Cardiovascular toxicity associated with adjuvant trastuzumab therapy: prevalence, patient characteristics, and risk factors

Abstract

Before the advent of the human epidermal growth factor receptor 2 (HER2)-targeted monoclonal antibody trastuzumab, HER2-positive breast cancers were difficult to treat and had a poor prognosis. Adjuvant trastuzumab is now an important part of the treatment regimen for many women with HER2-positive breast cancer and has undoubtedly resulted in a significant improvement in prognosis, but it is associated with a risk for cardiotoxicity. In this review, we describe the prevalence, patient characteristics, and risk factors for cardiotoxicity associated with use of adjuvant trastuzumab. Understanding risk factors for trastuzumab-induced cardiotoxicity and appropriate patient monitoring during trastuzumab treatment allows for safe and effective use of this important adjuvant therapy.

Introduction

Breast cancer is the most common cancer in women in the United States [Siegel et al. 2012] and worldwide [Coughlin and Ekwueme, 2009]. Breast cancer also occurs in men, but is approximately 100 times less common in men than in women [American Cancer Society 2013]. The National Cancer Institute and American Cancer Society estimate that in 2012 there were nearly 3 million breast cancer survivors in the United States, with an additional 226,870 new diagnoses expected [Siegel et al. 2102]. Since the 1970s, major advances in screening technologies and treatment options have resulted in significant declines in breast cancer mortality [Berry et al. 2005]. The development of trastuzumab for the treatment of human epidermal growth factor receptor 2 (HER2)-positive breast cancer in the late 1980s essentially turned the tables on the treatment of what was previously considered an aggressive and difficult to treat breast cancer with a poor prognosis. However, despite decreases in the last decade, morbidity associated with breast cancer treatment remains of major patient concern and often has a significant impact on quality of life in survivors [Binkley et al. 2012; Brem and Kumar, 2011]. In this review, we describe the prevalence, patient characteristics, and risk factors for cardiotoxicity associated with adjuvant trastuzumab.

Adjuvant trastuzumab for the treatment of breast cancer

In the initial report of HER2 overexpression in breast cancer by Slamon and colleagues [Slamon et al. 1987], amplification was noted in approximately 30% of breast cancers, although a range of 10–30% has been reported in the literature, with approximately 25% when standardized testing is applied [Bilous et al. 2003]. HER2 overexpression is associated with a more aggressive phenotype and poorer prognosis [Slamon et al. 1987]. Trastuzumab (Herceptin; Roche, San Francisco, CA) is a monoclonal antibody that binds specifically to the extracellular domain of the HER2 protein, disrupting the normal regulatory functions of HER2 in cell growth, differentiation, and survival. The addition of adjuvant trastuzumab to treatment regimens represents a major advance in the treatment of HER2-positive breast cancer. In a 2007 meta-analysis of five randomized controlled trials of adjuvant trastuzumab for the treatment of HER2-positive breast cancer, Viani and colleagues demonstrated a significant reduction in mortality, recurrence, and metastases with adjuvant trastuzumab compared with patients who did not receive trastuzumab [Viani et al. 2007]. Further follow up of these trials through 2010 demonstrated a 24–58% improvement in disease-free survival with adjuvant trastuzumab treatment as well as significant improvement in overall survival by 23–35% [Costa et al. 2010; Gianni et al. 2011]. As such, adjuvant trastuzumab therapy has become the standard of care for women with early-stage, HER2-positive breast cancer.

Currently, trastuzumab is indicated in the adjuvant setting for stage I–III breast cancer, which will be the focus of this review, and is also in use for metastatic breast cancer, as well as metastatic gastric and gastroesophageal cancers. Trastuzumab has also been shown to be effective in neoadjuvant settings [Moja et al. 2012]. Trastuzumab is given intravenously and may be given as a single agent or in combination with chemotherapy or hormone therapy. The standard duration of adjuvant trastuzumab treatment for breast cancer is for a total of 52 weeks given either weekly or every 3 weeks following an initial loading dose.

Cardiovascular side effects of trastuzumab treatment

Trastuzumab treatment is generally well-tolerated with the potential for trastuzumab-induced cardiotoxicity as the morbidity of primary concern. The mechanism by which trastuzumab causes cardiotoxicity is not completely understood, but it is thought to be related to blocking of the normal physiologic action of HER2 on cardiomyocytes [De Keulenaer et al. 2010], as well as potential effects on the function of resident cardiac stem cells [Barth et al. 2012]. In the absence of HER2 function, cardiomyocytes are unable to activate survival pathways and reactive oxygen species (ROS) accumulate resulting in cardiac dysfunction [Zeglinski et al. 2011]. Similarly, cardiac stem cells appear to lose their capacity for cardiogenic differentiation and formation of microvascular networks [Barth et al. 2012].

