Cardiac toxicity is one of the most concerning side effects of anti-cancer therapy. The gain in life expectancy obtained with anti-cancer therapy can be compromised by increased morbidity and mortality associated with its cardiac complications. While radiosensitivity of the heart was initially recognized only in the early 1970s, the heart is regarded in the current era as one of the most critical dose-limiting organs in radiotherapy. Several clinical studies have identified adverse clinical consequences of radiation-induced heart disease (RIHD) on the outcome of long-term cancer survivors. A comprehensive review of potential cardiac complications related to radiotherapy is warranted. An evidence-based review of several imaging approaches used to detect, evaluate, and monitor RIHD is discussed. Recommendations for the early identification and monitoring of cardiovascular complications of radiotherapy by cardiac imaging are also proposed.
Introduction
The two major contributors to radiation exposure in the population are ubiquitous background radiation and medical exposure. A high-dose radiation exposure on the thorax is mainly used in the context of adjuvant radiotherapy after conservative or radical breast surgery, adjuvant or exclusive radiotherapy of lung and oesophageal cancer, and as a complement to systemic treatment in lymphoma. Irradiation of the heart increases the risk of the so-called ‘radiation-induced’ heart disease (RIHD). RIHD is generated by total cumulative dosage of radiotherapy potentiated by the adjunctive chemotherapy. The total cumulative dosage of radiotherapy is a function of the number of treatments and the dose of irradiation. The manifestations of RIHD may acutely develop but most often become clinically apparent several years after irradiation. RIHD holds a wide range of deleterious effects on the heart including pericarditis, coronary artery disease (CAD), myocardial infarction, valvular heart disease, rhythm abnormalities, and non-ischaemic myocardial and conduction system damages. The number of patients at risk of developing RIHD is likely to increase as ∼40% of cancer survivors are at least 10 years past their radiotherapy treatment. The development of RIHD may be accelerated by the contribution of shared common risk factors of cardiovascular disease and cancer such as obesity, inactivity, and substance abuse (i.e. tobacco and alcohol). Several clinical trials and epidemiologic studies have revealed the adverse impact of RIHD on the outcome of long-term cancer survivors. Appropriate recognition of potential cardiac complications related to radiotherapy is warranted in our day-to-day clinical practice. Several imaging approaches can be used to detect, evaluate, and monitor RIHD. This document represents a consensus summary by experts of an extensive review of the literature regarding the role of cardiac imaging in the detection and serial monitoring of RIHD.
Radiation Effects on the Heart
Prevalence
Evidence of the dose-dependent increase in cardiovascular disease after chest radiotherapy has been documented in several studies, especially in the field of breast cancer and lymphoma ( Table 1 ). The estimated aggregate incidence of RIHD is 10–30% by 5–10 years post-treatment. Among these patients who have received radiation, cardiovascular disease is the most common non-malignant cause of death. Comparing the long-term benefits and risks, the positive effect of adjuvant radiotherapy may thus be partially offset by cardiac complications. However, the precise prevalence of RIHD is difficult to determine because currently available data mainly come from single-centre studies, often retrospective, in which old radiotherapy techniques were used, patients with a prior history of CAD were excluded, and baseline pre-radiotherapy imaging was lacking. The prevalence of RIHD in the setting of modern protocols of delivering adjuvant radiotherapy, reduction in doses, and field radiation size is still poorly defined.
| Types | Hodgkin’s disease relative risk | Breast cancer relative risk |
|---|---|---|
| RIHD | >6.3 | 2–5.9 |
| Ischaemic heart disease | 4.2–6.7 | 1–2.3 |
| Cardiac death | 2.2–12.7 | 0.9–2 |
Population Risk Factors
Despite considerable uncertainty, we are increasing our understanding of the factors that may influence the long-term risk of RIHD ( Table 2 ). However, risk factors modulating the acute effects of cardiac radiation are hardly known. It appears that the cumulative dose and its fractioning determine acute and chronic cardiac effects of radiation therapy. In the past, pericarditis used to be the most common side effect in patients receiving traditional radiotherapy for Hodgkin’s disease. Dose restriction to 30 Gy with lower daily fraction, different weighting of radiation fields, and blocking of the sub-carinal region have been reported to reduce the incidence of pericarditis from 20 to 2.5%. While, in doses >30 Gy, the risk of RIHD becomes apparent, the nature and magnitude of lower doses is not well characterized nor is it clear whether there is a threshold dose below which there is no risk. Radiation increases the risk of cardiotoxic effects of certain chemotherapeutic agents, such as anthracyclines. This interaction appears to be dependent on the total cumulative dose of anthracyclines. Other patients and disease-related factors may potentially influence cardiac risk after ionizing radiation. Age at irradiation for breast cancer has been shown to influence the risk; patients younger than 35 have a relative risk of 6.5 than the general population of RIHD. Similar observations have been made in the case of Hodgkin’s lymphoma. Smoking also increases the relative risk. Other risk factors such as diabetes, hypertension, overweight, and hypercholesterolaemia influence the overall risk. However, in some studies, no increase in cardiac risk, especially of myocardial infarction, has been observed after adjusting for pre-existing cardiovascular risk factors.
