CHAPTER 101: THE DIAGNOSIS AND MANAGEMENT OF CARDIOVASCULAR DISEASE IN PATIENTS WITH CANCER

With the advent of more effective cancer treatments and the increasing likelihood of an earlier cancer diagnosis, patients with many forms of cancer can expect to either be cured of their disease or have their disease stabilized by maintenance therapy. Although the overall rate of cancer incidence has declined since the early 2000s,¹ cancers necessitating aggressive chemotherapy, including melanoma, non-Hodgkin lymphoma, leukemia, and those of the pancreas and esophagus have been on the rise. Accompanying this trend, the length of cancer survival has increased for all cancers combined. The 5-year survival rate for all cancers combined improved from 66.7% in 2003 to 68% in 2009.¹ This implies that cancer survivors now live longer, allowing the manifestation of potential cardiac side effects of chemotherapeutic agents as well as the age-related increase in the risk of cardiovascular disease.

For individuals free of cardiovascular disease at 50 years of age, more than half of men and nearly 40% of women will develop cardiovascular disease during their remaining lifetime. Other than the extended survival of cancer patients and the aging population, there has been an increase in the recognition of chemotherapy-induced cardiotoxicity, adding to the linkage between patients with cancer and their risk for cardiovascular diseases. In addition, some patients with cancer may be at a higher risk for cardiovascular complications as compared with the general population.² Multiple classes of potentially cardiotoxic anticancer agents are currently being developed. Indeed, the combination of increased use of chemotherapy, overall increased survival, and development of newer agents have led to the emergence of chemotherapy-induced cardiotoxicity as a growing public health issue.

Moreover, one of the most significant late toxicities in survivors of childhood cancers is late-onset cardiotoxicity, largely as a result of anthracycline and chest-directed radiation exposure. Survivors also have an increased prevalence of traditional cardiovascular risk factors as they age, which potentiates the risk for major adverse cardiac events. Prevention of cardiotoxicity is essential in this population. Minimizing anthracycline dose exposure in pediatric cancer patients is a primary method of cardioprotection. Dexrazoxane and enalapril have also been studied as primary (pre-exposure) and secondary (post-exposure) cardioprotectant agents, respectively. Additionally, the Children’s Oncology Group has published exposure-driven, risk-based screening guidelines for long-term follow-up, which may be a cost-effective way to identify subclinical cardiac disease before progression to clinical presentation. Ongoing research is needed to determine the most effective diagnostic modality for screening (eg, echocardiography) and the most effective intervention strategies to improve long-term outcomes.³

Cardiologists dedicated to the care of patients with chemotherapy-induced cardiotoxicity and with concomitant cardiovascular problems in cancer patients are a part of an emerging subspeciality called “Cardio-Oncology” or “Onco-Cardiology.” Many cancer centers and tertiary care hospitals in the United States are now establishing dedicated Cardio-Oncology clinics where cardiovascular specialists play a dedicated role in managing heart disease in cancer patients. The reason for this evolving need for specialized cardiac care in cancer patients is related to the acute and chronic cardiovascular complications of cancer therapy, emerging data that early detection of cardiotoxicity improves patient outcomes, and finally the observation that cardiovascular interventions have to be balanced with cancer therapeutics to provide the patient with the best possible oncological and cardiovascular outcomes.

Prior to initiation of any cancer therapy, cardiologists should make every effort to optimize the management of any underlying cardiovascular conditions, including coronary artery disease (CAD), hypertension, and arrhythmias. Furthermore, cardiologists should recognize and promptly manage the cardiovascular complications from cancer treatment. Cardiovascular toxicity of chemotherapy and antibody-based therapy encompasses diverse clinical manifestations, including and not limited to heart failure, cardiac ischemia, blood pressure changes, and arrhythmias (Table 101–1). Early recognition of this spectrum of cardiotoxicities, such as with new imaging techniques to detect subclinical left ventricular (LV) dysfunction prior to the development of overt heart failure, will allow for early cardiovascular intervention to improve outcomes.

The Common Terminology Criteria for Adverse Events (CTCAE), developed by the National Cancer Institute,⁴ recognizes a broad array of cardiovascular events and subclinical laboratory and imaging-based functional changes that encompass under the broad definition of cardiotoxity.

Acute or subacute cardiotoxicity can include acute coronary syndrome (ACS), myocarditis, pericarditis, and/or supraventricular or ventricular arrhythmias. Such cardiotoxicity may manifest as electrocardiographic abnormalities (eg, QT prolongation) or fluctuations in blood pressure that may occur at the initiation of chemotherapy or up to 2 weeks following its termination.

Chronic cardiotoxicity includes the entire spectrum from asymptomatic systolic or diastolic LV dysfunction to severe symptomatic heart failure, accelerated CAD, chronic pericardial syndromes, and valvulopathy.

Several risk factors contribute to the development of cardiotoxicity in patients being treated for cancer. These include nonchemotherapy-related risk factors such as age, preexisting hypertension, CAD, and past radiation therapy as in the case of cardiotoxicity resulting from drugs such as anthracyclines (Table 101–2) and trastuzumab.^5,6,7,8,9

The chemotherapy-related risk factors include the dose of the drug administered during each chemotherapy cycle, the cumulative dose, the schedule of delivery, the route of administration, concomitant use of other cardiotoxic drugs, and the sequence of administration of these drugs. Anthracyclines potentially can induce cardiac damage with the first administration, but more typically affect cardiac myocytes at higher cumulative doses with a much higher incidence of heart failure at doses exceeding 450 mg/m².⁵ Cyclophosphamide causes cardiac dysfunction at high doses exceeding 180 mg/kg and when given concomitantly with other cardiotoxic drugs. Similarly, ifosfamide can induce low-grade arrhythmias at doses of 1.2 to 2 g/m²/d for 5 days, but may result in heart failure when administered at a higher dose of 10 to 18 g/m²/d for 5 days.

For some agents, the cardiac side effects depend on the schedule of administration. Interleukin (IL)-2, a T-cell growth factor, can cause body weight gain when given low dose as a continuous infusion at a rate of 9 × 10⁶ IU/m²/d and hypotension when given as a bolus at a dose of 600,000 IU/kg every 8 hours. Likewise, the administration schedule of anthracyclines and cyclophosphamide seems to play a role in the development of cardiac side effects. Administering anthracyclines by continuous infusion over 24 to 96 hours rather than by rapid intravenous (IV) infusion probably reduces the risk of cardiotoxicity of these drugs. Similarly, parenteral, though not oral, busulfan can result in tachyarrhythmias, hypertension, or hypotension, as well as LV systolic dysfunction.

Changing the sequence in which drugs are administered can also influence the risk or severity of cardiotoxicity. For example, the combination of IL-2 and interferon given simultaneously produces pronounced hypotension, but interferon treatment alone for 2 weeks followed by IL-2 treatment has less effect on blood pressure.

Among the various cardiac complications of cancer therapy, the most prominent one is LV systolic dysfunction as a direct result of myocardial damage. A large number of cancer therapeutics can cause direct myocardial damage (see Table 101–1). For example, trastuzumab (Herceptin), an antibody directed against the HER2-neu receptor in the treatment of selected patients with breast cancer, has been associated with an increased incidence of LV systolic dysfunction. Not all forms of cardiac dysfunction related to anticancer treatments are identical. LV dysfunction related to cancer therapy is currently defined by a decrease in cardiac function and is categorized into irreversible injury (type I) or reversible dysfunction (type II).^10,11

Among the anthracyclines, doxorubicin, daunorubicin, epirubicin, and idarubicin are approved by the US Food and Drug Administration (FDA) for the treatment of patients with a variety of hematologic malignancies and solid tumors. All anthracyclines are associated with both early and late toxicity. Anthracycline-induced cardiotoxicity has been classified as early or late onset using a cutoff of 1 year after anthracycline treatment. Cumulative incidences of cardiac events typically peaked at 1 year after anthracycline treatment. Early toxicity may be manifested as a myopericarditis with nonspecific electrocardiographic ST-segment and T-wave abnormalities; arrhythmias may be part of the clinical presentation. Late anthracycline cardiotoxicity is cumulative dose related, with heart failure incidences of 5%, 16%, and 26% estimated for cumulative doxorubicin doses of 400, 500, and 550 mg/m², respectively. Although the mechanism of such LV dysfunction was purported to be myocardial injury secondary to free radical formation, antioxidants have failed to prevent anthracycline-induced cardiotoxicity. Recent studies showed that anthracycline targets topoisomerase 2b in the cardiomyocytes to induce breaks in double-stranded DNA and changes in the transcription of antioxidative and mitochondrial electron transport chain genes, leading to increase in reactive oxygen species and mitochondrial dysfunction.¹⁶ Dexrazoxane, the only FDA-approved drug for prevention of anthracycline cardiotoxicity, is a potent topoisomerase 2b inhibitor (Fig. 101–1).¹²

