Cardiac Complications of Thoracic Irradiation

Pericardial disease

Acute radiation pericarditis might be one of the earliest cardiac complications of radiotherapy and was historically the most common cardiovascular manifestation of irradiation therapy, typically manifesting a few weeks after treatment. Modern methods of implementation, including lower doses (including equally weighted anterior and posterior beams) and shielding methods (like subcarinal blocking) have decreased the incidence from 20% to 2.5% (43,44), with the incidence being proportional to dose and treatment volume (17). Pericarditis is usually self-limiting, however, 10% to 20% of patients develop a chronic or constrictive pericarditis 5 to 10 years after treatment (16). Associated pericardial effusions are characterized by fibrous adhesions, high protein count, and an increase in serum inflammatory markers. Chronic pericarditis might occur even if the patient did not experience an acute event. Fibrotic fusion and thickening of the visceral and parietal layers in the presence of a tense effusion in the free pericardial space is referred to as effusive-constrictive pericarditis and has been well-described in radiation-treated cancer survivors (45,46).

CAD

Pathologically, vascular radiation injury is accompanied by an increase in capillary wall permeability, the generation of reactive oxygen species, and activation of an inflammatory cascade (38). This subsequently leads to intimal proliferation with collagen deposition and fibrosis (16). As such, initial endothelial dysfunction is believed to be a key factor in the development of cardiac sequelae (39). In the case of CAD, the stenoses might be due to intimal proliferation, damage to the intima, or atherosclerotic lesions (9). This initiates a cascade of events that is typical of atherosclerosis, including replacement of damaged cells with myofibroblasts along with platelet deposition. In vitro studies have also revealed the pro-thrombotic effect of radiation associated with higher release of von Willebrand factor which can provoke and contribute to coronary thrombosis (47,48). Ultimately, the mechanism involved in plaque formation is thought to mirror spontaneous atherosclerosis; however, plaques in irradiated patients have been found to be more fibrous with a decreased lipid component (17,49–51). Ostial stenoses are typical for radiation-induced CAD. Macroscopically, there is a significantly higher incidence of severe left main disease, followed by ostial right coronary artery and left anterior descending artery stenoses correlating with the anatomic location of radiation treatment (9). In general, the lesions are consistently longer, more proximal, smoother, concentric, and tubular (9). The incidence of radiation-induced CAD is increased by the standard cardiac risk factors. Hull et al. (52) studied 415 patients post-RT for Hodgkin’s lymphoma and found that 10.4% of patients developed significant CAD (defined as having history of myocardial infarction/CABG/>75% luminal stenosis on angiography or autopsy) at a median of 9 years of follow-up.

Myocardium

The lack of myocyte cell division makes the myocardium itself relatively radiation resistant. However, diffuse interstitial fibrosis occurs after relatively low doses of radiation through microvascular insufficiency and ischemia, involving damage to the capillary endothelial cells (53,54). The result is an alteration in the compliance of the myocardium, leading to both systolic and diastolic dysfunction. Most patients with myocardial involvement have some degree of interstitial fibrosis (55), typically developing at radiation doses above 30 Gy. It has been shown that diastolic dysfunction is seven-fold more common in individuals treated with thoracic radiotherapy population compared with community control subjects, and the presence of diastolic dysfunction is associated with stress-induced ischemia, a worse prognosis, and event free survival (56).

Valvular dysfunction

Radiotherapy might directly damage cardiac valves and surrounding myocardium, leading to fibrotic thickening, valvular retraction, and late calcification (5). The incidence of valvular dysfunction has been reported to increase during the second decade after mediastinal radiotherapy for HD (5). Wethal et al. (5) found that, of 116 patients observed approximately 10 years after treatment, 36 (31%) demonstrated moderate valvular regurgitation in 40 valves, primarily either the aortic or mitral valve, whereas none had evidence of stenosis. Interestingly, in the same cohort 39% of patients still available for follow-up 12 years later had developed aortic stenosis, including some of a moderate to severe degree (35%) (5). The valves were found to be thicker with reduced leaflet motion and calcification (5). These results suggest that valve retraction is the predominant early change that causes regurgitation and that it might take in excess of 20 years for the valves to become significantly thickened, calcified, and stenotic. It is hypothesized that a higher pressure system on the left side of the heart compared with the right accounts for why the mitral and aortic valves are affected more than the pulmonary and tricuspid. Multiple studies have supported the higher incidence of aortic and mitral valve disease (4,5,57). Consistent with these observations, another study found that, of asymptomatic patients previously treated with at least 35 Gy of radiation, 6% had clinically significant dysfunction, and 26% had >grade II aortic regurgitation. This equated to a 34-fold increased risk compared with the Framingham population (4). Also, 26% demonstrated marked calcification of the aortomitral curtain (4).