Trastuzumab-induced cardiotoxicity presents with a range of severity from asymptomatic decline in left ventricular ejection fraction (LVEF) to symptomatic heart failure [Chien and Rugo, 2010] and does not appear to be related to either dose or duration [Tripathy et al. 2004]. As such, in the early stages of trastuzumab-induced cardiotoxicity, withdrawal or discontinuation of the drug can reverse the effects with the support of standard cardiac therapy for heart failure [Ewer et al. 2005], including initiation of β blockers and angiotensin-converting enzyme inhibitors (ACEIs) [Perez and Morgan, 2013]. While the true reversibility of trastuzumab-induced cardiotoxicity is under debate, there is general agreement that conditions improve in most patients following trastuzumab withdrawal and treatment of cardiac symptoms [Telli et al. 2007]. Rechallenge with trastuzumab following discontinuation of trastuzumab and treatment of cardiac symptoms has been reported to be well-tolerated [Ewer et al. 2005]. In a 2005 study by Ewer and colleagues, 25 patients were rechallenged with trastuzumab after recovery of LVEF, and 22 (88%) maintained stable LVEF during a median follow up of 8.4 months. Even so, rechallenge is generally recommended primarily in the metastatic treatment setting or following careful evaluation of risks and benefits of continuation of the drug [Ewer and Ewer, 2008]. Episodes of trastuzumab-induced cardiotoxicity may have long-lasting effects on cardiac health, as evidenced by myocyte damage and indicated by cardiac troponin I elevation [Cardinale et al. 2010], suggesting that even if LVEF recovers, the damage may leave the patient susceptible to future insults [Telli and Witteles, 2011].

Prevalence of cardiotoxicity with trastuzumab treatment

Trastuzumab was first approved in 1998 for the treatment of HER2-positive metastatic breast cancer, and cardiac dysfunction was recognized early on as a potential adverse consequence of trastuzumab treatment. This recognition prompted major changes in the design of clinical trials, including more stringent cardiac exclusion criteria, rigorous prospective monitoring of cardiac function, and interim cardiac safety analyses [Seidman et al. 2002]. In a review of cardiac dysfunction in the pivotal phase II and III trials of trastuzumab for the treatment of metastatic HER2-positive breast cancer, the independent Cardiac Review and Evaluation Committee (CREC) reported rates of cardiac dysfunction ranging from as low as 8% with trastuzumab alone to nearly 30% in patients treated with concomitant anthracycline and trastuzumab [Seidman et al. 2002]. Safety data from the randomized clinical trials of trastuzumab as an adjuvant therapy for early stage and locally advanced HER2-positive breast cancer continue to demonstrate that trastuzumab treatment results in a significant increase in the risk of both LVEF decline and symptomatic heart failure [Moja et al. 2012].

In a 2012 Cochrane review, Moja and colleagues summarized findings from eight major randomized clinical trials that evaluated the efficacy of trastuzumab for patients with early and locally advanced breast cancer, including the B31, BCIRG006, Buzdar, FinHer, HERA, N9831, NOAH, and PACS-04 trials [Moja et al. 2012]. The Buzdar trial ended after enrollment of only 23 trastuzumab-treated patients due to demonstration of trastuzumab efficacy early on during recruitment and is therefore not discussed here further [Buzdar et al. 2005]. Study arms and the definition of cardiotoxicity used for study purposes are described in Table 1. Considerable variation in chemotherapy regimens used and differing definitions of cardiotoxicity make direct comparison of these trials difficult. However, the addition of trastuzumab appears to result in a consistent increase in the risk of both total and asymptomatic cardiotoxicity compared with the control arms (Table 2). The prevalence of cardiotoxicity, symptomatic heart failure, and trastuzumab discontinuation in the control and trastuzumab-treated arms of each of these trials is summarized in Table 2. All eight trials reported on the occurrence of congestive heart failure (CHF) as an outcome measure. The 2012 meta-analysis by Moja and colleagues of the over 10,000 patients included in the trials indicated a risk ratio (RR) for CHF of 5.11 [95% confidence interval (CI) 3.00–8.72, p < 0.0001] with trastuzumab treatment compared with controls. For CHF, heterogeneity between studies was minimal. However, definitions for significant decline in LVEF vary widely, resulting in a considerable amount of heterogeneity when LVEF decline was assessed in the seven trials in which it was reported. Consistent with findings for CHF, trastuzumab treatment also increased the risk for LVEF decline with an RR of 1.83 (90% CI 1.369–2.47, p = 0.0008) compared with the control group [Moja et al. 2012]. In this review, the trials focusing on neoadjuvant therapy (Budzar, NAOH, HERA) reported very low rates of symptomatic cardiotoxicity ranging from 0% to 1.7%, supporting tolerability of incorporation of trastuzumab into neoadjuvant treatment regimens. While these findings are informative, the results from randomized controlled trials include only patients meeting the stringent cardiac-related eligibility criteria prompted by original phase II and III trials and may not accurately reflect the prevalence of trastuzumab-induced cardiotoxicity in community practice where patients are not so carefully selected. Most trials excluded potential participants with an LVEF below 55%, older (> 65) patients, and patients with certain cardiac risk factors or with certain treatment experiences (e.g. mediastinal radiation therapy or hormone therapy) [Tan-Chiu et al. 2005; Slamon et al. 2011; Verma et al. 2010; Joensuu et al. 2006; Smith et al. 2007; Procter et al. 2010; Romond et al. 2005; Perez et al. 2008; Gianni et al. 2010; Spielmann et al. 2009].