The mechanism of dexrazoxane cardioprotection was originally attributed to the iron chelation properties of its EDTA-like hydrolysis product, leading to decreased levels of hydroxyl free radicals. This notion was challenged by subsequent studies documenting lack of protection against doxorubicin-induced cardiomyopathy provided by deferasirox (another iron chelator) and several other free radical scavengers.¹³ More recently, it was proposed that dexrazoxane-induced cardioprotection is related to its ability to antagonize the DNA damage caused by doxorubicin by means of interference with topoisomerase II (Top2). Mammalian cells express two types of Top2: Top2α and Top2β, both of which are targeted by doxorubicin. The efficacy of doxorubicin as a chemotherapeutic agent seems to result from the inhibition of Top2α, which is only expressed in proliferating and tumor cells. Contrariwise, Top2β is present in all cells, including postmitotic cells and the adult heart, in two possible states, free or DNA-bound. Dexrazoxane can bind to both free Top2β and DNA-bound Top2β. Binding of dexrazoxane to free Top2β stabilizes its conformation, preventing its interaction with chromosomal DNA. Likewise, binding of dexrazoxane to DNA-bound Top2β stabilizes the closed-clamp conformation of ATP-bound Top2β, which triggers proteasomal degradation of Top2β, finally resulting in Top2β downregulation.¹⁴ Because the dexrazoxane binding sites are the same for the two Top2 isozymes, dexrazoxane has been expected to possibly reduce to some extent the antitumor activity of doxorubicin. However, dexrazoxane does not degrade Top 2α and in some studies may even synergize with doxorubicin in cancer cell killing.

Although the incidence of cardiomyopathy increases significantly in patients receiving cumulative doses of doxorubicin that exceed 450 mg/m², cardiomyopathy can still occur at lower cumulative doses.⁵ The mortality in patients with advanced stages of heart failure secondary to anthracycline cardiotoxicity is as high as 30% to 60%. The prognosis can be greatly altered if cardiac dysfunction is recognized early, the offending agent is promptly discontinued, and optimal treatment is instituted with a regimen including β-blockers, angiotensin-converting enzyme (ACE) inhibitors, or angiotensin receptor blockers (ARBs). Anthracyclines cause a unique pattern of histologic cardiac changes, including the vacuolization of myocardial cells, myofibrillar disarray and loss, and necrosis.⁶ A direct relationship between biopsy grade of histopathologic change, the cumulative dose of the drug, and the clinical symptoms has been shown. Histopathologic and pathogenic information regarding the cardiotoxicity of other anticancer drugs is sparse.

On the other hand, trastuzumab-induced cardiac dysfunction seems to be different from anthracycline-induced cardiac dysfunction, leading to irreversible cardiac damage (type II). It has been shown that the HER2-HER4-neuregulin system in the heart may attenuate the response of cardiomyocytes to oxidative stress and is temporarily upregulated in response to cardiomyocyte stress. Inhibition of HER2 causes mitochondrial dysfunction in cardiomyocytes. Therefore, trastuzumab-induced cardiotoxicity may result from direct binding of trastuzumab to the HER2 receptor expressed on the myocardium and increased oxidative stress (eg, because of anthracycline use); causing HER2 upregulation increases the risk for trastuzumab-associated cardiotoxicity. Therefore, trastuzumab cardiotoxicity is not dose related, but depends on whether the drug is given alone or in combination with other cardiotoxic agents.⁷ The incidence of cardiotoxicity is also increased in older patients, those with preexisting cardiac disease, and those who previously received chemotherapy and radiation therapy. The incidence of toxicity associated with trastuzumab has been much lower, however, in more recent trials with closer monitoring and when the simultaneous administration of trastuzumab and anthracyclines is avoided.⁸

These are outlined in Table 101–1, and a detailed discussion of heart failure, myocardial ischemia, changes in blood pressure, arrhythmias and QT prolongation, and thromboembolism follows below.

Heart Failure and Left Ventricular Dysfunction

Heart failure related to cancer therapy is currently defined by a decrease in cardiac function and categorized into irreversible injury (type I) or reversible dysfunction (type II). Apart from the pathophysiological definition of cardiotoxicty, there is no consensus definition of cardiotoxicity in terms of cutoff for LV ejection fraction (LVEF) at present.^15,16 Trials studying cardiotoxicity in anthracylines and trastuzumab have long relied on the definition of cardiac dysfunction as a ≥ 5% decline in LVEF to result in an LVEF < 55% with symptoms of heart failure, or a ≥ 10% decline to result in an LVEF < 55% without symptoms of heart failure.^5,7,11

The CTCAE by the National Cancer Institute defines LV dysfunction as a reduction in LVEF > 5% to < 55% with symptoms of heart failure or an asymptomatic reduction in LVEF > 10% to < 55%.⁴ A most recent document released by the American Society of Echocardiography (ASE) defines chemotherapy-related cardiac dysfunction (CTRCD) as a decrease in the LVEF of greater than 10 percentage points, to a value < 53%, which is deemed the normal reference value for two-dimensional (2D) echocardiography as per this society.¹⁵ Centers such as MD Anderson and the European Society of Medical Oncolgy (ESMO)¹⁶ define CTRCD as a reduction in LVEF of 10% or more to < 50% with or without symptoms. As a consequence, at present, a consensus definition for cardiotoxicity is still lacking.

The basic pathophysiology of heart failure is essentially the same in both cancer and noncancer patients. The neurohormonal hypothesis that forms the basis of the diagnosis of heart failure and shapes the strategies for effective treatment is applicable in both groups of patients. Heart failure and cancer progression also share pathophysiologic characteristics. It is imperative to understand the cause of heart failure in these patients, because this is one of the most important comorbidities that affects the lifespan of a cancer patient. Although the triggering event of the heart failure may be the chemotherapy, the contribution of preexisting and coexisting cardiovascular risk factors may be difficult to quantify separately.

Strategies for Diagnosing Heart Failure and Left Ventricular Dysfunction

The importance of the physical examination in confirming the presence of heart failure and LV systolic dysfunction in cancer patients cannot be overemphasized (see Chap. 69). Certain physical findings, such as a third heart sound or jugular venous distension, can be highly predictive of heart failure (see Chap. 11).

Table 101–3 summarizes the typical causes of LV systolic dysfunction and heart failure in cancer patients. All of the major contributors to heart failure need to be investigated and treated appropriately to optimally affect the course of illness (see Chap. 70).

Among basic laboratory studies, the electrocardiogram (ECG) does not discern heart failure or LV dysfunction, although it can confirm suspected abnormalities or indicate potential underlying cardiac disease, such as ischemia, conduction abnormalities or chamber enlargement. Additional laboratory testing should include lipid panel to indicate risk for vascular disease and blood glucose monitoring to screen for diabetes, both clinical predictors for heart failure. Anemia, especially common in cancer patients, has been directly correlated with outcomes in heart failure patients and should be part of a basic laboratory screen (see Chap. 70). Other specific biomarkers that are crucial in the evaluation of patients with suspected heart failure, especially those with cancer, include cardiac troponin I as well as B-type natriuretic peptide (BNP). Troponins may have value in screening for LV systolic dysfunction that develops during the course of chemotherapy.¹⁷ The role of BNP levels to detect early cardiotoxicity, LV dysfunction, or to detect volume overload has not been validated in this population. It may be falsely elevated without the presence of LV dysfunction or heart failure in cancer patients. A careful physical examination is still the most reliable means of characterizing volume status.

LVEF, most commonly determined by echocardiography, is currently the mainstay of detecting heart failure or asymptomatic LV dysfunction during chemotherapy. Monitoring protocols for patients undergoing anthracycline-based chemotherapy in the 1970s and 1980s were developed and validated based on radionuclide multigated acquisition (MUGA) scans. These, though highly reproducible, have been largely replaced by echocardiography today and, as such, the algorithms validated for MUGA may be applied to echocardiography imaging. Widespread availability, feasibility, lack of radiation exposure, and acquisition of additional cardiac imaging information (valvular, pericardial, and hemodynamic data) make echocardiography an attractive option for serial imaging. Echocardiographic strain imaging has recently emerged as a promising method for detection of early cardiotoxicity prior to LV dysfunction. Three-dimensional echo imaging has shown promising results with least inter- and intraobserver variability and reproducibility, which could be comparable to MUGA. However, there are certain circumstances in which one technique is favored over the other. In pediatric patients, echocardiography is preferred because of the lack of ionizing radiation. In obese patients, however, adequate echocardiographic windows will be difficult to attain. Furthermore, the usefulness of echocardiography is limited in other variations of thoracic anatomy, such as emphysema, tight intercostal spaces, and heavily calcified ribs. Three-dimensional echocardiography is dependent on patient ability to hold the breath and requires a longer offline analysis. MUGA is less feasible in patients with arrhythmias as a result of poor electrocardiography-based triggering, which is more often seen in patients with heart failure. Regrettably, the choice of technique used is often not determined by the question of which method is more accurate or more suitable for the individual patient, but is rather governed by the availability of local resources and personal experience. Therefore, all the described noninvasive techniques should be used as complementary.