Conduction system

Abnormalities along the entire conduction system have been described in the setting of radiotherapy, including varying degrees of atrioventricular block, sick sinus syndrome, prolonged QTc, supraventricular arrhythmias, and ventricular tachycardia (53,58). Right bundle branch block is more commonly observed than left bundle branch block post-RT, likely due to higher radiation exposure of the right ventricle due to its anterior location. Ventricular ectopy might occur in up to 50% of patients exposed to mediastinal irradiation, often secondary to ventricular fibrosis (53,58). Crestanello et al. (59) found a high post-operative prevalence of conduction disease after valvular surgery in this patient cohort, with 27% of patients requiring permanent pacemaker placement. Dysfunction of the autonomic nervous system, leading to persistent tachycardia, loss of circadian heart rhythm, and respiratory phasic heart rate variability has also been described after irradiation (60). The latter clinical picture is similar to heart transplant patients who have a denervated heart (60). Persisting tachycardia might further increase the patient risk of developing a tachycardia-mediated cardiomyopathy.

At 5 years, then annually

B-type natriuretic peptide (BNP) and troponins might be useful serum biomarkers that might be able to act as early and perhaps pre-clinical markers of myocardial damage. This has been more widely researched in chemotherapy patients; however, Nellessen et al. (54) also evaluated plasma levels of troponin and BNP up to 6 weeks post-RT. Plasma levels of both biomarkers increased over the study period, although without measured impairment in left ventricular (LV) function. This finding supports the potential use of biomarkers in screening post-treatment to potentially identify high-risk patients. In addition to BNP and troponins, blood tests such as fasting serum glucose and lipid profiles should also be reviewed annually. Annual electrocardiography is also a simple investigation that might alert the treating physician to the development of progressive conduction abnormalities. A 24-h ambulatory Holter monitor might also be warranted, depending on the clinical scenario. In one study by Larsen et al. (58), 7% of childhood cancers survivors (mean age 15 years) were found to have ventricular tachycardia on 24-h Holter monitor. Given the sporadic and variable frequency of ventricular arrhythmias, this study might have underestimated the prevalence of ventricular arrhythmias in survivors of childhood cancer.

As outlined in the preceding text, the key late cardiovascular sequelae of thoracic irradiation are the result of coronary, myocardial, and valvular pathology. Most of these complications can be evaluated for using combined rest and stress echocardiography. This approach enables the clinician to assess LV systolic function, diastolic function, valvular function, pericardial integrity, and regional wall abnormalities, both at rest and with exercise. When possible, stress echocardiography should be chosen over a stress thallium/methoxyisobutylisonitrile for the radiation-free and functional advantages, particularly relevant in this patient cohort.

Vigilant screening for athletic participation might also be an important consideration in the pediatric cancer group. In the setting of documented arrhythmias and to further risk-stratify patients, cardiac magnetic resonance imaging (MRI) is an extremely valuable modality to further assess for regional scar, potentially serving as a substrate for ventricular arrhythmias. In patients with nonischemic cardiomyopathy, late gadolinium enhancement on cardiac MRI has been shown to strongly predict adverse cardiac outcomes such as hospital stays for heart failure, appropriate ICD firing, and death (68). There is a growing body of published data supporting a potential role for cardiac MRI and late gadolinium enhancement in the risk stratification and prognostication of patients with cardiomyopathy (68–70), which might have further value in determining the need for device therapy.

Furthermore, cardiac MRI is a highly sensitive modality that potentially provides clinical information beyond the capability of echocardiography, particularly with regard to tissue characterization. Cardiac MRI is particularly useful to assess for regional and, more recently, diffuse myocardial fibrosis with delayed gadolinium enhancement and T1 mapping respectively, with increasing evidence that T1 mapping correlates with diffuse fibrosis and diastolic dysfunction (71,72). Thereby, cardiac MRI might be particularly valuable to diagnose asymptomatic diastolic dysfunction and possibly guide the timing of initiation of anti-fibrotic agents like angiotensin-converting enzyme inhibitors (ACEI) and aldosterone inhibitors. Cardiac MRI might also be useful to evaluate possible diastolic dysfunction in the patient population that cannot undergo tissue Doppler imaging, limited by a heavily calcified cardiac skeleton (Fig. 1), mitral stenosis, or mitral valve replacement.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Cardiac Computed Tomography Images

Cardiac computed tomography images from a 56-year-old man, 27 years post-mantle irradiation for Hodgkin’s disease. Images demonstrate calcified aortomitral curtain and aorta (A), and mitral valve annulus (B).

