease as defined in the current study. This population was younger than that in the current study, with a median age of 23 years (range, 14 years-43 years) and a median of 10 years (range, 5 years-31 years) from the time of diagnosis, suggesting potential progression of CAD over time. In this study, 53% of those with coronary abnormalities and 47% of those without such abnormalities were also reported to have echocardiographic irregularities, although specific information regarding these abnormalities was not provided. In the current study, we focused on resting wall motion abnormalities or an ejection fraction <50% and found no positive echocardiograms in the 12 patients with CAD. However, 2 of the 19 patients without CAD had mildly reduced global left ventricular ejection fraction, suggesting a cardiomyopathy that was not attributable to epicardial coronary disease.

It is interesting to note that many of the ECG recordings in the current study were suggestive of ischemic heart disease. Frequently used to assess for arrhythmias or conduction disorders among survivors of HL, ECGs may have added benefit for CAD screening. Among the small population in the current study, the positive and negative predictive values of ischemic ECG findings were 63% and 25%, respectively. Formal assessment in the larger SJLIFE cohort is ongoing and likely to be more informative.

Unlike previous studies, the current study included concurrent treadmill stress testing, which identified a single abnormal test. Given initial concerns regarding maximally stressing at-risk survivors of HL, the current study did not use a Bruce Treadmill Test Protocol, thus potentially limiting the overall yield from treadmill stress testing. However, the use of more intensive stress testing in these patients should be approached with caution because this group has a higher incidence of stenotic aortic valvular disease and underlying conduction abnormalities such as left bundle branch block,[5] which could make the Bruce stress testing challenging. The Balke protocol has a moderate workload and is suitable for patients with left ventricular systolic dysfunction.

The findings of the current study should be interpreted within the context of evolving therapies for HL. The most severe disease (obstructive lesions) was identified among those individuals with the highest radiation exposures. Doses have been reduced, if not eliminated, in newer protocols, which will hopefully lead to a reduced risk of CAD among survivors of HL. However, data from survivors of breast cancer who were treated with RT would suggest that there is no safe exposure, with a 7.4% increase in coronary events per Gy of radiation to the heart.[26] Progression and the long-term outcome of this subclinical disease may not yet be evident.

Several limitations should be considered when interpreting the results of the current study. The small sample size limited the power for detailed analyses of treatment exposures and the potential contribution of traditional cardiovascular risk factors. Reflecting the trend in the general population, many of the patients in the current study cohort were overweight, hypertensive, or had dyslipidemia, potentially contributing to or significantly worsening CAD.[27] The current study did not characterize the atherosclerotic plaque beyond the presence or absence of calcium. More detailed descriptions of plaque character (eg, lipid component, surface integrity, and degree of vascular remodeling) are not part of routine clinical CCTA scanning and analysis. Historically, CCTA has required limited but not insignificant radiation exposures (on average, approximately 12 millisieverts).[28] Recent technical advances, however, have reduced this exposure, in some cases to <1 millisievert.[29]

The population of survivors of childhood HL in the current study had a significantly higher burden of CAD (39%) noted on CCTA than what has been reported among the similarly aged general population (8.5%-11%).[20-22] The lesions clustered in a proximal distribution, thereby increasing the risk of larger myocardial infarctions and coronary death. Given the rarity of ischemic heart disease among young people, few of these asymptomatic survivors of HL would typically have been screened for this condition. CCTA may offer an effective noninvasive tool for the management of at-risk cancer survivors, permitting the identification of those who might benefit from lifestyle and/or medication interventions and those for whom lower levels of monitoring may be appropriate.

FUNDING SUPPORT

Supported by Cancer Center Support (CORE) Grant CA21765 to St. Jude Children's Research Hospital and the American Lebanese Syrian Associated Charities (ALSAC) (Memphis, Tennessee).

CONFLICT OF INTEREST DISCLOSURES

Dr. Ness was supported by a grant from the National Institutes of Health for work related to the current study.

Ancillary