Of particular interest in detecting early cardiotoxicity is the application of strain imaging, which is a measure of regional deformation of the myocardium. It is mainly obtained by angle-independent 2D speckle-tracking echocardiography, which can evaluate all three domains of myocardial mechanics (longitudinal, circumferential, and radial) and derive data for deformation and rate of deformation for each myocardial segment (Fig. 101–2). Two-dimensional speckle-tracking echocardiography has been used in multiple independent studies, reporting changes in cardiac (mechanical) function before a decrease in LVEF and even before changes in diastolic function after chemotherapy.¹⁵ Based on numerous studies of strain rate imaging during cancer chemotherapy, a greater than 10% change in global longitudinal strain (GLS) after completion of anthracycline-based chemotherapy relative to baseline is predictive of a future decrease in LVEF (Fig. 101–3). Conceivably, but subject to further studies, abnormal strain values before cancer therapy may signal higher baseline risk for chemotherapy-induced cardiotoxicity. Based on the above discussion, it seems appropriate to include strain imaging in monitoring algorithms for cardiotoxicity. Thereby, the ASE recommends strain rate imaging as an inherent part of a compreshensive echocardiogram for patients undergoing cancer chemotherapy. It also recommends three-dimensional (3D) echocardiography as the preferred technique for monitoring of LV function and detecting CTRCD in patients with cancer, as LVEF determination by 2D echocardiography, with or without contrast, has inherent inter- and intraobserver variability that may vary by up to 10% (Table 101–4).

Other specific cardiac testing is frequently necessary to evaluate the extent of underlying heart disease that may be present in a patient with cancer. Exercise or pharmacologic stress testing can help investigate for CAD (see Chap. 13), and CAD angiographically the extent of coronary artery disease, LV systolic dysfunction, valvular abnormalities, and other hemodynamic disturbances. A myocardial biopsy may be appropriate when it is necessary to confirm a diagnosis of amyloidosis or myocarditis. Cardiac magnetic resonance imaging (MRI) may be extremely useful in the diagnosis of constrictive pericarditis, myopericarditis, or amyloidosis.

The Principles of Therapy for Heart Failure

The principles of therapy for heart failure and LV dysfunction in cancer patients are similar to those in patients without cancer. An often overlooked management component is the education of both the patient and family members regarding early recognition of the symptoms of fluid overload and decompensation. Management of patients who develop heart failure from cardio­toxicity during or after cancer therapy should be in keeping with the American Heart Association (AHA)/American College of Cardiology (ACC) heart failure guidelines (see Chap. 71). There is no reason to suspect that standard medical therapy would not be effective in most cancer patients as well.

ACE inhibitors and β-adrenergic blocking agents remain the cornerstone of therapy and have specifically been studied in cancer patients.^18,19,20 The timing from completion of anthracycline-based chemotherapy to initiation of heart failure therapy was identified as the crucial determinant of the response rate. Early initiation of these agents is key as the recovery of LV function is dependent on early intiation of heart failure treatment. The usefulness of ARBs, as alternatives to ACE inhibitors, for the treatment of heart failure can be extrapolated to the cancer population as well. The combination of nitrates and hydralazine is of special benefit in African American patients, as well as in patients with renal insufficiency not able to receive ACE inhibitors or ARBs. The precise role of aldosterone receptor antagonists (eg, spironolactone) in the treatment of chemotherapy-induced cardiomyopathy is currently unknown, but may be considered in those with New York Heart Association (NYHA) class > 1 symptoms and an LVEF of 35% or less.

Hemodynamic device support may become necessary if medical therapy fails. Acute hemodynamic support can be temporarily lifesaving, as in those with acute myopericarditis owing to drugs such as cyclophosphamide. Alternatively, chronic LV assist device support may become a bridge to transplant or destination therapy in some resported cases as reported recently.²¹ The most severe forms of heart failure are usually observed in patients receiving both chemotherapy and radiation therapy because of the profound injury and in those of young age.³ Two key processes have been noted in children developing cardiotoxicity: reduction in contractility and/or increase in afterload. These have not been described in adults, but may be of significance for the choice of therapy. At present, it is unknown whether the medical therapies outlined above are necessary for agents that cause type 2 cardiotoxicity. Initial studies on trastuzumab-associated cardiomyopathy reported improved cardiac function after withholding therapy alone.^10,22 This has remained the primary approach common to all published management algorithms for trastuzumab-associated cardiotoxicity. However, HER-2 inhibition is a vital element in the treatment of patients with breast cancer overexpressing this receptor, and thus “drug holiday” is of concern. It is currently undefined whether institution of cardioactive medication is necessary and would allow continuation of therapy without concerns for cardiac adverse effects. HER-2 upregulation, which is involved in trastuzumab-associated cardiotoxicity, merits investigation in targeted therapy of this pathway. Intriguingly, statins and nebivolol induce potent upregulation of this pathway. Whether such strategies would be counterproductive to the anticancer effects of therapy, however, remains an unanswered question.

In addition, risk factors for heart failure should also be aggressively treated in cancer patients. Moreover, the higher incidence of drug–drug interactions in cancer patients undergoing chemotherapy and heart failure therapy should be recognized. It is imperative that unnecessary medications are avoided in these patients and that the addition of a potentially beneficial heart failure medication is carefully weighed against the risk of possible interactions with anticancer agents. Several alternative approaches are being developed that may be particularly appropriate for use in cancer patients with heart failure. Stem cell therapy for LV dysfunction, for instance, may become an important strategy in the future; methods of delivery and the enhancement of engraftment remain subject of intense investigation. There are a variety of other potentially useful agents for treating heart failure in cancer patients. IV immunoglobulin, which is used in patients who have immune deficiencies or even an excessive graft-versus-host response after bone marrow transplantation, may have important future implications. However, one randomized, controlled study did not show any clear benefit.²³

Cardiac devices are an integral part of modern heart failure management. Placement of an implantable cardioverter-defibrillator is reasonable in patients with LVEF of < 30% to 35% of any origin with NYHA functional class II or III symptoms who are taking chronic optimal medical therapy and who have reasonable expectation of survival with good functional status of more than 1 year, as per the ACC/AHA heart failure guidelines. Similar considerations pertain to biventricular pacemakers in appropriate cancer patients.

Surgical therapy for heart failure has been intensely investigated, but as yet, there is no clearly established surgical technique for the management of heart failure (see Chap. 70). Coronary artery bypass grafting for the treatment of severe ischemic heart disease may be appropriate when dysfunctional, yet viable, myocardium can be revascularized, thereby restoring function. Other surgical techniques may be appropriate in selected situations resulting in heart failure, such as pericardial stripping for constrictive pericarditis or removal of intracardiac masses. Ethical concerns are also important when deciding on the appropriate therapy for a cancer patient with heart failure. One must consider to what extent the two entities interplay and to what extent they affect the patient’s overall prognosis. It is important to openly address end-of-life issues with patients so that the treatment approach incorporates the wishes of the patient as well as the reality of the clinical condition. The diagnosis of cancer should not imply that the patient’s heart failure cannot or should not be aggressively treated. Treatment decisions should be individualized.

In addition to standard treatment of heart failure, key management of LV dysfunction resulting from cancer chemotherapy includes correction of risk factors, early detection, and regular surveillance. Tables 101–5 and 101–6, adapted from the ESMO guidelines,¹⁶ include key points for consideration while evaluating patients before and during administration of potentially cardiotoxic chemotherapy.

Future Developments

The ability to detect a cardiotoxic insult continues to be a major challenge. Recent use of established biomarkers, such as cardiac troponin I and myeloperoxidase (MPO), in patients receiving chemotherapy has shown promise as an effective tool for identifying toxicity at its earliest stages.²⁴ Use of global longitudinal strain imaging is emerging as a useful echocardiographic tool, but vendor-specific measurements and protocols need to be validated. Prevention of LV dysfunction from cardiotoxicity by use of of specific drugs such as candesartan (see “Prevention of Cardiotoxicity” later in the chapter) are emerging. Future research will include mapping genetic polymorphisms to identify cancer patients who are at increased risk for cardiomyopathy as a consequence of chemotherapy or to serve as a marker of an impaired ability to recover from a cardiotoxic insult.²⁵ This may be especially important when considering antiangiogenic therapy, in which the main goal is to decrease microvascular blood flow to a tumor.

Myocardial Ischemia

Chest pain is a common symptom in cancer patients who are undergoing various forms of anticancer therapy. Chest pain often necessitates the interruption of chemotherapy, while serial cardiac enzyme assessments and ECGs are performed. If non–ST-segment elevation myocardial infarction (MI) or Q-wave MI is confirmed, patients should be managed according to the current ACC/AHA guidelines (see Chaps. 11, 39, and 41). However, thrombocytopenia and brain metastases pose a particular problem for cancer patients with an ACS because anticoagulant/antiplatelet therapies may be contraindicated under these circumstances. Similarly continuing dual antiplatelet therapy after percutanetous coronary intervention in cancer patients with impending cancer surgery, active bleeding, or thrombocytopenia, or in those undergoing bone marrow transplant, poses a clinical challenge where risks for coronary thromboses have to be balanced against those for bleeding.