| Anterior or left chest irradiation location |
| High cumulative dose of radiation (>30 Gy) |
| Younger patients (<50 years) |
| High dose of radiation fractions (>2 Gy/day) |
| Presence and extent of tumour in or next to the heart |
| Lack of shielding |
| Concomitant chemotherapy (the anthracyclines considerably increase the risk) |
| Cardiovascular risk factors (i.e. diabetes mellitus, smoking, overweight, ≥moderate hypertension, hypercholesterolaemia) |
| Pre-existing cardiovascular disease |
Pathophysiology
It is known that irradiation of a thoracic region encompassing the heart might be at the origin of acute and chronic RIHD. Current knowledge about acute radiation effects mainly derives from animal experiments, which do not necessarily reflect contemporary radiotherapy treatment strategies, neither in dosage nor in timing of irradiation. Furthermore, the processes from the acute injury to progressive cardiac disease and the relationship between short-term effects and long-term risks in each individual patient are still subject to investigations and not fully understood. Ionizing radiation might harm virtually all cardiac tissues and the underlying pathophysiological mechanisms may be related to micro- and macrovascular damages. Early events in the post-radiation cascade are loss of endothelial cells with subsequent inflammatory responses, driving the vascular damage. Microvascular damage (decrease in capillary density resulting in ischaemia) is associated with eventual fibrosis and diastolic dysfunction and heart failure. Primary radiation fibrosis is not related to the primary effect of radiation, but rather to a reparative response of the heart tissue to injury in the microvascular system ( Figure 1 ). This is a common pathological feature of late radiation tissue complications. Macrovascular damage includes accelerated atherosclerosis yielding endothelial dysfunction and coronary artery stenosis. The pathogenesis of this radiation-induced CAD shares common pathways with CAD driven by genetic and exogenous factors. As exogenous factors have been shown to result in genomic instability, and as low-dose radiation induces long-lasting genomic instability, a synergistic interaction between radiation-induced effects and pathogenic events unrelated to radiation exposure is highly probable.
Acute and Chronic Cardiovascular Toxicity
The clinical translations of the above radiation-induced pathophysiological changes are pericarditis, valvular heart disease, myocardial damage, microvascular dysfunction, CAD, myocardial ischaemia, and restrictive cardiomyopathy. These clinical entities differ with regard to latency, radiation exposure pattern, and clinical presentation. Acute radiation effects are commonly subtle, difficult to assess in patients, and clinically less relevant. Acute radiation effects must be suspected and investigated in patients with cardiovascular complaints early after radiotherapy. The late manifestations of RIHD usually become clinically overt several years after radiation. The symptoms and signs of RIHD are, for the most part, indistinguishable from those encountered in patients with heart disease due to other aetiologies. Table 3 gives a summary of the pathophysiological manifestations of RIHD for different radiosensitive structures within the heart.
