Pathophysiology of Heart Failure

Clinical heart failure occurs when the heart is unable to provide a sufficient oxygen supply to meet the metabolic needs of the body at normal cardiac filling pressures. In the setting of cancer therapy, the pathophysiological process leading to clinical heart failure is initiated by an extrinsic cardiac injury due to radiation, anthracyclines, or other toxins with consequent focal or segmental myocyte death (mediated by necrosis and/or apoptosis). The extrinsic cardiac injury triggers an intrinsic pathological process (ventricular remodeling) characterized by activation of cell repair and regenerative pathways, myocellular hypertrophy, and interstitial fibrosis in the surviving myocardium. The signal transduction pathways that regulate these processes in the heart overlap with pathways for cell growth and proliferation that are known to be important in tumor biology. Initial cellular responses to cardiac injury serve to preserve cardiac output (and oxygen delivery) and therefore are not typically associated with clinical manifestations of heart failure. The progression to chronic heart failure is an indolent process that frequently occurs 10 to 20 years after the initial injury. The development of clinical manifestations of chronic heart failure represents the end stage of a largely occult disease process. With the introduction of modern therapy, there has been a substantial reduction in mortality for patients with systolic heart failure. However, symptomatic heart failure continues to confer a worse prognosis than the majority of cancers in the United States, with 1-year mortality rate as high as 45% in the elderly patient population.²

Cardiotoxicity associated with radiation and chemotherapy can be categorized according to the type and clinical manifestations of cardiac injury.³ Five types of myocyte injury have been proposed with different pathophysiological mechanisms (Fig. 59-1). Among these, type I is primarily associated with anthracyclines and is attributable to direct myocyte toxicity with consequent cell death mediated by either apoptosis or necrosis. Type II is primarily associated with trastuzumab and related agents and is attributable to inhibition of specific cardiac repair pathways that may lead to transient myocellular dysfunction or cell death. The notion that trastuzumab-related cardiomyopathy (type II) can be reversed has been challenged by recent studies.⁴ Because myocytes have a limited repertoire of response to injury, the long-term clinical consequences of all five types of injury follow a final common pathway of pathological cardiac remodeling and progression to heart failure, as previously described.

Given this pathophysiological model of disease, the common terminology criteria for adverse events recorded in clinical trials may underestimate the risk of late sequelae of cardiotoxicity in long-term cancer survivors. This model also offers several important corollaries for the practicing oncologist. First, an observed change in left ventricular (LV) contractile function (most commonly assessed in clinical practice as ejection fraction [EF]) is not closely associated with clinical manifestations of heart failure. Second, a substantial decrease in EF (>10 percentage points from baseline, or to any value <55%) may or may not be associated with clinical heart failure in the acute setting but is likely associated with an increase in the long-term risk for later development of heart failure (with associated high mortality risk). Third, use of chemotherapeutic agents that specifically inhibit pathways related to hypertrophy and proliferation (trastuzumab and related agents) may have an adverse impact on the myocellular responses to injury after anthracycline therapy and increase risk of development of cardiomyopathy and clinical heart failure (CHF).

Anthracyclines (Type I)

The currently approved anthracyclines in the United States are doxorubicin (Adriamycin), liposomal doxorubicin and pegylated doxorubicin (Doxil), epirubicin (Ellence), daunorubicin (Cerubidine), and idarubicin (Idamycin PFS). Anthracyclines have played a major role in the treatment of hematologic malignancies and solid tumors, especially in the treatment of both metastatic and early breast cancer and sarcomas. Their dose-dependent cardiotoxicity is well documented, with the most common clinical presentation being dilated cardiomyopathy with chronic heart failure. Most available information on anthracycline-induced cardiomyopathy is derived from studies with doxorubicin and, to a lesser extent, with epirubicin and daunorubicin. However, other drugs of this class have also been associated with clinically indistinguishable cardiotoxic effects. The clinical presentations of cardiotoxicity can be acute, subacute, and/or chronic. No consensus exists for the definition of these terms. The previous definition for “acute” is the onset cardiac events within a few minutes after administration of anthracycline, which is mainly based on transient arrhythmias and rare cases of acute pericarditis and myocarditis.⁵ The elevation of troponin, new echocardiographic technology, and CMR allow the early detection of cardiotoxicity. We propose to define these terms as follows (Table 59-1).