It is well established that radiation therapy to the mediastinum promotes an increased risk for infarction, and it appears more prevalent in patients with left-sided chest radiation. Additionally, several chemotherapy agents are known to precipitate or exacerbate an ischemic response. Cisplatin infusions can cause chest pain; palpitations; and, occasionally, elevated cardiac enzyme levels indicative of an MI. Cisplatin can also cause cardiovascular complications such as hypertension, LV hypertrophy, myocardial ischemia, and MI as long as 10 to 20 years after the remission of metastatic testicular cancer.⁶ 5-Fluorouracil (5-FU) can also cause an ischemic syndrome that may range from subclinical ischemia to acute MI. Subsequent rechallenge with 5-FU frequently reproduces the initial ischemic event. Nevertheless, the ischemia is usually reversed when 5-FU treatment is stopped and anti-ischemic therapy implemented. In some patients, pretreatment with nitrates and calcium channel–blocking agents has allowed therapy deemed crucial to be continued. Capecitabine, which is currently used in the treatment of breast and gastrointestinal cancers, is believed to be less toxic than 5-FU, although its use has been associated with cardiotoxic effects including ischemic phenomena⁷ arrhythmias, ECG changes, and (rarely) cardiomyopathy. Vinca alkaloids, such as vinorelbine, also have been reported to cause angina associated with ECG changes and arrhythmias, as well as MI. Angina and MI are serious, but relatively rare, consequences of interferon-α therapy. Fatal MI and thrombosis have also been noted after the use of all-trans-retinoic acid (ATRA). Bevacizumab, a recombinant humanized monoclonal IgG₁ antibody that binds to and inhibits the activity of human vascular endothelial growth factor (VEGF), was recently shown to be associated with an increased risk of angina and MI, in addition to serious arterial thromboembolic events, including cerebrovascular accident and transient ischemic attacks. It is used for the treatment of metastatic colon carcinoma and is generally used in combination with other agents. A new class of chemotherapy agents, termed vascular disrupting agents, is currently undergoing clinical evaluation. These agents have been noted to cause asymptomatic creatinine kinase-myocardial bound release and may be associated with ACS.

Blood Pressure Fluctuations

The classic malignancy causing hypertension, often paroxysmal or episodic in nature, is pheochromocytoma, a rare catecholamine-secreting tumor most commonly benign, although malignant in 10% of cases. Pheochromocytomas may be isolated tumors, but can also be associated with multiple endocrine neoplasia (MEN) syndromes of the type 2 variety, particularly the MEN 2A and MEN 2B syndromes. Bilateral pheochromocytomas are also associated with von Hippel-Lindau disease and neurofibromatosis. Benign pheochromocytomas usually can be resected surgically, but patients with such tumors require special anesthetic considerations, including pretreatment with phenoxybenzamine followed by the administration of a β-adrenergic blocking agent. Other tumors may also cause hypertension, including hyperthyroidism associated with thyroid tumors.

Several cancer therapies may cause hypertension or hypotension (see Table 101–1). Newer targeted cancer therapies, which inhibit angiogenesis, frequently cause hypertension in cancer patients. Sorafenib is a potent inhibitor of the VEGFR-2, VEGFR-3, FLT-3, c-kit, and platelet-derived growth factor receptor in vitro and may affect the regulation of endothelial cell proliferation and survival. In clinical trials, sorafenib increases blood pressure in 17% to 43% of patients, whereas sunitinib is associated with slightly less hypertension, seen in 5% to 24% of patients. Similarly, hypertension has been observed in 4% to 35% of bevacizumab-treated patients. The mechanism by which these agents induce hypertension is not fully understood; however, it is thought to be related to VEGF inhibition, which decreases nitric oxide production in the wall of the arterioles and other resistance vessels. When treating hypertension in these patients, some antihypertensive agents may be preferred over others, because each agent affects angiogenesis in different ways. ACE inhibitors may be selected as first-line therapy, because of their ability to prevent proteinuria and plasminogen activator inhibitor-1 expression. In addition, in vivo, ACE inhibitors have demonstrated the potential to reduce microcirculatory changes, decrease the catabolism of bradykinin, and increase release of endothelial nitric oxide. Phosphodiesterase inhibitors and nitrates have also been suggested for the treatment of hypertension in these patients, because of their ability to increase nitric oxide levels. Lastly, the β-blocker, nebivolol, which has the unique mechanism of action of blocking β-adrenergic 1 receptors as well as causing vasodilation through the nitric oxide pathway, may also be beneficial in the treatment of hypertension caused by antiangiogenic cancer treatments.¹⁷ Antirejection regimen such as cyclosporine A, especially when used in conjunction with corticosteroids, are associated with hypertension; the incidence may be greater than 50%. The complex mechanism of cyclopsporin-induced hypertension includes alterations in vascular reactivity that cause widespread vasoconstriction, vascular effects in the kidney leading to reduced glomerular filtration and impaired sodium excretion, stimulation of endothelin, and suppression of vasodilating prostaglandins. Effective therapy includes use of vasodilating agents, often calcium channel blockers.

Hypotension is the most common side effect of etoposide. The infusion of monoclonal antibodies commonly causes hypotension as a result of the massive release of cytokines (an acute transfusion reaction); these agents may also cause fever, dyspnea, hypoxia, and even death. Careful monitoring for hypotension is especially important for patients with preexisting cardiac disease. Cetuximab, a human/mouse chimeric monoclonal antibody that binds to the human epidermal growth factor receptor, may cause severe, potentially fatal infusion reactions characterized by hypotension, bronchospasm, and urticaria; this phenomenon occurs in approximately 3% of patients.^26,27 Rituximab, a chimeric murine/human monoclonal antibody directed against the CD20 antigen, may cause infusion-related side effects that occur within the first few hours of the start of infusion. Supportive measures that are usually effective include IV fluids, vasopressors, bronchodilators, diphenhydramine, and acetaminophen. Interferon-α usually causes acute symptoms during the first 2 to 8 hours after treatment; these include flulike symptoms, hypotension, tachycardia, and nausea, and vomiting.²⁸ The retinoic acid syndrome appears in approximately 26% of patients who receive ATRA, typically within the first 21 days of treatment. This syndrome includes fever, dyspnea, hypotension, and pericardial and pleural effusions.²⁹ High-dose treatment with IL-2 may result in adverse cardiovascular and hemodynamic effects similar to those of septic shock²⁸ and may lead to hypotension, vascular leak syndrome (hypotension, edema, hypoalbuminemia), and respiratory insufficiency requiring pressor agents and mechanical ventilation support.

Blood pressure variations that are unrelated to chemotherapy are much more common in cancer patients; the combination of malnutrition and dehydration is one of the most common causes of hypotension. Other complications of malignancy, most notably sepsis, may be accompanied by profound and refractory hypotension.

Arrhythmias

Arrhythmia in cancer patients may be a consequence of cardiotoxic anticancer therapies, a response to an altered environment wherein the chemical, metabolic, or mechanical abnormalities promote abnormal impulse formation or propagation, or a manifestation of tumor spread. Cancer patients often require strategies for managing their arrhythmias that take into account the underlying condition responsible for the arrhythmia. Additionally, cancer-related arrhythmias encompass the spectrum from trivial to life threatening.

Types of Arrhythmia

Supraventricular Arrhythmias

Supraventricular arrhythmias are a common occurrence in cancer patients. Atrial premature complexes are a well-recognized manifestation of early anthracycline cardiotoxicity and may be a harbinger of the more serious late LV dysfunction. More sustained supraventricular arrhythmias are seen commonly in patients with a chest malignancy, especially lung cancer, which is often associated with increased pulmonary artery pressure. Pulmonary hypertension often precedes atrial flutter or fibrillation. Mechanical effects of an expanding tumor mass, as well as atelectasis or infection in areas distal to occluded bronchi, further increase right-sided pressures and cause arrhythmias. Hypoxia confounds the problem. Supraventricular arrhythmias are especially frequent early in the postoperative period after lung resection. Supraventricular arrhythmias also occur in association with high-dose chemotherapy and stem cell transplantation. Various specific chemotherapeutic agents also have been associated with atrial fibrillation and include 5-FU, gemcitabine, docetaxel, and alemtuzumab. Radiation to the heart has a well-known association with arrhythmia; however, radiation to sites distant from the heart also has been associated with supraventricular arrhythmia. Other forms of sustained supraventricular tachycardia, including reentrant supraventricular tachycardia and multifocal atrial tachycardia, are also seen commonly in the cancer patient.

Intracardiac masses, including benign tumors and malignant processes such as primary cardiac lymphomas and cardiac metastases from lung, breast, or melanoma, often present with supraventricular arrhythmias. Atrial fibrillation is often associated with acute pulmonary embolism (PE) and pericarditis, conditions that may be encountered in cancer patients.