| Acute | Long-term |
|---|---|
| Pericarditis | Pericarditis |
| • Acute exudative pericarditis is rare and often occurs during radiotherapy as a reaction to necrosis/inflammation of a tumour located next to the heart. | • Delayed chronic pericarditis appears several weeks to years after radiotherapy. In this type, extensive fibrous thickening, adhesions, chronic constriction, and chronic pericardial effusion can be observed. It is observed in up to 20% of patients within 2 years following irradiation. |
| • Delayed acute pericarditis occurs within weeks after radiotherapy and can be revealed by either an asymptomatic pericardial effusion or a symptomatic pericarditis. Cardiac tamponade is rare. Spontaneous clearance of this effusion may take up to 2 years. | • Constrictive pericarditis can be observed in 4–20% of patients and appears to be dose-dependent and related to the presence of pericardial effusion in the delayed acute phase. |
| Cardiomyopathy | Cardiomyopathy |
| • Acute myocarditis related to radiation-induced inflammation with transient repolarization abnormalities and mild myocardial dysfunction. | • Diffuse myocardial fibrosis (often after a >30-Gy radiation dose) with relevant systolic and diastolic dysfunction, conduction disturbance, and autonomic dysfunction. |
| • Restrictive cardiomyopathy represents an advanced stage of myocardial damage due to fibrosis with severe diastolic dysfunction and signs and symptoms of heart failure | |
| Valve disease | Valve disease |
| • No immediate apparent effects. | • Valve apparatus and leaflet thickening, fibrosis, shortening, and calcification predominant on left-sided valves (related to pressure difference between the left and right side of the heart). |
| • Valve regurgitation more commonly encountered than stenosis. | |
| • Stenotic lesions more commonly involving the aortic valve. | |
| • Reported incidence of clinically significant valve disease: 1% at 10 years; 5% at 15 years; 6% at 20 years after radiation exposure. | |
| • Valve disease incidence increases significantly after >20 years following irradiation: mild AR up to 45%, ≥moderate AR up to 15%, AS up to 16%, mild MR up to 48%, mild PR up to 12%. | |
| Coronary artery disease | Coronary artery disease |
| • No immediate apparent effects. (Perfusion defects can be seen in 47% of patients 6 months after radiotherapy and may be accompanied by wall-motion abnormalities and chest pain. Their long-term prognosis and significance are unknown.) | • Accelerated CAD appearing in the young age. • Concomitant atherosclerotic risk factors further enhance the development of CAD. • Latent until at least 10 years after exposure. (Patients younger than 50 years tend to develop CAD in the first decade after treatment, while older patients have longer latency periods.) |
| • Coronary ostia and proximal segments are typically involved. | |
| • CAD doubles the risk of death; relative risk of death from fatal myocardial infarction varies from 2.2 to 8.8. | |
| Carotid artery disease | Carotid artery disease |
| • No immediate apparent effects. | • Radiotherapy-induced lesions are more extensive, involving longer segments and atypical areas of carotid segments. |
| • Estimated incidence (including sub-clavian artery stenosis) about 7.4% in Hodgkin’s lymphoma. | |
| Other vascular disease | Other vascular disease |
| • No immediate apparent effects. | • Calcification of the ascending aorta and aortic arch (porcelain aorta). |
| • Lesions of any other vascular segments present within the radiation field. |
Radiation Effects on the Heart
Prevalence
Evidence of the dose-dependent increase in cardiovascular disease after chest radiotherapy has been documented in several studies, especially in the field of breast cancer and lymphoma ( Table 1 ). The estimated aggregate incidence of RIHD is 10–30% by 5–10 years post-treatment. Among these patients who have received radiation, cardiovascular disease is the most common non-malignant cause of death. Comparing the long-term benefits and risks, the positive effect of adjuvant radiotherapy may thus be partially offset by cardiac complications. However, the precise prevalence of RIHD is difficult to determine because currently available data mainly come from single-centre studies, often retrospective, in which old radiotherapy techniques were used, patients with a prior history of CAD were excluded, and baseline pre-radiotherapy imaging was lacking. The prevalence of RIHD in the setting of modern protocols of delivering adjuvant radiotherapy, reduction in doses, and field radiation size is still poorly defined.