Table 1.

Randomized clinical trials of trastuzumab.

Clinical trial	Reference	Study arms	Cardiotoxicity definition
B31	Tan-Chiu et al. [2005]	D + C → PT C → PT + H	LVEF decline >10% to <55%
BCIRG006	Slamon et al. [2011], Verma et al. [2010]	D + C → DT D + C → DT + H D + CP + H	LVEF decline >10%
FinHer	Joensuu et al. [2006]	Primary treatment (surgery, chemotherapy, or radiation therapy) → observation or H	LVEF decline >15% or >10% to <50%
HERA	Smith et al. [2007], Procter et al. [2010]	D + C → PT D + C → PT + H	LVEF decline ≥10% to <50%
N9831	Romond et al. [2005], Perez et al. [2008]	D + PT + C + M + FU D + PT + C + M + FU + H	Any change in LVEF from registration level
NOAH	Gianni et al. [2010]	Anthracycline-based chemotherapy Anthracycline-based chemotherapy + H Anthracycline-based chemotherapy + DT Anthracycline-based chemotherapy + DT + H	Asymptomatic LVEF decline to NYHA grade II or greater
PACS-04	Spielmann et al. [2009]	D + C → PT C → PT + H	LVEF decline to <50%

C, cyclophosphamide; CP, carboplatin; D, doxorubicin; DT, docetaxel; E, epirubicin; FU, fluorouracil; H, trastuzumab; LVEF, left ventricular ejection fraction; M, methotrexate; NYHA, New York Heart Association; PT, paclitaxel; V, vinorelbine.

Table 2.

Cardiotoxicity reported in control and trastuzumab arms of clinic trials.

Clinical trial	Reference	Control arm			Trastuzumab arm
		N	Total cardiotoxicity	Symptomatic heart failure	N	Total cardiotoxicity	Symptomatic heart failure	Trastuzumab discontinued*
B31	Tan-Chiu et al. [2005]	814	13 (1.6%)	5 (0.6%)	850	74 (8.7%)	31 (3.6%)	133 (15.6%)
BCIRG006	Slamon et al. [2011], Verma et al. [2010]	1050	115 (10.9%)	7 (0.7%)	2124	298 (14.0%)	25 (1.2%)	NR
FinHer	Joensuu et al. [2006]	116	4 (3.4%)	2 (1.7%)	117	7 (6.0%)	1 (0.8%)	NR
HERA	Smith et al. [2007], Procter et al. [2010]	1698	12 (0.7%)	3 (0.2%)	1703	97 (5.7%)	46 (2.7%)	72 (4.2%)
N9831	Romond et al. [2005], Perez et al. [2008]	664	81 (12.2%)	3 (0.4%)	1280	230 (18.0%)	38 (3.0%)	218 (17.0%)
NOAH	Gianni et al. [2010]	118	19 (16.1%)	1 (0.8%)	117	26 (22.2%)	4 (3.4%)	NR
PACS-04	Spielmann et al. [2009]	268	38 (14.2%)	11 (4.1%)	260	92 (35.4%)	37 (14.2%)	41 (15.8%)

^*Discontinued for cardiac-related reasons.

NR, not reported.