Acute Cardiac Toxicity

Acute cardiotoxicity may manifest as acute systolic dysfunction, arrhythmia, and/or clinical pericarditis-myocarditis during and/or soon after the administration of doxorubicin.6,7 These acute forms of cardiotoxicity are rare and usually transient or reversible. The most common form of acute cardiac toxicity is acute cardiomyocyte injury, which is associated with elevated serum troponin levels in the days or weeks after administration of anthracyclines. Elevation of troponin is associated with an increased risk for the development of chronic heart failure.⁸

Subacute and Chronic Cardiac Toxicity

Endomyocardial biopsy specimens have demonstrated that the dose-related cardiotoxicity associated with anthracyclines is biventricular. Retrospective studies indicate that the incidence of clinically recognizable heart failure with doxorubicin is 7% to 15% in patients who have received a cumulative dose greater than 450 mg/m² to 500 mg/m² without use of a cardioprotective agent (Table 59-2).^9,10 Above this dose, the incidence of CHF rises steeply, with an estimated incidence of CHF of 50% when the cumulative dose of doxorubicin reaches 700 mg/m². Cumulative cardiotoxic doses associated with the use of other anthracyclines vary, but the shape of the curve that plots cumulative dose against the incidence of CHF is comparable for all anthracyclines that have been studied. Careful surveillance for cardiotoxicity should be maintained from the start of therapy because isolated instances of CHF have been observed at lower cumulative doses.¹¹

The incidence of CHF in clinical practice may be higher than that reported from retrospective trials.12,13 Prospective clinical trials indicate that the number of patients with evidence of CHF at cumulative doses of 450 mg/m² may exceed 25%. Subclinical cardiotoxic effects of anthracyclines are likely present after each dose, with the likelihood of clinically apparent toxicity determined by the cumulative effects from prior doses of anthracyclines or from other co-morbid conditions. The natural history of the chronic anthracycline cardiotoxicity is highly heterogeneous. In its most virulent form, patients demonstrate refractory progressive symptoms despite optimal medical therapy with rapid onset of pump failure and death. Most patients present with permanent reduction in LVEF and persistent symptoms of CHF that progress slowly over time. In some patients, LVEF can improve over time, sometimes for years, with later recurrent deterioration of LVEF and eventually the appearance of CHF.

Risk Factors

The most important risk factor for cardiac toxicity is the total cumulative dose of anthracycline. Other factors associated with an increased risk of anthracycline-induced cardiotoxicity in adults include age greater than 70 years, exposure to ionizing radiation to the chest wall, prior exposure to anthracyclines, and preexisting cardiac disease or traditional cardiovascular risk factors including history of CHF, myocardial infarction, hypertension, and diabetes mellitus. Concomitant use of other agents with cardiotoxicity such as cyclophosphamide and paclitaxel also confer increased risk of anthracycline cardiotoxicity. Concurrent and sequential use of trastuzumab (Herceptin) with anthracycline-based adjuvant chemotherapy in patients with HER 2+ breast cancer significantly increases risk for cardiac toxicity (Table 59-2). The independent contributions of each of these factors (with the exception of trastuzumab) are not well defined, but having multiple risk factors appears to be associated with additive risk.

Heterogeneity in individual patient susceptibility to anthracycline cardiotoxicity suggests that genetic variation may also contribute to risk. Genetic variants in genes encoding proteins related to free radical metabolism, anthracycline transport, and drug biotransformation have been shown to be associated with increased risk in anthracycline-induced cardiotoxicity in both pediatric and adult populations of cancer survivors.16–16 These preliminary findings need to be further replicated and confirmed in larger prospective studies.

Mechanisms

The antitumor effects of anthracyclines are mediated by inhibition of DNA synthesis through direct interactions with DNA (intercalation) and inhibition of topoisomerase II (Top II).¹⁷ Top II has two isoenzymes in mammalian cells, Top IIα and Top IIβ. Doxorubicin or daunorubicin binds both DNA and Top IIα to form the topoisomerase IIα-doxorubicin-DNA ternary cleavage complex, which triggers cell death.¹⁸ These tumoricidal mechanisms do not likely contribute to cardiac toxicity because Top IIα is overexpressed in tumor cells and is not detectable in quiescent cells.¹⁹ Cardiomyocytes are terminally differentiated cells with very low rates of proliferation.

Anthracycline-induced cardiomyopathy is the result of a series of repeated chemical injuries to the myocyte that result in progressive cell dysfunction and, eventually, cell death. Early myocyte injury may be partly reversible via activation of intrinsic myocyte repairing signaling pathways, replacement of lost cells from resident myocyte progenitor cell populations in the heart (or possibly from bone marrow–derived stem cells), and the termination of the use of anthracycline. However, with repeated dosing and/or concomitant impairment of the intrinsic myocyte repair pathways (with use of trastuzumab or related drugs), cumulative myocyte loss will eventually activate pathological hypertrophy and fibrosis signaling in the remaining myocardium with eventual development of CHF. Once the limited reparative reserve of the myocardium has been exhausted, the process is irreversible with progression to CHF and death. This pathological process is consistent with the gradual onset, variability of course, risk factors for increased susceptibility, and laboratory and pathological findings of the cardiomyopathy that results from use of these drugs.