| Types | Hodgkin’s disease relative risk | Breast cancer relative risk |
|---|---|---|
| RIHD | >6.3 | 2–5.9 |
| Ischaemic heart disease | 4.2–6.7 | 1–2.3 |
| Cardiac death | 2.2–12.7 | 0.9–2 |
Population Risk Factors
Despite considerable uncertainty, we are increasing our understanding of the factors that may influence the long-term risk of RIHD ( Table 2 ). However, risk factors modulating the acute effects of cardiac radiation are hardly known. It appears that the cumulative dose and its fractioning determine acute and chronic cardiac effects of radiation therapy. In the past, pericarditis used to be the most common side effect in patients receiving traditional radiotherapy for Hodgkin’s disease. Dose restriction to 30 Gy with lower daily fraction, different weighting of radiation fields, and blocking of the sub-carinal region have been reported to reduce the incidence of pericarditis from 20 to 2.5%. While, in doses >30 Gy, the risk of RIHD becomes apparent, the nature and magnitude of lower doses is not well characterized nor is it clear whether there is a threshold dose below which there is no risk. Radiation increases the risk of cardiotoxic effects of certain chemotherapeutic agents, such as anthracyclines. This interaction appears to be dependent on the total cumulative dose of anthracyclines. Other patients and disease-related factors may potentially influence cardiac risk after ionizing radiation. Age at irradiation for breast cancer has been shown to influence the risk; patients younger than 35 have a relative risk of 6.5 than the general population of RIHD. Similar observations have been made in the case of Hodgkin’s lymphoma. Smoking also increases the relative risk. Other risk factors such as diabetes, hypertension, overweight, and hypercholesterolaemia influence the overall risk. However, in some studies, no increase in cardiac risk, especially of myocardial infarction, has been observed after adjusting for pre-existing cardiovascular risk factors.
| Anterior or left chest irradiation location |
| High cumulative dose of radiation (>30 Gy) |
| Younger patients (<50 years) |
| High dose of radiation fractions (>2 Gy/day) |
| Presence and extent of tumour in or next to the heart |
| Lack of shielding |
| Concomitant chemotherapy (the anthracyclines considerably increase the risk) |
| Cardiovascular risk factors (i.e. diabetes mellitus, smoking, overweight, ≥moderate hypertension, hypercholesterolaemia) |
| Pre-existing cardiovascular disease |
Pathophysiology
It is known that irradiation of a thoracic region encompassing the heart might be at the origin of acute and chronic RIHD. Current knowledge about acute radiation effects mainly derives from animal experiments, which do not necessarily reflect contemporary radiotherapy treatment strategies, neither in dosage nor in timing of irradiation. Furthermore, the processes from the acute injury to progressive cardiac disease and the relationship between short-term effects and long-term risks in each individual patient are still subject to investigations and not fully understood. Ionizing radiation might harm virtually all cardiac tissues and the underlying pathophysiological mechanisms may be related to micro- and macrovascular damages. Early events in the post-radiation cascade are loss of endothelial cells with subsequent inflammatory responses, driving the vascular damage. Microvascular damage (decrease in capillary density resulting in ischaemia) is associated with eventual fibrosis and diastolic dysfunction and heart failure. Primary radiation fibrosis is not related to the primary effect of radiation, but rather to a reparative response of the heart tissue to injury in the microvascular system ( Figure 1 ). This is a common pathological feature of late radiation tissue complications. Macrovascular damage includes accelerated atherosclerosis yielding endothelial dysfunction and coronary artery stenosis. The pathogenesis of this radiation-induced CAD shares common pathways with CAD driven by genetic and exogenous factors. As exogenous factors have been shown to result in genomic instability, and as low-dose radiation induces long-lasting genomic instability, a synergistic interaction between radiation-induced effects and pathogenic events unrelated to radiation exposure is highly probable.
Acute and Chronic Cardiovascular Toxicity
The clinical translations of the above radiation-induced pathophysiological changes are pericarditis, valvular heart disease, myocardial damage, microvascular dysfunction, CAD, myocardial ischaemia, and restrictive cardiomyopathy. These clinical entities differ with regard to latency, radiation exposure pattern, and clinical presentation. Acute radiation effects are commonly subtle, difficult to assess in patients, and clinically less relevant. Acute radiation effects must be suspected and investigated in patients with cardiovascular complaints early after radiotherapy. The late manifestations of RIHD usually become clinically overt several years after radiation. The symptoms and signs of RIHD are, for the most part, indistinguishable from those encountered in patients with heart disease due to other aetiologies. Table 3 gives a summary of the pathophysiological manifestations of RIHD for different radiosensitive structures within the heart.