In addition to the results from randomized controlled clinical trials, several relatively recent observational studies have reported rates of trastuzumab-induced cardiotoxicity during trastuzumab treatment in the community practice setting, as summarized in Table 3. Studies of community practice tended to apply less stringent cardiac exclusion criteria than those applied in the randomized controlled trials described above. These studies offer an important perspective regarding the generalizability of clinical trial findings to the less rigorously selected, more heterogeneous population treated in the community. As with randomized controlled trials, the definition of cardiotoxicity and study sample size varied considerably between observational studies, potentially contributing to the large range of cardiac dysfunction observed. Importantly, rates of total cardiotoxicity, symptomatic heart failure, and trastuzumab discontinuation were similar in both randomized controlled trials and observational community practice-based studies. Cardiotoxicity occurred in 5.7–35.4% of patients enrolled in randomized controlled trials compared with a similar range of 8.3–43.6% in community practice. Rates of symptomatic heart failure were also similar, ranging from 0.8% to 14.2% in clinical trials and from 0% to 8.9% in community practice. Trastuzumab discontinuation was not reported in all studies, but when it was reported, it ranged from 4.2% to 17.0% in clinical trials and from 2.0% to 14.6% in community practice. Of course, the results of observational studies must be interpreted with caution and cannot be directly compared with findings from clinical trials, but do provide important information related to the incidence of cardiac dysfunction in an unselected patient population. It is reassuring that estimates of cardiotoxicity are similar in community practice to those reported in major clinical trials of trastuzumab, suggesting that the results are largely generalizable to community practice.

Table 3.

Prevalence of trastuzumab-induced cardiotoxicity reported in community practice-based observational studies.

Reference	Country	Patients	Cardiotoxicity definition	Total cardiotoxicity*	Symptomatic heart failure	Trastuzumab discontinued
Bowles et al. [2012]	USA	554	Heart failure or cardiomyopathy defined by ICD9-based algorithm¶	63 (11.4%)	NR	NR
Tarantini et al. [2012]	Italy	499	LVEF decline >10% or to <50%	133 (26.6%)	16 (3.2%)	9 (1.8%)
Naumann et al. [2013]	UK	388	LVEF decline ≥10% to <50% or Reduction in NYHA cardiac function	61 (15.7%)	0 (0%)	NR
Piotrowski et al. [2012]	Poland	253	LVEF decline >15% or ≥10% to <50%	52 (20.5%)	6 (2.4%)	7 (2.8%)
Lemieux et al. [2013]	Canada	237	LVEF decline ≥10% to <50% or any decline to <45%	32 (13.5%)	16 (6.7%)	18 (7.6%)
Farolfi et al. [2013]	Italy	179	LVEF decline ≥ 15% or LVEF <50%	78 (43.6%)	4 (2.2%)	14 (7.8%)
Cochet et al. [2011] ||	France	118	LVEF decline ≥10%	18 (15.2%)	0 (0%)	3 (2.5%)
Tanz et al. [2011]	Morocco	53	LVEF measurements, NOS	11 (20.7%)	2 (3.8%)	9 (17%)
Onitilo et al. [2012b]	USA	49	LVEF decline ≥15% or to <50%	14 (28.6%)	0 (0%)	1 (2.0%)
Dent et al. [2012] ‡	Canada	48	LVEF decline ≥ 10% or to <50%	NA	NA	7 (14.6%)
Serrano et al. [2012] §	Spain	45	LVEF decline ≥10% to <50% or >20%	12 (26.7%)	4 (8.9%)	3 (6.7%)

^*Includes both symptomatic and asymptomatic cardiotoxicity.

^‡Included only patients referred to cardiology department for concerns regarding treatment-related cardiotoxicity.

^§Included only patients ≥ 60 years of age.

^||Excluded patients with baseline LVEF <50% or known heart failure.

^¶ICD-9 code-based algorithm developed by Go et al. (2006) and validated by Allen et al. (2012)

LVEF, left ventricular ejection fraction; NA, not applicable; NOS, not otherwise specified; NR, not reported; NYHA, New York Heart Association.

Identification and modulation of risk factors for trastuzumab-induced cardiotoxicity

Several risk factors for trastuzumab-induced cardiotoxicity have been described, though none can completely predict occurrence (Table 4). Duration of trastuzumab administration has been explored as an obvious potential risk factor, but optimal duration has still not been determined. Current practice guidelines recommend 52 weeks of treatment and a clinical trial comparing the efficacy of 1 versus 2 years of therapy found that 2 years offered no additional clinical benefit, but did increase the risk of cardiotoxicity. Two relatively small clinical trials examined efficacy and cardiotoxicity of trastuzumab treatment regimens of 6 months or less [Buzdar et al. 2005; Joensuu et al. 2006]. When Moja and colleagues compared the rates of cardiotoxicity in these two trials with rates in the six other major trials of 12 months or more, it appeared that prolonged administration did increase the risk of LVEF decline, but further study is needed to determine whether shorter duration treatment is truly advisable [Moja et al. 2012]. More recently, Joensuu and colleagues directly compared receipt of 9 versus 52 weeks of trastuzumab in 45 and 131 patients respectively [Joensuu et al. 2014]. No difference in recurrence-free survival was noted, with a median follow up of 6.7 years and rates of cardiotoxicity were approximately equivalent at 2.2–2.3%. The study authors note that caution should be exercised when the results of this small study are interpreted, but suggest that further study of short duration trastuzumab as an effective therapy with reduced risk of cardiotoxicity is warranted [Joensuu et al. 2014].