Coronary computerized topography (CCT) can noninvasively detect and quantify obstructive coronary artery stenoses. Novel prospective gating techniques have significantly reduced radiation dose, making cardiac computed tomography an increasingly attractive and complementary investigative tool for cancer survivors. In the general population, the coronary calcium score is particularly useful (with very little radiation exposure) to predict coronary artery stenosis and cardiac events and adds prognostic value to the Framingham risk score. Yeboah et al. (73) have shown that the coronary calcium score provides superior discrimination and risk reclassification compared with other risk markers in intermediate-risk individuals. However, its utility in the late effects patient cohort is still to be determined. Rademaker et al. (74) retrospectively evaluated the CCT findings of 9 HD survivors, treated with radiotherapy 12 to 35 years ago. This pilot study showed that 8 of the 9 patients were found to have at least mild atherosclerosis, 66% had calcium scores within the 90th percentile for their age, and 2 patients proceeded to coronary angiography (1 requiring PCI, and 1 requiring CABG) (74). Patients with the lowest radiation dose and shortest interval since radiotherapy had minimal or no findings of CAD. At this early stage, however, we do not recommend that baseline screening stress echocardiography should be substituted for CCT. However, this is certainly a potential focus for further research.

Acute and chronic pericarditis

As outlined in the preceding text, acute pericarditis is relatively uncommon with the use of more contemporary thoracic radiotherapy strategies. Similar to other forms of acute pericarditis, patients might present with fever, pleuritic chest pain, dyspnea, and tachycardia. Typical clinical, electrocardiographic, and biochemical features might be present. Transthoracic echocardiography should be performed to exclude tamponade. Cardiac MRI might also be useful to confirm the diagnosis and establish the degree of pericardial thickening along with an assessment for myocardial involvement. Treatment includes the use of nonsteroidal anti-inflammatory agents. Pericardiocentesis might be indicated if there is an associated large pericardial effusion and if the patient is significantly hemodynamically compromised. Pericardectomy might be required for constrictive pericarditis. It is important to exclude other etiologies in the setting of recurrent pericardial effusion such as infection, tumor invasion or recurrence, and hypothyroidism, which might be seen after mantle irradiation. Constrictive pericarditis has been described as a marker for greater radiation injury to the heart, associated with diastolic dysfunction and high mortality (58). Early pericardial stripping might be supported at the time of planned cardiac surgery to improve postoperative cardiac physiology (62).

CAD

In the setting of flow limiting disease of the major epicardial coronary arteries, most lesions have been found to be amenable to treatment with either CABG or PCI dependent upon the specific anatomy. Some early studies have suggested that percutaneous angioplasty seems to have a higher rate of restenosis (75); however, there are no data to date incorporating outcomes of newer, second-generation drug-eluting stents. A study conducted by Schomig et al. (76) found significant bare-metal stent restenosis at 6-month follow-up (86%) in patients with radiation-induced CAD (p < 0.001) compared with control subjects (76). The group subjects were also found to have significantly narrower coronary lumen diameters (1.0 ± 0.6 mm vs. 1.9 ± 0.9 mm, p < 0.001). Caution with appropriate patient selection must be exercised for PCI, given that certain post-irradiation complications might preclude surgical candidature for failed PCI. Of note, the internal mammary arteries might be chronically scarred or more friable, making them unsuited for use in bypass grafting. In conjunction, the presence of damaged target vessels might add increasing complexity and risk in patients undergoing CABG (77,78). In one study, the internal mammary artery was an appropriate conduit less than one-half the time, routinely harvested by the same authors in their standard practice (62). Routine angiography of the internal mammary arteries might be useful to guide conduit utility and subsequent treatment options.

To date there are few studies, mostly with small patient numbers, on cardiac surgical outcomes and the post-operative course in this specific patient population. Siregar et al. (79) showed that previously treated patients for Hodgkin’s lymphoma had relatively good early post-operative outcome with a 1-year survival of 86%; however, the long-term survival at 4 years was only 46%. The most common post-operative complication after cardiac surgery (operations included: CABGs, valvular repair, valvular replacement) was atrial fibrillation, however, with a slightly higher incidence than described by large databases in the general population [33% vs. 20%] (79,80). In the postoperative setting these patients are also at increased risk of pulmonary complications, particularly pleural effusions, pneumonia, and respiratory insufficiency, due in part to the presence of pre-existing radiation-induced lung disease and smoking (62,79). Thus pre-operative smoking cessation assumes even greater importance in this patient cohort. The presence of diastolic dysfunction is another challenging issue in this patient group, with high left atrial pressures contributing to pulmonary congestion, thereby increasing post-operative morbidity. The largest study performed by Chang et al. (62) included 70 patients with previous HD. These authors found that extensive exposure of the heart to radiation increased perioperative morbidity and decreased both short- and long-term survival. The cardiac risk factors included higher central venous pressure, higher left atrial pressure, and lower LV mass (62). Increased central venous pressure as a risk factor also suggests that concomitant right ventricular dysfunction is a complicating factor, and in particular the anterior location of the right ventricle makes it vulnerable to dysfunction post-radiotherapy.