| Acute | Long-term |
|---|---|
| Pericarditis | Pericarditis |
| • Acute exudative pericarditis is rare and often occurs during radiotherapy as a reaction to necrosis/inflammation of a tumour located next to the heart. | • Delayed chronic pericarditis appears several weeks to years after radiotherapy. In this type, extensive fibrous thickening, adhesions, chronic constriction, and chronic pericardial effusion can be observed. It is observed in up to 20% of patients within 2 years following irradiation. |
| • Delayed acute pericarditis occurs within weeks after radiotherapy and can be revealed by either an asymptomatic pericardial effusion or a symptomatic pericarditis. Cardiac tamponade is rare. Spontaneous clearance of this effusion may take up to 2 years. | • Constrictive pericarditis can be observed in 4–20% of patients and appears to be dose-dependent and related to the presence of pericardial effusion in the delayed acute phase. |
| Cardiomyopathy | Cardiomyopathy |
| • Acute myocarditis related to radiation-induced inflammation with transient repolarization abnormalities and mild myocardial dysfunction. | • Diffuse myocardial fibrosis (often after a >30-Gy radiation dose) with relevant systolic and diastolic dysfunction, conduction disturbance, and autonomic dysfunction. |
| • Restrictive cardiomyopathy represents an advanced stage of myocardial damage due to fibrosis with severe diastolic dysfunction and signs and symptoms of heart failure | |
| Valve disease | Valve disease |
| • No immediate apparent effects. | • Valve apparatus and leaflet thickening, fibrosis, shortening, and calcification predominant on left-sided valves (related to pressure difference between the left and right side of the heart). |
| • Valve regurgitation more commonly encountered than stenosis. | |
| • Stenotic lesions more commonly involving the aortic valve. | |
| • Reported incidence of clinically significant valve disease: 1% at 10 years; 5% at 15 years; 6% at 20 years after radiation exposure. | |
| • Valve disease incidence increases significantly after >20 years following irradiation: mild AR up to 45%, ≥moderate AR up to 15%, AS up to 16%, mild MR up to 48%, mild PR up to 12%. | |
| Coronary artery disease | Coronary artery disease |
| • No immediate apparent effects. (Perfusion defects can be seen in 47% of patients 6 months after radiotherapy and may be accompanied by wall-motion abnormalities and chest pain. Their long-term prognosis and significance are unknown.) | • Accelerated CAD appearing in the young age. • Concomitant atherosclerotic risk factors further enhance the development of CAD. • Latent until at least 10 years after exposure. (Patients younger than 50 years tend to develop CAD in the first decade after treatment, while older patients have longer latency periods.) |
| • Coronary ostia and proximal segments are typically involved. | |
| • CAD doubles the risk of death; relative risk of death from fatal myocardial infarction varies from 2.2 to 8.8. | |
| Carotid artery disease | Carotid artery disease |
| • No immediate apparent effects. | • Radiotherapy-induced lesions are more extensive, involving longer segments and atypical areas of carotid segments. |
| • Estimated incidence (including sub-clavian artery stenosis) about 7.4% in Hodgkin’s lymphoma. | |
| Other vascular disease | Other vascular disease |
| • No immediate apparent effects. | • Calcification of the ascending aorta and aortic arch (porcelain aorta). |
| • Lesions of any other vascular segments present within the radiation field. |
Role of Imaging in Assessing ‘RIHD’
In oncological patients, cardiac imaging is classically dictated either by the symptomatic status or by the presence of suggestive physical examination findings. Echocardiography takes a central role in evaluating the morphology and function of the heart and represents the first imaging modality in the majority of cases. Other imaging modalities, including cardiac computed tomography (CT), cardiac magnetic resonance (CMR), and nuclear cardiology, are used to confirm and evaluate the extent of RIHD. Although their use is often complementary, their clinical utility depends on the type of pathological features. For instance, the role of nuclear cardiology for assessing pericardial structures, myocardial fibrosis, or valvular heart disease associated with RIHD is limited by its suboptimal spatial resolution. Conversely, the sensitivity of cardiac CT to detect localized pericardial effusion and pericardial thickening and the accuracy of CMR in characterization of myocardial oedema, inflammation, and fibrosis are superior to echocardiography.