Table 4.

Risk factors for trastuzumab-induced cardiotoxicity reported in the literature.

Reference	Study design	Patients	Risk factors for cardiotoxicity
Farolfi et al. [2013]	Retrospective chart review	179	Higher cumulative dose of anthracycline
Lemieux et al. [2013] *	Retrospective cohort study	132 73	Heavy alcohol use during treatment Ile655Val HER2 polymorphism
Russo et al. [2012]	Retrospective chart review	499	Older age Lower estimated glomerular filtration rate Higher baseline LVEF Doxorubicin exposure
Serrano et al. [2012] $	Retrospective chart review	45	Diabetes mellitus History of cardiac disease
Cochet et al. [2011]	Prospective	118	Impaired left ventricular diastolic function at baseline Higher cumulative dose of epirubicin
Perez et al. [2008]	Randomized controlled trial	1280	Older age Lower baseline LVEF Use of antihypertensive medication
Beauclair et al. [2007]	Prospective	61	Ile655Val HER2 polymorphism
Tan-Chiu et al. [2005]	Randomized controlled trial	850	Older age Higher baseline LVEF
Seidman et al. [2002] ‡	Prospective	1219	Anthracycline exposure Older age

^*Study included a total of 237 women, of whom 132 completed surveys regarding alcohol use during treatment and 73 underwent genetic testing for HER2 polymorphisms of interest.

^$Included only women ≥ 70 years of age.

^‡Cardiac Review and Evaluation Committee (CREC) assessment of cardiotoxicity in pivotal phase II/III trials of trastuzumab.

HER2, human epidermal growth factor receptor 2; LVEF, left ventricular ejection fraction.

Perhaps the most certain risk factor for trastuzumab-induced cardiotoxicity is the additive effect observed during concomitant anthracycline and trastuzumab administration [Seidman et al. 2002]. Risk can be lowered by sequential administration of anthracycline followed by trastuzumab, but previous anthracycline exposure also appears to increase the risk of trastuzumab-induced cardiotoxicity [Seidman et al. 2002; Farolfi et al. 2013; Russo et al. 2012; Cochet et al. 2011]. In the initial 2002 review of trastuzumab-induced cardiotoxicity in the original phase II and III clinical trials, the incidence of cardiotoxicity was highest in patients receiving concomitant trastuzumab and anthracycline plus cyclophosphamide at 27%. Rates were lower in patients who received a combination of paclitaxel and trastuzumab (13%) or trastuzumab alone (3–7%), although most received prior anthracycline therapy [Seidman et al. 2002]. Evaluation of the cumulative evidence suggested that the rate of cardiotoxicity and symptom severity and responsiveness to therapy was less severe with temporal separation of anthracycline and trastuzumab administration [Seidman et al. 2002]. These findings informed the design of future clinical trials, necessitating both a formal safety evaluation phase as well as avoidance of concomitant anthracycline and trastuzumab administration [Seidman et al. 2002; Moja et al. 2012]. Of note, in the BCIRG006 trial first published in 2009, the rates of cardiotoxicity were approximately equivalent in the control arm, which consisted of doxorubicin and cyclophosphamide followed by docetaxel, and one of the two trastuzumab-containing arms that did not include an athracycline (docetaxel, carboplatin, and trastuzumab), with 11% LVEF decline and 0.7% symptomatic heart failure in the control arm, and 9% LVEF decline and 0.4% symptomatic heart failure in the trastuzumab-containing arm. In contrast, the trastuzumab-containing arm that also contained an anthracycline had 19% LVEF decline and 2.0% symptomatic heart failure, clearly demonstrating the importance of anthracycline exposure as a risk factor for the development of trastuzumab-induced cardiotoxicity [Slamon et al. 2011; Verma et al. 2010].