An important risk factor for sustained supraventricular arrhythmias is intrinsic cardiac disease. In particular, conditions resulting in increased atrial size or that are associated with inflammation contribute to elevated atrial pressure and result in atrial rhythm disturbances. Age, hypertension, lung disease, thyrotoxicosis, surgery, and other states that trigger increased catecholamine levels may all also contribute to elevated atrial pressure.

Ventricular Arrhythmias

Cancer patients are more prone to ventricular arrhythmias than the noncancer population. These patients are exposed to stressful and potentially cardiotoxic anticancer treatments, the increased ingestion of multiple noncancer medicines that may be proarrhythmic, and coexisting hormonal and metabolic abnormalities. Hypokalemia, alkalotic states, hypomagnesemia, thyrotoxicosis, pheochromocytomas, and the release of mediators such as serotonin and kinins associated with the carcinoid syndrome are all associated with potentially life-threatening ventricular tachycardia.

Prolonged QT Interval and Torsade de Pointes

Torsade de pointes, a form of ventricular tachycardia associated with prolongation of the QT interval, is related to an increasing list of medications, many of which are used in the management of malignancy and in their supportive care. Prolongation of the QT interval may also occur in association with the supplements that cancer patients sometimes ingest, among them cesium chloride, which is commonly used as an alternative therapy for various types of malignancies. Several anticancer agents may also lead to QT prolongation. Arsenic trioxide is associated with QT prolongation in more than 50% of treated patients.³⁰ Other cardiac side effects include sinus tachycardia, nonspecific ST-T changes, and torsade de pointes.³⁰ In one study, the most common acute side effect was fluid retention with pleural and pericardial effusions. Complete heart block and sudden cardiac death³¹ have also been reported in patients receiving arsenic trioxide. Several newer cancer treatments such as dasatinib, lapatinib, nilotinib, and vorinostat have also been shown to prolong the QT interval. It is important to monitor the QT interval with serial ECGs, with particular attention to patients who are simultaneously receiving other drugs that have the potential to further prolong the QT interval.

Bradyarrhythmias and Atrioventricular Heart Block

Thalidomide may cause sinus bradycardia that may be mitigated by adjusting the dosage and is often asymptomatic, but occasionally permanent pacing may be required. Paclitaxel is used extensively in the treatment of many solid tumors and has been reported to cause sinus bradycardia, heart block, premature ventricular contractions, and ventricular tachycardia.

The conduction system of the heart can be interrupted at all levels by primary and metastatic tumors or by infiltrative processes such as amyloidosis. Pheochromocytomas, thymomas, high-dose chemotherapy, and stem cell transplantation are all associated with block at the level of the atrioventricular (AV) node. Individual chemotherapy agents associated with transient AV block include paclitaxel and octreotide.

Treatment of Arrhythmias

In managing arrhythmias in cancer patients, one must first consider possible cancer-related causes for the rhythm disturbance. In many instances, correcting metabolic abnormalities or removing arrhythmogenic agents is the most successful approach. The arrhythmia may be sufficiently serious that the permanent elimination of the offending agent from the treatment regimen must be recommended. In some settings, the arrhythmia may be controlled by the use of standard medications or procedures, thereby permitting completion of highly effective anticancer treatment. The use of implantable devices should not be denied to appropriate patients who have malignancy and who have reasonable expectation of survival with good functional status of more than 1 year, as per the ACC/AHA heart failure guidelines.

Thromboembolism

General Considerations

The incidence of thrombosis is higher in patients with cancer than in the general population and portends a worse prognosis. The prevention, recognition, and treatment of thromboembolism are clinical challenges for physicians treating patients with cancer. Venous thromboembolism (VTE) occurs in 5% to 7% of patients with malignancy, an incidence that is much greater than that for the general population (approximately 0.1%). Patients with cancer constitute nearly 20% of all cases of thrombosis. Approximately 10% of all noncancer patients with VTE will be diagnosed with malignancy within 2 years.

The incidence of arterial thromboembolism is less than that of VTE in cancer patients, but occurs at a much higher incidence than in the general population. The source of arterial thromboembolism is most likely atherosclerosis, but the general inflammatory condition observed in patients with cancer and treatment with the new antiangiogenic agents are also important determinants. With the more widespread use of such agents, the incidence of arterial thrombosis likely will increase. Thrombotic events, such as deep venous thrombosis (DVT), PE, and arterial thrombosis have also been observed in approximately 11% of patients treated with IL-2.³² The risk of death for cancer patients diagnosed with a DVT or PE is substantially higher than that in cancer patients without such an event. The same is true for arterial thrombi. The in-hospital mortality may approach 20% to 30%, especially in those with pulmonary emboli. Nonbacterial endocarditis (marantic endocarditis) and disseminated intravascular coagulation are additional examples of life-threatening thrombotic conditions associated with cancer.

Another agent most recently linked to increased risk of life-threatening arterial thrombotic events is ponatinib, a third generation tyrosine-kinase inhibitor (TKI), which was developed for the treatment of chronic myeloid leukemia (CML) and Philadelphia chromosome–positive (Ph+) acute lymphoblastic leukemia (ALL). Although ponatinib, similar to the other TKIs, acts as a platelet antagonist, paradoxically, in contrast to the bleeding side effects common to TKI chemotherapies, it can cause severe prothrombotic complications,³³ which resulted in the temporary suspension of the drug by the FDA in late 2013. The pathogenic mechanisms underlying the cardiovascular events observed in patients taking ponatinib remain poorly understood, but can cause arterial thrombosis in up to 20% of patients. The FDA has since lifted the suspension of ponatinib after issuing a “black box warning” regarding the risks of arterial thrombosis to be weighed against the oncological benefits.

Pathophysiology of Thrombosis

The pathophysiology of thrombosis is discussed in Chaps. 32, 33, and 75, and varies somewhat depending on the underlying cause. This section focuses on venous and arterial thrombosis in the cancer patient. However, there is an incomplete understanding of the initial triggers for thrombosis, and no single explanation pertains to all conditions in which thrombosis occurs (Table 101–7).

VTE in cancer patients occurs as a result of the classic triad of stasis, endothelial disruption, and hypercoagulability. In each situation, one component may predominate and, if identified, can be effectively treated. If, for example, a patient has an abdominal tumor that is compressing the inferior vena cava, causing stasis, and this tumor can be removed or reduced by chemotherapy or radiation, then the likelihood of VTE may be markedly reduced. Indwelling IV catheters may cause thrombus formation (Fig. 101–4). Removal of an indwelling central catheter is also likely to be effective in reducing the risk of subsequent recurrence or progression of a thrombus after the stimulus for hypercoagulability has been removed.

Arterial thrombosis in cancer patients is most commonly associated with platelet activation either in the presence or absence of atherosclerosis. The malignancy itself can lead to platelet activation. The exact mechanism also may include microvessel constriction and platelet activation.

Clinical Manifestations

The location of a thrombosis is of paramount importance with regard to how the thrombus will manifest itself. The extent and position of collateral vessels is also crucial in determining how the thrombus will affect the patient. A DVT in a lower extremity, for example, is likely to cause pain, swelling, and erythema of that extremity. A DVT of an arm may be revealed by unilateral arm swelling that develops fairly suddenly or a bluish discoloration caused by poor venous return. Arterial thrombi usually result in pain, pallor, and pulselessness in an extremity. An arterial thrombosis of an internal organ may be manifested by an episode of unstable angina, an ACS, a cerebral vascular accident, or intestinal ischemia that may be difficult to detect.

The most common presentation of VTE is unilateral leg edema. This can be associated with erythema, but typically it is difficult to discern VTE from cellulitis. Unilateral extremity swelling is, therefore, a hallmark of VTE. Bilateral lower extremity edema may be caused by a proximal venous thrombosis or even congestive heart failure, malnutrition, or other causes of edema. PEs are most commonly manifested by a sudden onset of shortness of breath, chest discomfort, and tachypnea. In addition, patients may report hemoptysis and tachycardia. Syncope, hypotension, and sudden cardiac death are manifestations of PEs emboli. Long-term effects of PEs and DVT include pulmonary hypertension and the postphlebitic syndrome (see Chaps. 74 and 75).

The diagnostic approaches used to detect thrombosis, either venous or arterial, in cancer patients are similar to those used in patients without malignancy; these techniques are highlighted elsewhere. Computed tomographic (CT) angiography (see Chap. 17) is a reliable and sensitive tool for detecting PE, and peripheral extremity ultrasonography remains important in the evaluation of venous flow and thrombus (Fig. 101–5). The techniques used to detect arterial thrombus may vary and are determined mainly by the location of the arterial insufficiency. A combination of ultrasonography, sequential pressure measurements, angiography, CT, and MRI forms the basis of diagnostic testing for patients with suspected arterial insufficiency.