Specific Technical Considerations
Echocardiography
Detection of any cardiac structure abnormality, measurement of left ventricular (LV) performance, and evaluation of valvular disease severity are critical components of the assessment and management of RIHD. Several echocardiographic approaches (M-mode, Doppler, two-/three-dimensional (2D/3D) transthoracic or transoesophageal, contrast, or stress echocardiography) can be used according to the clinical indications. Unless 3D echocardiography is used, the 2D biplane disk summation method (biplane Simspon’s) is recommended for the estimation of LV volumes and ejection fraction. Contrary to 2D, 3D echocardiography makes no assumptions about the LV shape and avoids foreshortened views resulting in a better accuracy regarding the assessment of LV mass and volumes. A common limitation of 2D/3D is the suboptimal visualization of the endocardial border. This happens particularly in patients with obesity, respiratory disease, thoracic deformity, or previous open-chest cardiac surgery. When more than two segments are not adequately visualized, the use of contrast agents for endocardial border definition improves inter-observer variability to a level comparable with CMR. New currently available techniques (tissue Doppler imaging and 2D speckle tracking) may yield complementary information for the assessment of LV function. Although tissue Doppler-derived velocity parameters are easier to obtain, deformation imaging (strain and strain rate) appears more sensitive to detect subtle function changes and may become a valuable clinical tool to assess myocardial function in oncology patients. 2D speckle tracking echocardiography is an accurate angle-independent modality for the quantification of strain, a measure of LV systolic function, while tissue Doppler imaging is angle dependent and its derived velocities are widely affected by tethering to adjacent segments and the overall motion of the heart. Due to its high degree of automation, 2D speckle tracking is particularly suited for repetitive follow-up examinations by different echocardiographers. The main drawback of the 2D speckle tracking approach is that the results are affected by the image quality. Further, to guarantee comparability, serial studies should be performed on the same platform and software release. For valve analysis, transthoracic Doppler echocardiography is the recommended first-line imaging, whereas transoesophageal echocardiography is advocated in the absence of contraindications when transthoracic echocardiography is non-diagnostic or when further diagnostic refinement is required. 3D echocardiography is reasonable to provide additional information in patients with complex valve lesions.
Cardiac Magnetic Resonance
CMR physics and image acquisition strategies are discussed elsewhere. Black-blood T 1 -weighted fast spin-echo CMR provides an excellent morphologic view of the heart, pericardium, great vessels, and adjacent structures. T 2 -weighted fast spin-echo imaging, using a short-tau inversion-recovery (STIR) sequence (triple inversion-recovery), depicts increased free water as areas of high signal intensity. This sequence allows the visualization of myocardial oedema in the setting of acute myocarditis, or pericardial oedema in patients with inflammatory pericarditis. More quantitative data can be obtained using T 2 -mapping techniques. Gadolinium-based paramagnetic contrast agents are routinely used in clinical CMR. Following intravenous injection, the first pass of contrast agent can be used for single-phase or time-resolved 3D MR angiography, and for myocardial perfusion imaging. The latter can be performed during infusion of a vasodilator (e.g. adenosine and dipyridamole) to visualize LV segmental perfusion abnormalities due to haemodynamically significant coronary artery stenosis. Normal myocardium is typically characterized by a rapid wash-in and wash-out. Conversely, in an abnormal myocardium, such as necrotic or fibrotic myocardium, the concentration of gadolinium increases over time owing to an increased extracellular volume distribution with decreased wash-out. These regions are typically hyper-intense (i.e. bright). With the advent of the inversion-recovery-based CMR sequences, the so-called late-/delayed- (gadolinium) enhancement (LGE) imaging technique, irreversible myocardial damage as small as 1 g, can be depicted. The pattern, location, and extent of myocardial enhancement enable the differentiation of ischaemic from non-ischaemic causes. To depict diffuse myocardial fibrosis, T 1 -mapping techniques have been recently proposed. These calculate the post-contrast T 1 relaxation time. Bright-blood cine CMR imaging, using balanced steady-state free precession (SSFP) gradient-echo sequences, provides dynamic information to quantify ventricular volumes, function, and mass, to assess regional myocardial function, and to visualize valvular heart disease. In addition, myocardial deformation patterns can be assessed by CMR tagging techniques. A final, important CMR technique is velocity-encoded or phase-contrast cine CMR. This sequence measures the degree of ‘dephasing’ caused by through-plane motion of protons. This versatile sequence can be used to measure flow velocities (and volumes) in blood vessels, to calculate severity of shunts, to quantify velocities and regurgitation through valves, and to possibly assess diastolic function. The main limitation of CMR is that it is impractical in patients with pacemakers, claustrophobia, and anxiety attacks, and may present some difficulties in children and very obese patients. Moreover, the inability to carry out repeated breath holds and the presence of arrhythmias might represent additional problems. Finally, CMR may not be available in some community hospitals and access to CMR is limited in some institutions.