More recent studies designed to specifically examine risk factors have supported the association between anthracycline exposure and trastuzumab-induced cardiotoxicity and have demonstrated a clear effect of cumulative anthracycline dose on risk [Farolfi et al. 2013; Russo et al. 2012; Cochet et al. 2011]. Farolfi and colleagues performed a retrospective study of 179 consecutive patients who started adjuvant trastuzumab between 2007 and 2010. Trastuzumab-induced cardiotoxicity occurred in 78 patients and a higher previous cumulative anthracycline dose (>240 mg/m² doxorubicin or >500 mg/m² epirubicin) was associated with a threefold increase in risk of cardiotoxicity [Farolfi et al. 2013]. Russo and colleagues observed a similar threefold increase in risk of trastuzumab-induced cardiotoxicity with doxorubicin exposure, but did not observe a significant association with epirubicin [Russo et al. 2012]. However, Cochet and colleagues examined only epirubicin at cumulative doses of 0, 300, or 600 mg/m² and found a direct relationship between risk of trastuzumab-induced cardiotoxicity and epirubicin dose increasing from 9% with no exposure to 35% at the highest dose [Cochet et al. 2011]. Interestingly, the recently published TRYPHAENA study demonstrated reasonable safety of concurrent or sequential administration of anthracycline and trastuzumab with the addition of the HER2-targeting monoclonal therapy pertuzumab [Schneeweiss et al. 2013]. However, in the absence of any difference in the primary endpoint of pathologic complete remission, this approach is not recommended as standard of care.

Anthracycline exposure results in cardiomyocyte apoptosis via production of ROS and other mechanisms. It is thought that trastuzumab exposure may worsen sublethal anthracycline-related damage in addition to causing HER2-specific damage and inhibition of cardiomyocyte repair [Farolfi et al. 2013] and a ‘dual-hit’ hypothesis has been proposed to explain the role of ROS and signal regulation in the case of trastuzumab-induced cardiac dysfunction potentiated by anthracycline exposure [Zeglinski et al. 2011]. Other factors known to be related to the production of ROS, besides anthracyclines, have also been identified as potential risk factors for trastuzumab-induced cardiotoxicity, including heavy alcohol use during treatment and comorbid diabetes mellitus [Lemieux et al. 2013; Serrano et al. 2012]. The relationship between alcohol consumption and trastuzumab-induced cardiotoxicity has been hypothesized to be related to effects on hypertension or general cardiac health [Lemieux et al. 2013], but it is possible that alcohol consumption may also increase underlying oxidative stress and promote ongoing cardiac damage in the context of anthracycline or trastuzumab exposure and ROS production [Bondia-Pons et al. 2012; Wu et al. 2006; Addolorato et al. 2008]. Similarly, oxidative stress has been suggested as a pathway important to the development of both diabetes mellitus and cancer [Onitilo et al., 2012a], and the chronic state of oxidative stress and inflammation that exists in the context of diabetes mellitus may also influence the risk of trastuzumab-induced cardiotoxicity [Onitilo et al. 2012a].

Older age has also been associated with an increased risk of trastuzumab-induced cardiotoxicity [Russo et al. 2012; Perez et al. 2008; Tan-Chiu et al. 2005], which is not particularly surprising considering that older age is also associated with increased risk of cardiovascular disease and other comorbidities in general. Additionally, several comorbidities associated with older age have been identified as risk factors for trastuzumab-induced cardiotoxicity, including diabetes mellitus [Serrano et al. 2012], decreased estimated glomerular filtration rate as an indicator of impaired kidney function [Russo et al. 2012], use of antihypertensive medications [Perez et al. 2008], and a history of heart disease [Serrano et al. 2012].

A history of heart disease is a clear predisposing factor for trastuzumab-induced cardiotoxicity, and the majority of clinical trials exclude such patients [Seidman et al. 2002; Serrano et al. 2012]. However, even patients with no history of diagnosed heart disease may have underlying left ventricular dysfunction, and two studies suggest that patients with impaired left ventricular dysfunction are at increased risk of trastuzumab-induced cardiotoxicity [Cochet et al. 2011; Perez et al. 2008]. Cochet and colleagues examined several measures of left ventricular function and found time to peak filling rate to be an independent predictor of trastuzumab-induced cardiotoxicity [Cochet et al. 2011]. Similarly, Perez and colleagues identified a low baseline LVEF as a risk factor for development of cardiotoxicity [Perez et al. 2008]. In contrast, a higher baseline LVEF was associated with increased risk of trastuzumab-induced cardiotoxicity in a 2012 study published by Russo and colleagues, but the authors suggest that this may have been caused by an overestimation of baseline LVEF in some women, resulting in a false diagnosis of trastuzumab-induced cardiotoxicity [Russo et al. 2012].