Treatment of Thromboembolism

The optimal treatment of VTE is a challenge in cancer patients. Typically, these patients are at increased risk for bleeding and because they are often concurrently anemic, the consequences of bleeding are increased and disproportionate. Additional aggravating factors may be thrombocytopenia, hepatic failure, renal dysfunction, or development of a thrombus in a location that may predispose to bleeding complications. Antithrombotic therapy may consist of medications to prevent or limit the propagation of a thrombus or occasionally thrombolytic agents to promote thrombus dissolution. Furthermore, the risks associated with thrombus extension, distal embolization, or recurrence must be weighed against the risks of hemorrhage as part of the decision-making process.

Guidelines for antithrombotic therapy for VTE have been published by the Ninth American College of Chest Physicians Conference on Antithrombotic and Thrombolytic Therapy.³⁴ Standard treatment for VTE traditionally consists of low-molecular-weight heparin (LMWH), IV unfractionated heparin (UFH), or adjusted-dose subcutaneous heparin followed by long-term therapy with an oral anticoagulant. UFH has been superseded by LMWH as the initial treatment in most cancer patients with VTE in both inpatients and outpatients. Long-term treatment with warfarin may be complicated by several problems, including drug–drug interactions, malnutrition, nausea or vomiting during chemotherapy, and thrombocytopenia. Additionally, hepatic dysfunction in cancer patients may lead to unpredictable levels of anticoagulation and result in increased bleeding complications. LMWH is more effective than oral anticoagulant therapy with warfarin for the prevention of recurrent VTE in patients with cancer who have had acute, symptomatic proximal DVT, PE, or both.³⁵ Further advantages of LMWH are that the doses are more easily adjusted, the pharmacokinetic properties are more predictable, laboratory monitoring is minimized, and fewer drug interactions occur than is the case with oral anticoagulants. Also in favor of heparin therapy in cancer patients are reports that the heparins have antiproliferative, antiangiogenetic, and antimetastatic effects and that LMWH may increase the response to chemotherapy, thereby prolonging the survival of cancer patients.³⁶ For these several reasons, secondary prophylaxis with LMWH may be more effective and feasible than oral anticoagulant therapy in cancer patients with VTE, although the risk of bleeding is the same for both treatment approaches. More recently, low-dose aspirin was safely used to prevent thrombotic complications in patients with polycythemia vera and has been advocated as an alternative option in patients with cancer presenting with paraneoplastic thrombocytosis.

Newer oral antithrombotic agents or NOACs (novel oral anticoagulants) include thrombin inhibitors, such as dabigatran, and factor Xa inhibitors, such as rivaroxiban, edoxaban, and apixaban. These seem to be as effective and safe as conventional treatment of VTE in patients with cancer, as suggested in some small studies and meta-analyses.³⁷ Further clinical trials are warranted to validate effectiveness and safety of NOACs in patients with cancer-associated VTE. Moreover, mechanical treatments, such as compression stockings used to prevent postthrombotic syndrome, have value in the treatment of cancer patients with thromboembolism, as do selected surgical techniques to treat venous ulceration. Venacaval filters may be used instead of anticoagulation in patients with recurrent PEs who also have a high risk of serious bleeding.

Prevention of Thromboembolism

Prompted by the high incidence of VTE in patients undergoing ambulatory chemotherapy, the 2013 American Society of Clinical Oncology (ASCO) guidelines recommend a scoring method (commonly called the Khorana score) as a risk calculator for chemotherapy-associated VTE in the ambulatory setting.³⁸ They recommend that patients with cancer be assessed for VTE risk at the time of chemotherapy initiation and periodically thereafter. The prophylactic anticoagulation of high-risk patients is feasible, safe, and effective as suggested by several randomized clinical trials, including the Prophylaxis of Thromboembolism During Chemotherapy Trial (PROTECHT) and the SAVE-ONCO investigation, the largest thromboprophylaxis study ever conducted in cancer patients.

The key to prevention of cardiotoxicity is a standardized risk assessment (see Table 101–2). Such an approach allows for a universal standard of care for all patients, facilitates multidisciplinary communication, and aids in treatment decisions and follow-up planning. Recent meta-analyses support operational models that incorporate underlying patient-related risk factors, the cardiac toxicity potential of cancer therapies, and the anticipated therapeutic benefit of the anticancer regimen. These models can then be used to individualize the best oncological therapy for the patient balanced against the fewest cardiovascular risks.

Two approaches for primary prevention of anthracycline-induced cardiotoxicity are to (1) reduce cardiotoxic potency by administering via continuous infusion, liposome encapsulation, or using a less cardiotoxic derivative (eg, epirubicin or idarubicin); and (2) use a cardioprotective agent (eg, dexrazoxane) in conjunction with treatment.¹² Other investigated agents include β-blockers, ACE inhibitors, and ARBs. Earlier trials, such as the OVERCOME trial,³⁹ have studied the role of carvedilol and enalapril in prevention of LV dysfunction in patients with hematological malignancies undergoing stem cell transplantation with high-dose chemotherapy, while newer trials such as PRADA⁴⁰ have demonstrated the role of ARBs such as valsartan in primary prevention of LV dysfunction in patients receiving anthracyclines for breast cancer.

The Effects of Radiation Therapy on the Heart

The basic mechanism of cardiotoxicity seems to involve damage to blood vessels, likely mediated by generation of reactive oxygen species, which causes DNA strands disruption and subsequent secondary inflammatory changes, finally resulting in extensive fibrosis. Common histologic changes observed in radiation-associated cardiotoxicity include diffuse interstitial fibrosis of the myocardium with significantly reduced ratio of capillaries to myocytes; arterial and capillary narrowing, leading to myocyte cell death, latent ischemia, and thrombosis; endothelial cell damage, triggering precocious atherosclerosis; replacement of the pericardial fat with dense connective tissue, leading to pericardial fibrosis, effusion, or even tamponade; as well as fibrosis and calcification of valvular leaflets/cusps, more commonly left-sided, which suggests a mechanistic role of blood pressure levels in the pathogenesis of the lesions (Fig. 101–6).

Multiple predisposing factors facilitating the development of radiation-induced cardiotoxicity have been recognized thus far, including total radiation dose and dose per fraction, the volume of heart irradiated, and the concurrent administration of additional cardiotoxic agents, such as anthracyclines and trastuzumab. Younger age at the time of irradiation and the presence of other risk factors for coronary heart disease, such as hypertension, diabetes, and smoking, are patient-specific dynamics that may heighten the risk of radiation-induced cardiotoxicity.⁴¹

Cardiac manifestations of irradiation may be seen even before completion of the anticipated exposure or may occur years or decades later. Patients treated for both Hodgkin and non-Hodgkin lymphoma and patients treated for left-sided breast cancer are the most likely to show these manifestations, because treatment for these tumors sometimes involves portals that overlie the heart and as a result of incremental risks imposed by concomitant use of cardiotoxic drugs such as anthracyclines.⁴²

Radiation Effects on the Pericardium

The pericardium is the most sensitive cardiac structure with regard to irradiation, because it consists of tissue with rapid cell turnover. Initially, exposure can cause an acute radiation-associated pericarditis presenting with signs and symptoms that are clinically undistinguishable from those associated with acute pericarditis from other causes. Sometimes the symptoms may be masked by the underlying disease or the applied chemotherapy.

Acute pericarditis used to be the most frequent complication, but advances reduced its incidence from 25% to 2%.⁴³ Chronic pericarditis still develops with a clinical incidence of 3% at 20 years and 12% at 30 years in those who underwent chest radiation at a dose of 35 Gy or greater.⁴³ Fibrinous exudates, fibrous adhesions, and collagenous thickening (predominantly of the parietal pericardium) are characteristic features. The probability to develop radiation-induced pericarditis is influenced by the applied source, dose, its fractionation, duration, radiation exposed volume, form of mantle field therapy, and the age of the patients.

A normal ECG does not rule out acute radiation-associated pericarditis. The changes appear over a period of days and then dissipate. Imaging should start with echocardiography, followed by cardiac CT or MRI if necessary to rule out constriction. A chest radiograph is usually normal unless a large pericardial effusion is manifested as an increased cardiac silhouette.

Pericarditis without tamponade may be treated conservatively with 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. 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. Pericardial constriction may happen in up to 20% of patients, requiring pericardiectomy. The operative mortality is high (21%) and the postoperative 5-year survival rate is very low (1%), mostly the result of concomitant myocardial fibrosis.⁴⁴

Radiation Effects on the Myocardium, Heart Valves, and Cardiac Vasculature

The myocardium, heart valves, conduction system, and cardiac vasculature are all less frequently affected by radiation than is the pericardium. Rarely, congestive heart failure from myocardial fibrosis or ischemia stemming from small vessel injury, valve sclerosis with or without calcification, and conduction abnormalities that include prolongation of the PR interval and other degrees of AV nodal dysfunction, as well as intraventricular conduction delays, may all occur in cancer patients treated with radiation, especially in doses exceeding 35 Gy. The cardiac effects of radiation are worse in patients who were also treated with anthracyclines, such as survivors of Hodgkin lymphoma.⁴²

Radiation-associated vascular injury is a well-described autopsy finding and is also frequent in young patients in the form of premature coronary occlusion. Radiation may also induce thickening of the arterial wall related to intimal thickening, thereby reducing the luminal area. Additionally, radiation may accelerate atherosclerosis and enhance cholesterol deposition and luminal ulceration. Usually the CAD is 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. Several of these patients with radiation-associated CAD and valvular heart disease end up requiring cardiothoracic surgery, and the prior radiation exposure protends an increased long-term mortality in these patients compared to matched control populations undergoing similar surgical procedures.⁴⁴ The increased cardiovascular risk underscores the importance of identifying individuals who might benefit from primary prevention and aggressive cardiovascular risk factor modification.