Cardiac CT
Cardiac CT offers detailed cross-sectional anatomical imaging of the chest. Intravenous injection of contrast medium opacifies the cardiac cavities and vessels and allows differentiation from the surrounding tissues. By synchronizing the acquisition or reconstruction of images to the electrocardiogram (ECG), motion-free, and phase-consistent images of the heart can be obtained, which is important for robust depiction of the coronary arteries and functional analyses. Advantages of cardiac CT in comparison with other imaging modalities include high-spatial resolution, short-exam times, and high sensitivity for calcified tissues. CT is the only non-invasive technique that can reliably image the coronary arteries. Drawbacks are the need for iodine-containing contrast media, ionizing radiation, breath holding, lower heart rates, and the inter-machine variation in radiation dose. Contemporary CT systems are equipped with 64 or more detector rows, which allow imaging of the entire heart in five heart cycles or fewer. ECG synchronization is accomplished by retrograde ECG gating for spiral scanning, or prospective ECG triggering for axial scan modes. By limiting exposure to the phase of interest (generally, the motion sparse diastolic phase), the radiation dose can be reduced. Contemporary scanner technology and scan protocols for coronary imaging are associated with an average radiation dose of <5 mSv. Cardiac CT examinations that include full-cycle exposure are associated with a higher radiation dose, which represents a significant limitation, especially if follow-up is the goal of the examination.
Nuclear Cardiology
Cardiac radionuclide imaging (single-photon emission CT, SPECT and positron emission tomography, PET) encompasses a variety of techniques designed to provide valuable information in detecting the presence and extent of cardiac disease. Two radioisotopes are routinely used in SPECT perfusion imaging: 201 Tl and 99m Tc (bound to either sestamibi or tetrofosmin). Imaging can be performed at rest and during stress (exercise or pharmacological), which allows the determination of regional perfusion defects (ischaemia or infarction/scar). ECG-gated SPECT ventriculography by either myocardial perfusion or by blood pool techniques provides highly accurate, reproducible, and prognostically validated measurements of LV end-systolic volume, end-diastolic volume, and ejection fraction. Technetium-based tracers are preferred over thallium for gated acquisitions due to the higher count statistics. Limitations of the techniques relate to its radiation exposure, ability to reproduce the same position on initial and delayed (or rest) images, and the need to select the longest cardiac cycles during ECG-gated imaging to optimize the assessment of LV ejection fraction and volume indices in cases of unstable rhythm. Radiation exposure depends on the radionuclide agents, ranging between 3 and 22 mSv, but with current cadmium zinc telluride (CZT) SPECT technology these exposures can be readily reduced to the <12 mSv range. PET myocardial perfusion with 13 NH 4 or 82 Ru has attractive features as a screening tool in survivors of mediastinal irradiation. Its intrinsic higher resolution, higher count rate, and more robust attenuation correction allow for accurate quantification of myocardial blood flow. However, the availability of PET is more restricted, because the majority of PET tracers (except for 82 Ru) require an onsite cyclotron.
Of note, the current generation of new CZT SPECT gamma cameras provide superior spatial resolution compared with traditional sodium iodide SPECT systems (spatial resolution 8–10 mm) and approach effective spatial resolution of PET (spatial resolution 4–5 mm) cameras. ECG-gated myocardial perfusion SPECT imaging and equilibrium-gated radionuclide angiocardiography (ERNA) provide an quantitative assessment of LV volume indices, ejection fraction, and diastolic peak filling rate, which are all of proven value for risk stratification in patients with ischaemic, valvular, and myocardial diseases. In valvular heart disease, the inability to assess valve morphology and its severity limits the use of these techniques. Moreover, these techniques have not yet been tested in patients with known or suspected RIHD.
Imaging Findings
Pericarditis
In radiation-induced pericarditis, heart imaging is useful for evaluating the degree of pericardial thickening, the extent of pericardial calcification, the presence of constrictive physiology, the presence and quantification of a pericardial effusion, and for patient follow-up.
