Cardiac surveillance for anti-HER2 chemotherapy

HER2 EFFECTS, TESTING, AND THERAPY

About 1 in 4 patients with breast cancer has an aggressive tumor that overexpresses a tyrosine kinase receptor protein called human epidermal growth factor receptor 2 (HER2, also known as HER2/neu, CD340, Erbb2, and proto-oncogene Neu). It is encoded by the ERBB2 oncogene on chromosome 17.¹ Signaling through this receptor promotes cell proliferation and opposes apoptosis; when it is over-expressed, uncontrolled cell growth results.

Patients with breast cancer undergo HER2 testing to assess prognosis and determine candidacy for personalized therapy. Over the past 20 years, agents targeted against HER2 have been developed, contributing to a halving of breast cancer mortality rates, in what is widely regarded as a phenomenal success story.² Anti-HER2 agents include the following^3,4:

Monoclonal antibodies, eg, trastuzumab, which targets the extracellular domain of HER2, and pertuzumab, which prevents HER2 receptor homodimerization and heterodimerization, which are necessary for activation

Ado-trastuzumab emtansine, an antibody-drug conjugate

Small-molecule inhibitors that block the HER2 receptor intracellularly, eg, lapatinib, available in oral formulations.

Other targeted anti-HER2 therapies are continually being developed.

CARDIAC ISSUES WITH TRASTUZUMAB DISCOVERED EARLY

Current FDA recommendations regarding the frequency of surveillance of left ventricular function with anti-HER2 therapy are conservatively based on historical data involving patients receiving concomitant anthracycline therapy.⁵

The pivotal trastuzumab randomized controlled trial in patients with metastatic breast cancer, published in 2001, reported a 27% rate of cardiac dysfunction and a 16% rate of New York Heart Association (NYHA) class III or IV heart failure in patients who received trastuzumab with anthracycline chemotherapy.⁶ These findings prompted the FDA to issue a stern package-insert warning of cardiomyopathy for anti-HER2 treatments, and recommendations for cardiac surveillance.

Trastuzumab’s package insert recommends measuring left ventricular ejection fraction (LVEF) before starting therapy, every 3 months during treatment, and at the completion of therapy. If drug therapy is withheld for cardiotoxicity, studies should be repeated monthly. Furthermore, after completion of therapy, LVEF should be measured every 6 months for at least 2 years.⁵ Thus, a minimum of 9 echocardiograms is recommended for patients undergoing a standard 12-month adjuvant dosing schedule, with an indefinite (and potentially lifelong) number of 3-monthly echocardiograms for those with metastatic disease on continual anti-HER2 therapy.

RECENT DATA PUT RECOMMENDATIONS IN QUESTION

Subsequent clinical trials^7–16 have generally indicated a more favorable cardiac profile (Table 1).

The 2007 Herceptin Adjuvant (HERA) trial found a 3% rate of cardiac dysfunction and a 0.6% rate of NYHA III or IV heart failure.¹⁷ A 2019 trial⁸ found that only 1.2% of patients discontinued dual therapy because of decreased ejection fraction, while adjudicated cardiac events occurred in less than 1%. The relationship between decreased ejection fraction assessed by cardiac monitoring and the development of clinical heart failure was not discussed.

Although the risk of cardiac dysfunction from anti-HER2 therapy now appears low, the FDA package-insert warning and recommendations remain. Extensive cardiac monitoring and echocardiographic testing regimens are still part of the standard protocols of clinical trials involving this drug.^8,11,18,19 In a 2017 trial,¹¹ up to 13 imaging studies (preferably echocardiograms) were scheduled using the following protocol: at baseline, during treatment (at chemotherapy cycles 2, 6, 10, and 14) and during follow-up (months 3, 6, 12, 18, 24, 36, 48, 60), resulting in a potential total of 19,318 studies for 1,486 patients.

WHAT ACCOUNTS FOR DIFFERENT RESULTS BETWEEN TRIALS?

Several factors may help explain different event rates between clinical trials of the same drug.

Concomitant vs sequential therapy. In early studies, trastuzumab was given concomitantly with an anthracycline and cyclophosphamide. It has since been realized that cardiotoxicity rates are much lower if trastuzumab is given sequentially with other drugs. This is likely the most important explanation of the differences between the early and late anti-HER2 clinical trials.

More surveillance in the drug arm. A 2019 long-term study²⁰ found a higher rate of cardiotoxicity in patients treated with trastuzumab than in those treated with chemotherapy alone. But LVEF was measured 5 times in the trastuzumab group vs no routine testing in the control group. Because cardiotoxicity is more likely to be revealed if more LVEF measurements are taken, more surveillance usually results in findings in the more tested arm.

Exclusion criteria. Cardiac event rates may be underestimated in clinical trials that exclude high-risk patients who are more likely to experience such events.

Problems of definition. Duplicative, inconsistent, and sometimes contradictory consensus criteria to classify cardiotoxicity can affect event rates. For example, a study participant experiencing an asymptomatic drop in LVEF from 60% to 35% might be reported as having either grade 0 left ventricular dysfunction, grade 1 heart failure, or a grade 3 ejection fraction decrease.^21,22

Method of event reporting. Variability in reported outcomes data can arise if studies only include adverse events that are “site reported.”²³ But this is less relevant for objective findings, such as drop in ejection fraction, which should be documented in the primary data. Ideally, all events, whether or not they are thought to be treatment-related, are reported, with details provided for events that are believed not to be treatment-related.²³

Findings from screening using surveillance echocardiography would probably not be confused with acute events associated with other temporary or persisting causes of left ventricular dysfunction (eg, sepsis, acute coronary syndrome, acute arrhythmia including atrial fibrillation, takotsubo cardiomyopathy).