As such, it is clear that, for patients with a prior history of chest irradiation in whom cardiothoracic surgery is contemplated, full consideration should be given to the potential impact of alterations in the structure and function of the skin, mediastinal tissues, lungs, intra-thoracic vessels, and the heart per se require careful consideration (59,81). On balance, an appropriate choice of surgical versus percutaneous interventions balancing the relevant risk/benefit ratio of each approach can then be offered to the patient in the setting of the specific clinical context.

Conduction disturbance

A permanent pacemaker is indicated in the setting of high-degree block or symptomatic sick sinus syndrome, following the routine American College of Cardiology/American Heart Association guidelines. Additionally, after a life-threatening arrhythmia or aborted sudden cardiac death, an ICD is indicated for secondary prevention, upon exclusion of ischemia. For some patients, an electrophysiological study and ventricular tachycardia ablation might be useful.

Valvular dysfunction

As outlined in the preceding text, aortic and mitral valve regurgitation and stenosis are well-recognized complications of thoracic irradiation. Crestanello et al. (59) were one of the first groups to investigate whether conventional reparative techniques could be applied to irradiation related valvular disease. These investigators found that the durability of valve repair in long-term survivors was limited, because severe dysfunction developed in one-third of the patients and with reoperation being required in 16% (59). Accordingly, valve replacement might be preferable to valve repair when considering long-term cardiovascular outcomes. In some cases of severe aortic stenosis with significant extracardiac late sequelae of radiotherapy, transcatheter aortic valve implantation might be the best treatment option.

Heart failure

Although no specific therapies for the treatment of radiation-induced heart failure have been identified, therapy of systolic heart failure should follow American College of Cardiology/American Heart Association guidelines, with slow upward titration of ACEI, beta blockade, and aldosterone inhibitors, particularly in patients with reduced LV systolic function. In patients with diastolic dysfunction, the principles of treatment broadly mirror those of patients with heart failure with preserved ejection fraction. Although no treatment has been shown to convincingly improve morbidity and mortality in patients with heart failure with preserved ejection fraction (82), diuretic agents and, more recently, exercise training might be introduced to improve symptoms (83). Hypertension, sinus rhythm, and myocardial ischemia must also be optimized. The same indications for insertion of ICD cardiac resynchronization therapy should apply to patients with radiation-induced heart failure and might additionally prolong survival. If significant subcutaneous skin involvement is observed as a late effect of thoracic irradiation, a sub-pectoral approach for device implantation might be necessary to avoid poor wound healing and infection. In the instance of refractory heart failure, temporary inotropic support for periods of decompensation might be needed. In patients with refractory heart failure the use of LV assist devices and heart transplantation might be considered; however, the decision with regard to the appropriateness of such therapies clearly requires multidisciplinary input, particularly from oncologists.

Heart transplantation

For a small patient subgroup with biventricular dysfunction, calcified cardiac skeleton (Fig. 1), previous open-heart surgery, and restrictive or constrictive hemodynamic status, heart transplantation might be the only possible and reasonable treatment option. The largest reported study to date was conducted by Uriel et al. (84). This was a retrospective study of 9 patients with radiotherapy induced cardiac disease who underwent orthotropic heart transplantation. Of these patients, there were 3 in-hospital perioperative deaths, 3 second cancers, and 3 life-threatening infections. Five of the 9 patients (56%) were still alive at a mean follow-up of 10 years post-transplantation (however, 2 with a second cancer) (84). Orthotropic heart transplantation also raises the question of potential cancer recurrence in an immunosuppressed population, with a paucity of data in this area to date.

Future research

Large prospective studies still need to be conducted to explore the benefit of prophylactic treatment strategies in asymptomatic patients. For example, the effect of statins, ACEI, or aldosterone inhibitors in asymptomatic radiation-treated cancer survivors to slow/halt accelerated atherosclerosis, myocardial dysfunction, or myocardial fibrosis is not yet known. The clinical benefit of such pharmacological agents in those both with and without additional cardiac risk factors is of particular clinical interest. The impact of newer-generation drug-eluting stents on in-stent restenosis and particularly stent thrombosis in this patient cohort still needs to be clarified. And clearly, the long-term effects of contemporary and refined radiotherapy techniques mandate further analysis with time.