As none of the risk factors described above have been able to completely predict which patients are likely to develop trastuzumab-induced cardiotoxicity, additional risk prediction efforts are warranted. Methods for calculating a cardiac risk score in the context of trastuzumab therapy have been proposed, with one such equation based on age, LVEF, and use of antihypertensive medication [Rastogi et al. 2007]. However, risk score calculations have not been easily incorporated into standard guidelines or clinical practice. Clear definition of the mechanism of anthracycline-induced cardiotoxicity has allowed for significant progress in determining a pharmacogenomic risk profile using a candidate-gene approach. At present, there is significantly less known about the underlying mechanism of trastuzumab-induced cardiotoxicity, and as such, the candidate-gene approach has been much less fruitful [Jensen and McLeod, 2013]. However, two small studies have described HER2 polymorphism Ile655Val as a risk factor for trastuzumab-induced cardiotoxicity [Beauclair et al. 2007; Lemieux et al. 2013]. As progress is made toward a better understanding of the mechanisms responsible for trastuzumab-induced cardiotoxicity, molecular and genome-wide association studies are performed, and candidate genes are identified, it is likely that pharmacogenomic profiling of genetic risk may provide a more specific approach to identifying patients at high risk for trastuzumab-induced cardiotoxicity [Jensen and McLeod, 2013].

Based on the risks for trastuzumab-related cardiotoxicity described above, appropriate risk modification strategies include selection of chemotherapy regimens to avoid combining trastuzumab with cardiotoxic chemotherapy agents, such as anthracyclines. Also, because the cardiotoxic effect of trastuzumab is idiosyncratic and there is no cumulative dose reported at which this side effect occurs, trastuzumab should be administered for the minimum effective duration, which current recommendations suggest is a 1-year course. Careful consideration of appropriateness of trastuzumab therapy should occur for patients over 50 with previous cardiotoxic chemotherapy use, cardiac comorbidities, or a high body mass index. Strict monitoring for cardiotoxicity is also critical to ensuring patient safety while receiving trastuzumab.

Monitoring the safety of trastuzumab treatment

The goal of safety monitoring during trastuzumab administration is to detect cardiac dysfunction early, before the development of clinical symptoms, when therapy can be provided to prevent permanent changes. The trastuzumab package insert calls for monitoring of LVEF at baseline/pretreatment and at 3–4-month intervals throughout trastuzumab treatment, in combination with clinical evaluation for symptoms of heart failure as standard practice [Genetech, Inc., 2003]. Trastuzumab use should be avoided, or at least very carefully considered, in patients with baseline LVEF less than 50%. Trastuzumab should be held for at least 4 weeks in cases when LVEF declines more than 15% from baseline or more than 10% to below the institutionally defined normal value, which is usually 50–55%. Trastuzumab can be restarted when LVEF returns to within the normal limits following careful evaluation of risk and benefit by the provider and a full disclosure of risks to the patient.

Current treatment protocols call for measurement of LVEF by echocardiogram or multigated acquisition scan to evaluate the potential cardiotoxic effects of trastuzumab administration [Oreto et al. 2012]. Such techniques are limited, however, by the fact that significant myocardial damage must occur before changes in LVEF are observed, making measurement of LVEF an insensitive tool for the detection of early stage cardiotoxicity. In addition, definitions of cardiotoxicity based on LVEF or symptomology vary from study to study, which can make comparisons difficult. As such, several additional tools, including advanced imaging techniques and serum biomarkers, are currently under development for improved detection of trastuzumab-induced cardiotoxicity [Tocchetti et al. 2012; Braden et al. 2013].

Prevention of trastuzumab-induced cardiotoxicity and future directions

The potential for trastuzumab-induced cardiotoxicity is well-recognized. However, several areas of investigation regarding trastuzumab-induced cardiotoxicity require further attention, including identification of patients at high risk, early detection, prevention, and innovations in the use of trastuzumab. As described above, despite the identification of several risk factors for trastuzumab-induced cardiotoxicity, we are still unable to completely predict which patients are likely to develop trastuzumab-induced cardiotoxicity and better risk stratification measures are needed. Ongoing work to better understand the mechanisms responsible for trastuzumab-induced cardiotoxicity may pave the way for future genetic risk-profiling strategies [Jensen and McLeod, 2013]. When cardiotoxicity cannot be prevented by means of risk stratification, early detection becomes paramount to prevent irreversible cardiac damage. Several ongoing studies are focused on the investigation of advanced imaging techniques, such as echocardiogram with tissue velocity imaging, strain imaging, and cardiac MRI to better detect subclinical left ventricular dysfunction [Tocchetti et al. 2012]. In addition, several serum biomarkers of cardiotoxicity are also under investigation, including C-reactive protein, B-type natriuretic peptide, cardiac troponin I, interleukin 1 receptor-like 1, and neuregulin 1 [Braden et al. 2013].

