Long-term outcomes of prophylactic cardioprotection in early-stage breast cancer patients on non-anthracycline anti-HER2 therapy

Abstract

Anti-HER2 therapy has improved survival in HER2-positive breast cancer patients for decades but is associated with well-established cardiotoxicity. Our prior randomized controlled trial PROTECT HER2 found no short-term benefit of prophylactic beta-blocker cardio protection (CPT) based on changes in global longitudinal strain (GLS) in a breast cancer population treated with non-anthracycline anti-HER2 therapies. The long-term cardiovascular (CV) effects of CPT remain unknown. Utilizing the PROTECT HER2 cohort with 5–10 years follow-up, we performed a post hoc analysis to assess long-term outcomes. We evaluated the incidence of major adverse cardiovascular events (MACE; defined as all-cause mortality, stroke, myocardial infarction (MI), heart failure hospitalization (HFH), and significant arrhythmias) based on CPT. We also assessed associations between CV risk factors (hypertension, hyperlipidemia, diabetes and BMI≥25 kg/m²) and the occurrence of MACE. Among 109 patients the mean (SD) age was 51 (11) years, and the median (IQR) follow up was 7 (6–9) years post-RCT. Eight (7.3%) patients developed a total of 10 events, including 3 (2.8%) non-CV mortalities, 1 (0.9%) non-fatal stroke, and 6 (5.5%) arrhythmias. There were no significant associations observed between CV risk factors and MACE overall. However, 40% of patients (44/109) developed at least one new CV risk factor during follow-up. These findings demonstrate that while prophylactic CPT did not significantly impact long-term CV outcomes, a substantial proportion developed new risk factors over time, highlighting an important area for longitudinal risk monitoring and intervention.

Introduction

Cardiotoxicity remains a clinically significant concern in patients receiving anti-human epidermal growth factor receptor (anti-HER2) therapy for breast cancer, despite well-documented improvement in disease-free and overall survival [1]. Cardiotoxicity from anti-HER2 therapy is characterized from early subclinical changes in global longitudinal strain (GLS) to asymptomatic left ventricular systolic dysfunction or heart failure [2]. In response to cardiotoxicity from antiHER2 related therapies, cardioprotective therapy (CPT) has been proposed as a strategy to mitigate cancer related-therapy cardiac dysfunction (CTRCD) [3]. CPT refers to the prophylactic use of a guideline-directed cardiovascular medication, most commonly beta-blockers and renin-angiotensin system inhibitors to prevent or mitigate cancer therapy-related cardiotoxicity [4]. Numerous trials have been conducted to evaluate if cardioprotective therapies may mitigate these cardiotoxic effects, but most of these trials have been in the context of anthracyclines or sequential therapies with anthracyclines and antiHER2 therapies [3, 5, 6]. The data from these trials have been mixed and is strongest for breast cancer patients receiving sequential therapy with anthracycline followed by antiHER2 therapies [5]. PROTECT HER2 was the first randomized controlled trial (RCT) of early-stage breast cancer patients to evaluate CPT for non-anthracycline anti-HER2 therapies, and further risk stratify patients on these treatment regimens with GLS [7]. The findings from PROTECT HER2 results did not demonstrate any short-term benefit of prophylactic beta-blocker cardioprotection (CPT) in GLS stability, cancer-therapy related cardiac dysfunction (CTRCD), or cancer therapies held [7].

The long-term cardiovascular (CV) outcomes associated with prophylactic beta-blocker use in this population remain unknown. To inform cardiovascular risk stratification, guide long-term surveillance, and management of these patients post breast cancer treatment, we performed a post hoc analysis of the PROTECT HER2 RCT up to 9 years of follow-up. Our aims were (1) to evaluate the incidence of major adverse cardiovascular events (MACE), and (2) to assess whether CV risk factors were associated with the occurrence of MACE during the follow-up period.