The primary mechanism for anthracycline cardiomyocyte injury was thought to be free radical damage to vital cell structures and subsequent cell death via apoptosis. Reduction of the quinone groups on the B ring of the anthracene structure results in a semiquinone radical before further reduction to the alcohol (e.g., doxorubicinol). Interaction of the semiquinone with oxygen yields superoxide anion (O₂⁻). Dismutation of superoxide anion yields H₂O₂, which may react further with cardiac iron sources or the semiquinone to produce a reactive intermediate with the chemical characteristics of the hydroxyl radical. These reactions can proceed in either an iron-dependent or iron-independent fashion (Fig. 59-2).²⁰ Doxorubicin also directly increases the labile intracellular iron pool by inhibiting iron regulatory protein–1 with consequent changes in transcription of transferrin receptors and ferritin.^21,22 Formation of a Fe³⁺–doxorubicin complex can also enhance free radical generation via the Fenton reaction. Free radicals can cause damage at a variety of intracellular sites, including the nuclear envelope, cell membrane, mitochondria, DNA, and sarcoplasmic reticulum.²³ Increased susceptibility of myocardial cells to free radical injury may be attributable to very low myocardial levels of catalase, one of major cellular enzymatic defenses against free radical injury, in cardiomyocytes when compared with other cells types. Doxorubicin also decreases levels of other antioxidant enzymes, including selenium-dependent glutathione-peroxidase-1 and cytosolic CuZn superoxide dismutase. Recently, Top II–mediated doxorubicin-induced DNA damage was shown to play an important role in cardiotoxicity. As previously mentioned, Top IIα is only expressed in proliferating and tumor cells and is the target of the antitumor effect of anthracyclines. Cardiomyocytes express Top IIβ, which is also a target for doxorubicin. Top IIβ-doxorubicin-DNA ternary complex can induce DNA double-strand breaks, leading to cell death. A recent study showed that cardiomyocyte-specific deletion of Top IIb (encoding Top IIβ) protects cardiomyocytes from doxorubicin-induced DNA double-strand breaks and transcriptome changes that are responsible for defective mitochondrial biogenesis and reactive oxygen species formation. Furthermore, cardiomyocyte-specific deletion of Top IIβ protects mice from the development of doxorubicin-induced progressive heart failure.²⁴ Anthracyclines may also mediate cardiotoxicity through other iron-independent mechanisms related to changes in myofibrillar structure and function (see Fig. 59-2), changes in calcium and energy metabolism, and secondary changes in nitric oxide signaling in response to altered redox state in the myocardium.²⁵

Diagnosis

The majority of cardiotoxicity is subclinical and is manifested solely by elevated troponin or left ventricular dysfunction as detected by echocardiography or multiple gated acquisition (MUGA) scans. As discussed in the section on early detection of cardiac toxicity, early detection of cardiotoxicity is an important strategy for the prevention of subsequent development of cardiomyopathy/CHF. Consensus does not exist about whether an elevated troponin level after anthracycline use requires cessation of chemotherapy. However, an early rise in the troponin level may be useful for identification of patients who are more likely to derive benefit from the concurrent use of a cardioprotective agent (e.g., dexrazoxane) and early use of angiotensin-converting enzyme (ACE) inhibitors and β-blockers. A significant reduction of LVEF (20 percentage points or 10 percentage points and LVEF less than 55%) should prompt consideration of cessation of anthracycline therapy. However, when the chemotherapy is being given for curative intent, the oncologist and cardiologist must carefully balance the individual risk/benefit ratio to ensure that reduction of cardiotoxicity is not achieved at the expense of the curative potential of chemotherapy.