There are currently several pharmaceutical options in use for the treatment and prevention of heart failure, including ACEIs, angiotensin receptor blockers (ARBs), and β blockers. Several studies have examined the use of these drugs as prophylaxis for the prevention of left ventricular dysfunction and heart failure in the context of anthracycline treatment with considerable success [Kalam and Marwick, 2013]. However, differences in the pathways by which anthracyclines and trastuzumab cause cardiotoxicity suggest that treatment strategies for the prevention of anthracycline-induced cardiotoxicity need to be evaluated separately for the prevention of trastuzumab-induced cardiotoxicity [Kalam and Marwick, 2013]. To date, only one study has been published examining pharmacological prevention of trastuzumab-induced cardiotoxicity. Seicean and colleagues retrospectively examined 920 consecutive patients with early stage breast cancer receiving adjuvant trastuzumab and anthracycline [Seicean et al. 2013]. Patients who received continuous β blocker treatment during cancer treatment (n = 106) were matched with 212 patients who did not receive β blocker treatment and admissions for heart failure were compared between the two groups. A fivefold reduction in heart failure events was observed in patients treated with β blockers, providing preliminary evidence to suggest that β blockers may be valuable as a prophylaxis to prevent trastuzumab-induced cardiotoxicity in addition to the previous demonstrated role in the prevention of anthracycline-induced cardiotoxicity [Seicean et al. 2013; Kalam and Marwick, 2013]. The Prevention of Cardiac Dysfunction During Adjuvant Breast Cancer Therapy (PRADA) study is a randomized, placebo-controlled, double-blind trial that is currently underway to assess the use of ARBs, β blockers, or their combination to prevent the development of left ventricular dysfunction in patients receiving standard adjuvant treatment for early breast cancer and will provide important information regarding the use of pharmacologic prophylaxis for the prevention of trastuzumab-induced cardiotoxicity [Heck et al. 2012].

In addition to ongoing research regarding risk stratification, early detection, and prevention of cardiotoxicity, continuous monitoring of the potential cardiotoxic effects of trastuzumab is necessary as advances and innovations are made in the way trastuzumab is used. For example, very little is known about the risk of cardiotoxicity in the context of trastuzumab monotherapy as most patients are treated with additional chemotherapeutic agents either before or during treatment with trastuzumab. A recently described ongoing randomized controlled trial seeks to determine both the safety and efficacy of trastuzumab monotherapy in older patients with HER2-positive breast cancer treated with postoperative adjuvant trastuzumab and will likely provide important data regarding cardiotoxicity with trastuzumab monotherapy [Sawaki et al. 2011]. Additionally, little is known about the incidence of trastuzumab-induced cardiotoxicity in male breast patients or patients cancer treated with trastuzumab for gastroesophageal malignancies. Despite the use of trastuzumab for the treatment of HER2-positive breast cancer for many years, only one report of the potential benefit for the treatment of men with breast cancer has been published, and safety data are nonexistent [Carmona-Bayonas, 2007; Onami et al. 2012]. Similarly, trastuzumab was approved by the US Food and Drug Administration for the treatment of malignant, HER2-positive, gastroesophageal malignancies in 2010 [Toomey et al. 2013], yet cardiotoxicity data specific to trastuzumab use in this context have not been published.

Additional innovations in the way trastuzumab is used are coming in the form of chemical modification and route of administration. For example, trastuzumab emtansine is an antibody–drug conjugate under development for the treatment of HER2-positive breast cancer that combines the HER2-targeted monoclonal antibody trastuzumab with the cytotoxic drug mertansine [Diéras and Bachelot, 2013]. In the open-label phase I–III clinical trials performed to date, cardiotoxicity has not been reported in association with trastuzumab emtansine [Diéras and Bachelot, 2013]. Additionally, a subcutaneous form of trastuzumab has been developed to overcome the challenges associated with intravenous administration and has been found in early clinical trials to provide a dosing and safety profile similar to that found with intravenous administration [Bittner et al. 2012; Ismael et al. 2012].

Finally, the success of HER2-targeted therapy has led to the development of other novel HER2-specific agents, including the well studied tyrosine kinase inhibitor lapatinib and other newer targeted drug therapies such as pazopanib, afatinib, and neratinib, which may provide potential alternatives for patients when HER2-targeted therapy is indicated, but trastuzumab is not well-tolerated [Incorvati et al. 2013; Sendur et al. 2013]. Continuous monitoring for cardiotoxicity will remain an important feature of ongoing studies as these and other innovations in the use of trastuzumab and other HER2-targeted therapies are tested further.

Conclusion

Adjuvant trastuzumab is an important part of the treatment regimen for many women with HER2-positive breast cancer and has undoubtedly resulted in a significant improvement in the prognosis of HER2-positive breast cancers. However, the risk for trastuzumab-induced cardiotoxicity is an important consideration. Understanding risk factors for trastuzumab-induced cardiotoxicity and appropriate patient monitoring during trastuzumab treatment allows for safe and effective use of this important adjuvant therapy.