Methods

PROTECT HER2 was a single-center prospective randomized clinical trial approved by the institutional review board of Northwestern University (NCT02993198). The original study participants included 110 female patients aged > 18 years with a new diagnosis of HER2-positive early stage (stage I-III) breast cancer. All participants were recruited for the clinical trial at Northwestern Memorial Hospital, Chicago, Illinois between June 2015 and February 2020. Participants were actively followed up for one year after initiation of anti-HER2 therapy, with longitudinal data collected from the electronic medical record after cessation of the RCT. The design and methods of trial, including eligibility, inclusion and exclusion criteria have been described previously [7]. Baseline echocardiography was performed in all patients. Upon initiation of anti-HER 2 therapies, patients were grouped into four arms based on left ventricular ejection fraction (LVEF) and GLS assessed by echocardiography. Arm A included patients with normal LVEF and normal GLS throughout the study period. Arms B and C consisted of patients with normal LVEF and reduced GLS, who were randomized in a 1:1 ratio to receive carvedilol (Arm B) or standard-of-care monitoring without carvedilol (Arm C). Arm D comprised of patients who had reduced LVEF. Normal LVEF was defined as LVEF ≥ 53% throughout the study period while reduced GLS was defined as > 15% decrease from baseline GLS or absolute GLS < 15% strain units. Patients were followed up every three months for 12 months, or for the duration of the anti-HER2 therapy, with serial transthoracic echocardiography (TTE). Patients initially assigned to Arms A, B, or C who developed cancer therapy-related cardiac dysfunction (CTRCD) during follow up crossed over to arm D and were initiated on guideline-directed medical therapy (GDMT). One patient who experienced non-fatal stroke during trial was excluded from the post-RCT analysis (Fig. 1).

Fig. 1.

CONSORT diagram of patient enrollment, treatment arm assignment, and post-RCT outcomes

Baseline CV risk factors were defined as the presence of any of the following conditions prior to anti-HER 2 therapies: elevated body mass index (BMI ≥ 25 kg/m²), hyperlipidemia, hypertension, diabetes mellitus or use of any cardiovascular medication (including ACE inhibitors, angiotensin receptor blockers (ARB), calcium channel blockers, diuretics or statin therapy). The primary outcome was incidence of MACE across groups. MACE was defined as mortality (all-cause and CV-related mortality), non-fatal stroke, non-fatal myocardial infarction (MI), heart failure hospitalizations (HFH), and development of arrhythmias. Other survivorship variables collected during follow-up included breast cancer recurrence, receipt of additional chemotherapy, endocrine therapy, radiation, and subsequent HER-2 directed therapies. An additional outcome was development of other CV risk factors, defined as an increase in BMI of ≥ 3 kg/m² to a peak of ≥ 25 kg/m², new diagnosis of hyperlipidemia, hypertension, diabetes mellitus, and initiation of new CV medication prescription. Among patients randomized to carvedilol (Arm B), duration of therapy and discontinuation during follow-up were recorded. Additionally, subsequent use of ACE inhibitors or ARBS following cessation of carvedilol, or completion of anti-HER2 therapy, was noted. Finally, echocardiographic data obtained after the end of the trial was collected for each patient in search of cancer therapy-related cardiac dysfunction (defined as a new LVEF decline of > 10% to a nadir of < 50%). Baseline characteristics were summarized using count and percentages for categorical variables and mean (standard deviation) for continuous variables and were compared between groups using ANOVA or Fisher’s exact tests. Cumulative incidence of MACE was estimated using the Kaplan-Meier function (Fig. 2). Associations between CV risk factors and MACE were assessed using Fisher’s exact test. All statistical tests were two-sided and a p-value of < 0.05 was considered statistically significant. Analyses were performed using SAS (Base 9.4).

Fig. 2.

Kaplan-Meier curves for MACE-free survival (A) and overall survival (B)

Results

A total of 109 women (mean age 51yrs, SD 11) from the PROTECT HER2 trial were included. Of these, 83 patients were assigned to arm A (normal LVEF and GLS), 10 patients were randomized to arm B (normal LVEF, reduced GLS, on CPT), 10 patients were randomized to Arm C (normal LVEF, reduced GLS, standard care), and 6 patients were assigned to Arm D (reduced LVEF). The median (IQR) length of follow-up was 7 (6–9) years with a range of 5 to 10 years. Baseline CV risk factors including hypertension, hyperlipidemia and diabetes mellitus were lower in arm D compared with the other arms. Baseline use of CV medications was low and comparable across groups (Table 1).