Echocardiography
Pericardial thickening appears as increased echogenicity of the pericardium on 2D echocardiography and as multiple parallel reflections posterior to the LV on M-mode recordings. However, the distinction between the normal and thickened pericardium may be difficult. Pericardial effusion is visualized as an echo-free space, external to the myocardial wall. Small amounts of fluid (<20 mL) can be detected with a high sensitivity. Pleural effusion and epicardial fat may be sometimes mistaken for pericardial effusion. As a rule, fluid appearing in the parasternal long-axis view anterior to the descending aorta is typically pericardial, while pleural effusion is usually localized posterior to the aorta. Fat is naturally distinguished from effusion by a higher density (brighter echoes). As for pericardial thickening, distinction between fat and pericardium may require the use of other imaging techniques.
Echocardiographic features suggestive of cardiac tamponade may occur, but are rare. They are discussed elsewhere. Characteristic echocardiographic findings of constrictive pericarditis include thickened pericardium, prominent respiratory phasic diastolic bounce of the inter-ventricular septum, restrictive diastolic filling pattern (E/A ratio of >2 and deceleration time of the mitral E-velocity of <140 ms), significant inspiratory variation of the mitral E-wave velocity (>25%), diastolic flattening of the LV posterior wall, inferior vena cava plethora, and expiratory diastolic flow reversal in the hepatic veins. Typically, tissue Doppler interrogation of the medial mitral annulus reveals a normal or increased velocity that can be higher than the lateral annulus velocity. The systolic pulmonary pressures are not significantly elevated. This condition may be differentiated from restrictive cardiomyopathy (also a complication of radiation) by the normal mitral tissue Doppler velocity and a systolic pulmonary artery pressure <50 mmHg.
Cardiac MR
In acute pericarditis, pericardial layers are typically thickened and strongly enhance following contrast administration. Pericardial enhancement reflects inflammation and correlates with elevated inflammatory markers ( Figure 2 ). The presence, location, and extent of pericardial effusion, as well as associated cardiac tamponade, can be well assessed using a combination of dark-blood and bright-blood CMR sequences, and to some extent the characterization of pericardial effusion can be achieved. The location and severity of pericardial abnormalities is well visualized using black-blood, T 1 -weighted fast spin-echo CMR, though it should be emphasized that pericardial calcifications might be missed. CMR allows the detection of indirect signs of constrictive pericarditis, such as unilateral or bilateral atrial enlargement, conical deformity of the ventricles, dilatation of caval/hepatic veins, pleural effusion, and ascites. End-stage chronic forms of constrictive pericarditis may not demonstrate pericardial LGE on CMR, whereas pericardial enhancement is suggestive of residual inflammation. Although pericardial thickness is traditionally considered an important criterion for constrictive pericarditis, it is important to note that the range of pericardial thicknesses is highly variable (1–17 mm, mean of 4 mm) with up to 20% of patients showing a normal thickness (<2 mm). Two recent studies showed that pericardial thickness in end-stage constrictive pericarditis was significantly lower than in those with persistent chronic inflammation and no signs of constriction. Real-time cine imaging is of great value to assess the impact of respiration on the inter-ventricular septal shape and motion, allowing to easily depict pathological (increased) ventricular coupling. Furthermore, tagging the sequence detects the presence of pericardial adhesion. Recently, real-time phase-contrast imaging has been proposed to assess the effects of respiration on cardiac filling.
Cardiac CT
The pericardial cavity and membranes are between the epicardial and pericardial fat and can be recognized on cardiac CT images even without injection of contrast media. The normal pericardium is clearly visible near the right ventricle (RV) and generally does not measure >3 mm in thickness ( Figure 3 A). Thickening of the pericardium ( Figure 3 B) may be difficult to distinguish from small pericardial effusions. Inflamed pericardial membranes may have increased attenuation, compared with the pericardial fluid ( Figure 3 C). Pericardial calcifications ( Figure 3 D), as well as larger pericardial effusion, are readily identified, and also, on non-enhanced CT images. Based on the measured attenuation, serous transudates (0–25 HU) and non-serous exudates (>25 HU) may be differentiated. Cardiac tamponade may be suggested by large fluid accumulation, compression of the cardiac cavities, and right-sided venous congestion. Constrictive pericarditis is not an anatomical diagnosis, although certain CT characteristics are associated, such as pericardial calcification, pericardial thickening (>4 mm), narrowing or tubular deformation of the RV, as well as manifestations of venous congestion. Pericardial abnormalities may be regional ( Figure 4 ).