Screening echocardiography is advised at 5 years post-exposure in high-risk patients (who have had > 35 Gy of radiation and/or have multiple other cardiovascular risk factors) and at 10 years post-exposure for all others. Alternatively, stress echocardiography, myocardial perfusion imaging, coronary calcium scoring, and coronary CT angiography (CCTA) have been used as more sensitive functional screening methods and should be considered at 5 to 10 years after radiation treatment in high-risk patients (Fig. 101–7).

Benign and Malignant Primary Pericardial Tumors

Both primary and metastatic tumors occur in the pericardium. Tumors are found to have spread to the pericardium in approximately 9% of cancer patients who come to autopsy. In the general population, approximately 7% of all patients who experience acute pericarditis either have a history of malignancy or are later found to have a malignancy related to their pericardial inflammation. Not surprisingly, 35% of invasive pericardial interventions are therefore performed on cancer patients. Among primary pericardial tumors, approximately 25% are malignant. Benign tumors present more frequently in infancy and childhood, and malignant tumors are more common after the age of 30 years; lipoma, a slow-growing benign tumor, is often seen later in life. Pericardial cysts, the most common pericardial tumor, are usually small and are most frequently seen along the right ventricular (RV) border. Pericardial cysts are often discovered as incidental findings at autopsy, but they may present with chest pain or arrhythmias or as a supplementary finding on chest radiography obtained for other purposes. Their presence can be confirmed by ultrasonography, CT scanning, or MRI. These lesions are not malignant, and patients have an excellent prognosis. Table 101–8 enumerates the incidence of tumors (benign and malignant) and cysts of the heart and pericardium, and Table 101–9 lists the general manifestations of neoplastic heart disease.

Pericardial teratomas are usually benign, usually occur in children, and are more common in girls. Compression of the right atrium is sometimes seen with these tumors, and large associated pericardial effusions are frequent. Surgery is the treatment of choice, and the prognosis is usually good when teratomas are benign.

The most common primary malignant tumor of the pericardium is mesothelioma; it usually presents in middle-aged adults, and men are afflicted more often than women. It is uncertain if this gender-related difference is linked to toxic exposure. Pericardial mesothelioma may present with pericardial effusion, pericarditis, constriction, and constitutional symptoms.

Angiosarcomas are the second most common primary pericardial malignancy; these tumors may originate in the right atrium or the pericardium, are usually seen in young adults, and are more common in young men. As with mesothelioma, the presentation may include pericarditis and pericardial effusion. Pericardial effusion is usually hemorrhagic; surgical treatment may be successful. When surgery is not possible, treatment options are limited; palliation with radiation and chemotherapy is suboptimal.

Pericardial lipomas are usually benign and surgically resectable; transformation to malignant liposarcoma is possible. These tumors may also be surgically resected, and repeated surgical resection is sometimes necessary. Thymomas may arise from the parietal pericardium without evidence of anterior mediastinal involvement and may be benign or malignant; they are not associated with myasthenia gravis.

Metastatic spread of tumors to the pericardium is common; it has been estimated that 10% of all malignancies spread to some portion of the heart, and 85% of these eventually involve the pericardium. Lung cancer, breast cancer, and hematologic malignancies together account for approximately two-thirds of the cases of metastatic disease to the pericardium. Melanoma is the tumor most likely to spread to the pericardium, with 70% of patients with metastatic melanoma having pericardial involvement at death. Metastatic breast cancer invades the pericardial space in approximately 21% of cases and lung cancer in approximately 19%.

The clinical presentation of metastatic disease to the pericardium is variable (Fig. 101–8). Sometimes pericardial involvement is an incidental finding at autopsy. It may also present as a catastrophic and life-threatening pericardial tamponade. More commonly, however, patients present with acute pericarditis, chronic effusion, and constrictive disease.

Echocardiography is the most helpful test for evaluating such patients (see Chap. 15), but CT scanning and MRI (see Chaps. 15, 16, 17) are helpful adjuncts. The diagnosis is often confirmed by cytologic analysis of pericardial fluid removed for diagnostic or therapeutic pericardiocentesis; however, cytologic findings are not always positive in patients with malignant effusions, and not all pericardial effusions in cancer patients are malignant. Effusions in these patients may be the result of heart failure, renal failure, inflammation, interruption of lymphatic or venous drainage, infection, hypoalbuminemia, toxicities of chemotherapeutic agents, and chest irradiation.

Cardiac Myxomas

Intracardiac myxoma (Fig. 101–9) is the most frequent benign tumor of the heart. Although most (75%) are located in the left atrium, myxomas are also found in the right atrium (18%), right ventricle (4%), and left ventricle (3%). Cardiac myxomas usually originate from the region of the fossa ovalis, but may arise from a variety of locations within the atria. The DNA genotype of sporadic myxomas is normal in 80% of patients. Tumors are likely to be associated with other abnormal conditions and have a low recurrence rate. Approximately 5% of myxoma patients show a familial pattern of tumor development with autosomal dominant inheritance; 20% of those with sporadic myxoma have an abnormal DNA genotype chromosomal pattern. Carney complex, is characterized by cardiac and mucocutaneous myxomas, lentiginosis, and endocrine dysfunction including bilateral adrenal micronodular hyperplasia. The cardiac tumors are often multicentric, rarely metastasize and are amenable to surgical resection. There at least three different genetic loci with two identified genes for this complex. Most true myxomas arise only from the mural endocardium despite isolated reports that they arise from the cardiac valves, pulmonary vessels, and vena cava.

Pathology of Cardiac Myxomas

Attached to the endocardium by a broad base, myxomas are usually pedunculated, polypoid, and friable, although some may have a smooth surface and are rounded. A myxoma appears as a soft, gelatinous, mucoid, usually gray-white mass, often with areas of hemorrhage or thrombosis. Myxomas vary from 1 to 15 cm in diameter, with most measuring 5 to 6 cm.

On microscopic examination, the myxoma consists of an acid mucopolysaccharide myxoid matrix in which polygonal cells and occasional blood vessels are embedded. Channels, often containing red blood cells, communicate from the surface to deep within the tumor and are lined by endothelial-like cells resembling multipurpose mesenchymal cells, from which the tumor is purported to arise. Similar endothelial cells line the surface of the tumor; however, fibrin, erythrocytes, and organized thrombi also may be present on the surface. Cystic areas; focal or gross hemorrhage; calcification; glandular elements; rarely, bone formation; and even hematopoietic tissue constitute the multiple, less common, variations.

Although asymptomatic patients with myxoma have been reported, most present with one or more effects of a triad of constitutional, embolic, and obstructive manifestations. Cardiac myxomas provoke systemic manifestations in 90% of the patients, characterized by weight loss, fatigue, fever, anemia (often hemolytic), elevated erythrocyte sedimentation rate, and elevated serum immunoglobulin concentration formed in response to tumor embolization, degenerative changes within the tumor, or overproduction of IL-6 by the tumor.

Constitutional manifestations and embolic potential are relatively common in patients with myxoma in any intracavitary location. The cardiac manifestations, symptoms, and physical findings are the consequence of the intracavitary mass and the particular location of the tumor. Myxomas of the left atrium may obstruct either the mitral or pulmonary venous orifices and produce pulmonary venous hypertension, secondary pulmonary hypertension, and right-sided heart failure. The clinical symptoms include dyspnea on exertion, orthopnea, paroxysmal nocturnal dyspnea, acute pulmonary edema, cough, and hemoptysis, along with palpitations, chest pain, fatigue, and peripheral edema. Episodes of syncope or dizziness are frequent, and sudden death may occur. A marked change in the severity of any symptom caused by a change in position of the patient, especially if recumbency relieves dyspnea, is suggestive of myxoma. On physical examination, the S₁ is loud and frequently split, with the second component corresponding to the tumor’s expulsion from the mitral orifice in the case of left atrial myxoma (see Chap. 11). P₂ is accentuated, and an early diastolic sound, the “tumor plop,” is usually heard 80 to 120 ms after the A₂, resembling an opening snap. The tumor plop may be confused with either an opening snap or a third heart sound and follows A₂ at an intermediate interval between these events.