Table 1.

Baseline clinical characteristics of study participants (N = 109)

Characteristics, No. (%)	Assigned RCT Arm
	Arm A N = 83	Arms (B + C) N = 20	Arm D N = 6	p-value*
Age, years, mean (SD)	51 (12)	55(7)	44(9)	0.10
Race
White	59 (71.1)	16 (80.0)	1 (16.7)	0.028
Black	14 (16.9)	3 (15.0)	2 (33.3)
Asian	7 (8.4)	0 (0)	2 (33.3)
Other	3 (3.6)	0 (0)	1 (16.7)
Cardiovascular Risk Factors
BMI, mean (SD)	26.1 (5.7)	27.3 (6.3)	22.8 (4.4)	0.26
Elevated BMI (≥ 25 kg/m2)	44 (53.0)	10 (50.0)	1 (16.7)	0.26
Hyperlipidemia	8 (9.6)	2 (10.0)	0 (0)	1.00
Hypertension	10 (12.0)	2 (10.0)	0 (0)	1.00
Diabetes Mellitus	6 (7.2)	0 (0)	0 (0)	0.71
Cardiovascular Medications
Beta-blockers	1 (1.2)	0 (0)	0 (0)	1.00
CCB	6 (7.2)	2 (10.0)	0 (0)	0.78
Diuretics	4 (4.8)	1 (5.0)	0 (0)	1.00
Statin therapy	2 (2.4)	2 (10.0)	0 (0)	0.34
Patients With ≥ 1 CV Risk Factor	48 (57.8)	12 (60.0)	1 (16.7)	0.17

Categorical variables are represented by counts and percentages. Continuous variables are represented by mean (SD)

ACEi Indicates Angiotensin converting enzyme inhibitors, ARB Angiotensin receptor blocker, BMI Body mass index, CCB Calcium channel blockers

*Based on ANOVA and Fisher’s Exact Test

Among all 109 patients, 10 total events occurred within 8 (7.3%) patients. Three (2.7%) mortalities from non-CV related causes, one (0.9%) non-fatal strokes, and 6 (5.5%) new diagnoses of arrhythmia were observed. Specifically, patients in Arm A experienced 2 mortalities, 1 stroke, and 3 arrhythmias. Arms B and C saw 1 atrial fibrillation. Arm D patients experienced 1 mortality and 2 arrhythmias. No non-fatal MI, CV-related deaths, or heart failure hospitalizations (HFH) were observed. Notably, in the two randomized arms of interest (Arm B: strain-triggered CPT vs. Arm C: strain-triggered observation), no deaths or stroke occurred in either subgroup (Fig. 1). Additionally, among the 10 patients who were randomized to carvedilol, the median duration of therapy was 285 days following cessation of their involvement in the trial, and 1 patient continued carvedilol to present day. Following discontinuation of carvedilol or completion of anti-HER2 therapy, only 1 (10%) patient was transitioned to an ACE inhibitor or ARB to continue cardioprotection. No patients from Arms A, C, or D were started on carvedilol during the follow-up period, although 5 patients from Arm A did begin other beta-blockers (3 metoprolol, 1 atenolol, 1 propranolol).