PRINCIPLES OF CARDIAC SURVEILLANCE

More than 50 years ago, Wilson²⁴ wrote about the attributes of an ideal screening test and advised caution: “In theory, screening is admirable, but in practice there are snags; the central idea is simple and may appear deceptively straightforward.” Wilson’s screening criteria and their applications to surveillance echocardiography during chemotherapy are presented in Table 2.²⁴

LVEF WITH ECHOCARDIOGRAPHY IS RECOMMENDED FOR SCREENING

Currently, LVEF is the screening variable of choice.¹⁹ Strain assessment is a nonactionable supportive tool. However, it is the focus of ongoing research and is increasingly being used, especially as it received a formal Current Procedural Terminology code by the US Centers for Medicare and Medicaid for reimbursement to Medicare providers.²⁵

Echocardiography is the preferred screening method, although cardiac magnetic resonance imaging is considered to be the gold standard and is advised in selected cases (ie, if echocardiographic images are inadequate or yield equivocal findings). Another option, multigated acquisition radionuclide scanning, is not a first-line test, as it involves radiation and introduces cross-modality error.

BALANCING THERAPY RISKS AND BENEFITS

The net benefit of therapy refers to balancing the risks of toxicity with prognosis and available treatment options. Potential cardiotoxicity may be more acceptable in the setting of a cancer with a poor prognosis and few treatment possibilities. On the other hand, cardiotoxicity is less likely to be an acceptable risk for a later-generation drug in a cancer with multiple existing therapies and a generally good prognosis.

Regarding breast cancer, regimens without an anthracycline have been shown to be as effective as those with an anthracycline, especially for women at low risk of recurrence. Strategies without an anthracycline involve much lower rates of cardiotoxicity, with rates of NYHA class III and IV heart failure being close, if not equal, to those with placebo (0.4% over 5 years or fewer than 1 per 1,000 patients per year).¹⁸ They have also demonstrated improved survival and favorable cardiac safety for metastatic cancer.²⁶

Because anti-HER2 treatment is used against a particularly aggressive cancer, decisions regarding interrupting or stopping it based on side effects have especially important implications. Whether such decisions should be made based on a surrogate echocardiographic end point, possibly in the absence of symptoms, needs careful consideration.

CARDIAC MANAGEMENT AND ANTI-HER2 THERAPY

Anti-HER2 treatment in patients with preexisting cardiac dysfunction has been associated with a worse prognosis and higher rate of symptomatic heart failure compared with patients with preserved ejection fraction at baseline.²⁷ However, preexisting cardiac dysfunction is a relative rather than an absolute contraindication to starting anti-HER2 treatment. The FDA recommends extreme caution in treating such patients,⁵ and a cardiologist should be involved in management.

For patients who develop clinically significant congestive heart failure, discontinuing anti-HER2 therapy should be strongly considered. For patients without symptoms, treatment-specific LVEF thresholds for stopping medications have been developed, with slightly different recommendations between FDA-approved labeling, clinical trial protocols, and professional society guidelines. Criteria from clinical trials that do not involve anthracycline therapy tend to be a little less stringent because anti-HER2-associated toxicity is considered to be dose-independent, nonapoptotic, and potentially reversible (type 2 cardiotoxicity). In contrast, anthracycline-mediated cardiotoxicity is regarded as type 1 (ie, irreversible and related to cumulative dose).²⁸

A commonly used threshold defining cardiotoxicity is a decrease in LVEF of more than 10% to a value below the lower limit of normal. Hussain et al,²⁹ in a study of 23 patients with asymptomatic LVEF decline who continued trastuzumab, found that 14 patients (61%) tolerated it without a cardiac event, 6 (26%) developed further worsening of LVEF, 1 (4%) developed heart failure, and 2 (9%) died of a possible or probable cardiovascular cause.

Strategies to prevent or attenuate cardiotoxicities include participation in cardiooncology programs (particularly for symptomatic or high-risk patients being considered for anti-HER2 treatment, including anyone with baseline low LVEF), early recognition of cardiac side effects, active cardiac surveillance, and cardioprotective medical therapy.

INTERPRETING SERIAL TESTING IS A CHALLENGE

Accurate diagnosis of cardiotoxicity is critical, as false-positive results may lead to inappropriate stopping of potentially lifesaving chemotherapy.

Serial echocardiography in patients with cancer can be difficult and measurement variability may be high. Reasons may be technical (eg, concomitant lung disease, high or low body mass, postoperative status) or involve confounding factors (eg, variable hemodynamics, medications, fluid status).³⁰ Published test-retest variability data have generally been derived between 2 tests rather than multiple tests and conducted under optimized experimental settings in academic centers. Even optimized test-retest variability remains close to echocardiographic thresholds used to define real interval change representative of true cardiotoxicity, especially with multiple tests.

Outside of trial settings, false-positive results are not infrequent. In addition, pre-chemotherapy studies may manifest hyperdynamic function. Teasing out whether serial changes are related to cancer therapy vs comorbid illness (eg, concomitant arrhythmia, ischemia, stress cardiomyopathy, and myocarditis) may be challenging.

A PATH FORWARD

We suggest a more focused cardiac surveillance approach to low-risk, asymptomatic patients receiving anti-HER2 treatment. Routine serial LVEF measurement by echocardiography should be done only if patients have received anthracyclines or are considered to be at high risk (eg, concomitant hypertension, borderline low LVEF). For these patients, studies should be carried out at baseline, post-anthracycline (if appropriate), and every 3 months while on anti-HER2 treatment. Less frequent testing may be justified for patients with metastatic disease who have repeatedly normal LVEF test results. Patients should be informed about potential symptoms of cardiotoxicity and advised to report them promptly.