The value of transthoracic echocardiography in the noninvasive diagnosis of intracavitary tumors is well documented (see Chap. 15). M-mode recordings in patients with a prolapsing left atrial myxoma typically demonstrate a diminished ejection fraction slope of the anterior leaflet of the mitral valve, behind which a dense array of wavy tumor echoes is seen (see Chap. 15). The tumor plop coincides with the completion of this anterior movement of tumor echoes. A similar array of tumor echoes may be seen in the left atrium during ventricular systole. Transthoracic echocardiography and transesophageal echocardiography identify the size, shape, point of attachment, and motion characteristics of left atrial myxomas. Transesophageal echocardiography permits superior imaging of the posterior cardiac structures and left atrial myxomas, especially their point of attachment. Visualization of all four chambers permits recognition of multiple tumors, as well as tumors in less common locations. Doppler assessment of the flow patterns of the mitral valve and pulmonary vein provides further information regarding the hemodynamic consequences of left atrial myxomas (see Fig. 101–9).

Surgical resection of a myxoma is the only acceptable therapy and, in view of the dangers of embolization and sudden death, should be performed promptly. For complete removal of left atrial myxoma, we use a biatrial approach, excising a full thickness of interatrial septum if the tumor is attached to the region of the fossa ovalis. Right atrial myxomas are commonly attached to the fossa ovalis, and with right-sided tumors, a full thickness of atrial septum should also be resected. If a large portion of the septum is removed, a patch of knitted Dacron cloth should be used for repair to avoid distortion, arrhythmias, or possible atrial septal defect. Ventricular standstill with cardioplegia solution is induced before manipulating the heart to reduce the possibility of fragmentation of the gelatinous tumor. Left atrial myxomas have been removed successfully during pregnancy, using cardiopulmonary bypass, with subsequent uncomplicated completion of a full-term pregnancy. Surgical removal of a RV myxoma in a neonate has been reported.⁴⁵

Extensive literature has been published over the past 30 years on identifying and minimizing the risks for cardiac complications related to noncardiac surgery (see Chap. 98). Because the surgical treatment of patients with cancer can involve lengthy and complicated procedures, the attendant risks of surgery are often increased.

Preoperative Evaluation

The overall preoperative risk assessment for cancer patients being considered for noncardiac surgery is largely similar to that applied to other groups of patients. The AHA/ACC guidelines (see Chap. 98) that have been developed and updated over the past decade are sufficiently extensive and detailed in this regard. Additional assessment of cancer patients undergoing surgery should also include a review of recent chemotherapy, bleeding risk in case of stents on antiplatelet therapy etc which are summarized in Table 101–10. Chemotherapy affects the immune system and may result in significant thrombocytopenia or neutropenia, which may predispose the patient to infection or bleeding. Also, chemotherapy may result in ECG prolongation of the QT interval and the associated increase in risks for ventricular arrhythmias. It is important to acquire information on such interactions preoperatively because of the bearing it can have intraoperatively and postoperatively. Radiation therapy may have an impact on surgical outcome as well.

A similar approach to the basic laboratory testing and stress testing for risk stratification is used in both patients with and without cancer. However, patients who are undergoing surgery for cancer may have greater debility and thus may not be able to exercise adequately. Consequently, pharmacologic stress testing (see Chap. 52) may be used more frequently in these patients. Although this provides information about myocardial perfusion and reversible ischemia, it does not provide insight into the overall physical capacity of these patients. This information may be relevant after surgery because a poor functional state will likely contribute to a protracted recovery. In several cases, preoperative testing may not influence the timing and choice of surgery, especially if the oncological benefits outweigh the risks. However, it may help the physician to assess the burden of perioperative cardiovascular risk.

Cancer surgery is a particular challenge in patients who have coexisting CAD. Preoperative coronary revascularization does not reduce the risk of the surgical procedure and may actually make patients more prone to complications. This is especially true with regard to percutaneous coronary interventions as stent thrombosis is a feared, albeit rare, complication of premature discontinuation of dual antiplatelet therapy. Conservative medical therapy, instead of percutaneous coronary intervention, may be a preferred management approach for select cancer patients with CAD and may eliminate the need for prolonged dual antiplatelet therapy. In patients with LV dysfunction, optimal management of their volume status perioperatively is equally necessary.

The perioperative management of implanted cardiac devices in patients undergoing surgery for cancer is another important consideration. This holds true especially for those who received prior radiation therapy with the pacemaker in close proximity of the radiation field, thereby making them susceptible to radiation induced device malfunction. The physical distance between the location of the surgery and the device is also important, with a greater likelihood of pacemaker malfunction from electromagnetic interference in case of patients undergoing procedures in the head and neck or chest. If electrocautery is to be used, pacemakers should be placed in a triggered or asynchronous mode; implantable cardioverter-defibrillators should have arrhythmia detection suspended before surgery. If defibrillation is to be used, the current flow between the paddles should be kept as far away from and perpendicular to the lead system as possible.

Patients with certain tumors with a predilection for mediator release present major preoperative challenges that may concern cardiologists. Specific strategies are required in patients with pheochromocytomas to prevent potentially life-threatening blood pressure elevations during the perioperative period, as noted above. Patients with carcinoid tumors may also demonstrate wide fluctuations in blood pressure during surgical resection, albeit through the release of different mediators. The volume status in these patients may also be difficult to discern, especially if concomitant tricuspid regurgitation is present. Transesophageal ultrasonography (see Chap. 15) during surgery may be particularly useful in assessing the intracardiac volume status in patients who are undergoing surgery to remove these challenging tumors.

Postoperative Care

The postoperative management of patients who undergo cancer surgery is similar to that of the noncancer population. For some surgeries, the incidence of postoperative atrial fibrillation, estimated to be as high as 33%, is disproportionately higher than in the noncancer population⁴⁶; a number of these patients have persistent arrhythmias. Restoration of sinus rhythm as quickly as possible following the onset of atrial fibrillation may obviate the need for anticoagulation (see Chap. 81). It is also imperative to reinitiate preoperative cardioprotective medications as early as possible in the postoperative period to preserve stable cardiovascular status. The general principles of early ambulation and “pulmonary toilet” remain standard postoperative surgical principles that should be used aggressively in cancer patients as well.

As per the Cancer Trends Progress Report 2009/2010 Update,¹ the length of cancer survival has increased for all cancers combined. The overall 5-year survival rate for all types of cancer increased from 49% in 1977 to 68% in 2008. Today there are 14 million Americans alive with a history of cancer.

This represents a growing aging population vulnerable to the late risks of potentially cardiotoxic chemotherapy and chest radiation. These patients should be aware of this risk and should subsequently undergo regular monitoring and treatment of cardiovascular risk factors such as hypertension, diabetes, and dyslipidemias. Risk factor modification and symptom awareness may play pivotal roles. Even those who are asymptomatic, but are high risk for developing cardiovascular disease as a result of their past exposure to cardiotoxic chemotherapy such as anthracyclines, combined cardiotoxic chemotherapy regimens (such as doxorubicin with trastuzumab), high doses of chemotherpy (eg, doxorubicin > 350 mg/m²) or radiation (> 35 Gy), and age greater than 65 years, should have a cardiac assessment. This may include screening with imaging such as stress testing, echocardiography, or carotid artery imaging, although definitive guidelines to inform such testing do not exist.

Cancer genomics has focused on the discovery of mutations and chromosomal structural rearrangements that either increase susceptibility to cancer or support the cancer phenotype. Protein kinases are the most frequently mutated genes in the cancer genome, making them attractive therapeutic targets for drug design. However, the use of some of the kinase inhibitors has been associated with toxicities to the heart and vasculature, including ACS and heart failure. There are ongoing studies on the genetic basis of cancer, focusing on mutations in the kinase genome (kinome) that lead to tumorigenesis. This will allow an understanding of the real and potential power of modern cancer therapeutics and the underlying mechanisms that drive the cardiotoxicity of these kinase inhibitors.²⁵

Advances in cancer therapy and the outcome-limiting effect of their cardiovascular adverse effects have generated a growing need for the field of cardio-oncology. With this, the paradigm has shifted toward early recognition and treatment of cardiotoxicity and even precancer therapy, cardiovascular risk assessment, and prevention. The key practical steps in the cardio-oncology approach outlined in this chapter comprise the following:

1. A multidisciplinary approach with ongoing interaction between dedicated specialists comprising cardiologists, oncologists or hematologists, and radiation oncologists in a “cardio-oncology team” approach that evaluates potential cardiotoxic adverse effects of chemotherapeutic agents prior to the start of therapy.

2. Early and routine screening of cardiovascular diseases in cancer patients.

3. Risk stratification for the development of cardiotoxicity prior to the initiation of chemotherapy.

4. Institution-wide algorithms should be formulated based on current evidence-based medicine to guide use of cardiotoxic drugs like anthracyclines and trastuzumab.

5. Evolving imaging techniques such as longitudinal strain may establish themselves as a future modality for detection of early cardiotoxicity.

6. Awareness of cardiotoxic manifestations beyond LV dysfunction of various chemotherapeutic agents and radiation therapy.

We would like to thank Dr. Courtney L. Bickford and Dr. Michael S. Ewer, who contributed to the previous version of this chapter in the 13th edition.