Despite its limitations (discussed in the next section), serial measurement of LVEF remains the most commonly used method for detection of anthracycline cardiotoxicity. CHF may occur with preserved or decreased EF. In patients with preserved EF, heart failure is thought to be primarily attributable to changes in diastolic properties of the heart (increased diastolic LV wall stiffness and/or increased end-diastolic LV volume as a consequence of cardiac injury or overload). The classic clinical signs of congestion associated with heart failure (rales and edema) have both low specificity and low sensitivity. More than half of patients with documented extreme elevation of cardiac filling pressures do not present with rales or radiographic signs of pulmonary congestion. Accordingly, the clinical oncologist must maintain a high level of suspicion in patients receiving chemotherapy associated with potential cardiotoxicity. New onset of otherwise unexplained exertional dyspnea and/or fatigue may be the sole manifestation of heart failure regardless of change in LVEF. Unexplained sinus tachycardia at rest and after exercise in an otherwise oncologically stable patient is another potential early sign of cardiotoxicity. A system of staging anthracycline-induced hemodynamic changes in humans was developed by Bristow.²⁶ Early referral to cardiology is recommended in the setting of suspected cardiotoxicity, even in asymptomatic patients, because accumulating evidence suggests that medical therapy may forestall disease progression. The relationship between staging criteria for heart failure developed by the American Heart Association, American College of Cardiology, and New York Heart Association²⁷ in comparison with the Common Terminology Criteria for Adverse Events is summarized in Table 59-3.

Routine Monitoring of Cardiac Toxicity

Serial determination of LVEF, by either echocardiography or MUGA radionuclide scans, is the most commonly used strategy for detection of anthracycline cardiotoxicity. Because reduction in EF occurs only after a large cumulative cardiac injury has occurred, both imaging techniques lack sensitivity for detection of early subclinical cardiac injury. It is also important to note that LVEF is highly dependent on loading conditions of the heart and thus must be interpreted with knowledge of the blood pressure (BP) at the time of the study. Echocardiography offers a more comprehensive assessment of cardiac structure and function, including detection of co-morbid valvular and pericardial disease, but MUGA scans provide a more reliable quantitative assessment of LVEF.

A commonly used strategy for patient monitoring is to obtain a baseline study before therapy commences and then obtain repeat studies at intervals as the cumulative dose increases, with more frequent observations at higher cumulative doses. Scans are examined for sequential changes in wall motion and global LVEF. Any decrease in LVEF, including those that remain within the “normal” range and those that decrease to clearly abnormal values, may indicate anthracycline myocyte damage.26,28 Careful monitoring with serial MUGA scans may permit treatment with greater cumulative doses of doxorubicin beyond the empiric stopping dose of 450 to 500 mg/m.

Trastuzumab is frequently given in association with anthracycline-based chemotherapy in patients with HER2-positive breast cancer. Its cardiac toxicity does not appear to be dose related. The same ventricular function monitoring strategy should be used during the trastuzumab treatment period because of the significant increase of occurrence of trastuzumab-related cardiotoxicity.

It is important to bear in mind that the likelihood of CHF increases disproportionately with increased exposure, that early toxicity may be difficult to quantify with imaging alone, and that an additional two or three cycles of chemotherapy may expose patients to life-threatening cardiac complications. Decreases in the absolute value of LVEF by more than 10 percentage points or to less than the lower limit of normal for the laboratory may warrant discontinuation of therapy, because a decrease in LVEF usually precedes the development of clinical heart failure. This algorithmic approach has been validated as a clinical practice guideline, but it is not fail-safe. When clinical circumstances surrounding treatment of the primary cancer indicate that exceeding recommended cumulative dosages is possible, cardioprotective strategies should be considered.

Endomyocardial Biopsy

Endomyocardial biopsy remains the gold standard for diagnosis because of its high sensitivity and specificity. Anthracyclines cause a unique pattern of histologic damage as originally demonstrated in animal models and confirmed by endomyocardial biopsies in patients.²⁹ Billingham grade has been used as a histopathological scale for doxorubicin-induced cardiomyopathy (Box 59-1).²⁹ A continuum of change is well described, from dilation of vacuoles to mitochondrial swelling, myofibrillar dropout, interstitial fibrosis, and, ultimately, cell death.^31–31 The changes seen on endomyocardial biopsy appear to parallel clinical findings.³² The consistency of these findings in nonmammalian species has provided the basis for a number of animal models of anthracycline cardiomyopathy.³³ However, endomyocardial biopsy is rarely used in clinical practice because of the risks associated with an invasive procedure, absence of data to demonstrate that the results of the test alter treatment approach or outcome, and the availability of noninvasive alternatives to monitor cardiac toxicity.

Early Detection of Cardiac Toxicity

A substantial amount of myocardial damage may have already occurred by the time a decrease in LVEF is evident by echocardiogram or MUGA imaging. Earlier detection of cardiac toxicity can significantly reduce the risk for future development of clinical manifestations of heart failure and could allow individualization of treatment regimens (including use of cardioprotective agents) and frequency of cardiac imaging testing.