A large proportion of the cohort developed cardiovascular risk factors during follow-up, with 40% of patients (44/109) diagnosed with at least one risk factor that had not been present during the trial time period. 10% (11/109) of patients were diagnosed with hypertension, 10% (11/109) with hyperlipidemia, 2.7% (3/109) with type 2 diabetes mellitus, and 14% (15/109) with an increased BMI (average increase of 5.1 kg/m²). 26% of patients (28/109) initiated ≥ 1 new prescription of a cardiovascular medication (including ACE inhibitors, angiotensin receptor blockers (ARB), calcium channel blockers, diuretics or statin therapy). There were no significant associations observed between development of CV risk factors and MACE overall (Table 2). During follow-up, 7/109 (6.4%) of patients experienced breast cancer recurrence, all of whom received additional cancer-directed therapy, including chemotherapy (3 patients), radiation (2 patients), endocrine therapy (2 patients), and/or subsequent HER-2 directed therapy (6 patients). 53 of the patients had at least one post-RCT echocardiogram during follow-up, as echocardiograms after trial completion were obtained at the discretion of each patient’s care team. Among these patients, no CTRCD was observed when referencing baseline LVEF from the original RCT. Of note, all 6 patients in Arm D who developed CTRCD during the trial demonstrated rapid LVEF recovery by the subsequent echocardiogram at a 3-month interval. None went on to re-develop CTRCD in subsequent echocardiographic assessments.

Table 2.

Association of newly developed cardiovascular risk factors with MACE

Cardiovascular Risk Factors, No (%)	Overall N = 109	Occurrence of Cardiac Event
		No n = 101	Yes n = 8	p-value*
Increase in Body mass index***	15 (13.8)	14 (13.9)	1 (12.5)	1.00
Hyperlipidemia	11 (10.1)	11 (10.9)	0 (0)	1.00
Hypertension	11 (10.1)	10 (9.9)	1 (12.5)	0.59
Diabetes Mellitus	3 (2.8)	2 (2.0)	1 (12.5)	0.21
New CV Medication Prescription**	28 (25.7)	25 (24.8)	3 (37.5)	0.42
Patients with ≥ 1 CV Risk Factor	44 (40.4)	40 (39.6)	4 (50.0)	0.71

*Based on Fisher’s Exact Test. ** CV medications include Angiotensin converting enzyme inhibitors (ACEi), Angiotensin receptor blocker (ARB), Betablockers, Calcium channel blockers,(CCB), Diuretics and Statin therapy. *** Defined as an increase in BMI of ≥3 kg/m2 to a peak of ≥25 kg/m2

Discussion

In the observational post hoc follow-up of participants from the PROTECT HER2 RCT, prophylactic administration of beta-blocker for CPT was not associated with differences in long-term cardiovascular outcomes. During long-term observational follow up, the overall survival at 8 years was 95.2% (95% CI: 87.9–99.2%). Similarly, the 5-year cumulative incidence of MACE was low at 5.6% (95% CI : 2.0 -10.7%), and the 8-year cumulative incidence of MACE was 9.7% (95% CI : 3.8–17.9%), which is roughly in-line with the 10-year risk in the general US adult population [8]. Prophylactic CPT was also not associated with other CV benefits, such as lower rates of arrhythmia or hypertension. Indeed, 9 of 10 patients (90%) randomized to carvedilol discontinued therapy after cessation of their portion of the trial, with a median exposure of 650 total days, which likely limits our ability to detect a durable cardioprotective effect. Therefore, any potential protective effects may not persist, particularly for outcomes such as hypertension or arrhythmia which develop over longer periods of time. In fact, this patient cohort developed high rates of new hypertension during the follow-up period. This data suggests that, in a relatively low-cardiac-risk population treated with non-anthracycline-based anti-HER2 regimens, routine administration of prophylactic CPT medication may not provide long-term advantage.

These results build upon existing evidence suggesting that cardiotoxicity from trastuzumab in early-stage HER2-positive patients is often transient [9, 10]. For example, in one 7-year study, 84% of trastuzumab-related drops in LVEF recovered and only 2% of patients experienced a new cardiac event after therapy [11]. These findings suggests that patients without notable risk factors rarely develop cardiac pathology from trastuzumab alone, which may partly explain why we did not observe a statistically significant long term clinical benefit of prophylactic CPT despite changes in GLS. Indeed, other studies have also found modest or mixed results for preventative cardioprotection, with clear utility only seen in high risk settings [12, 13]. Therefore, given the insufficient evidence, expert consensus does not endorse routine cardioprotective therapy for all patients on antiHER2 therapies [14]. Our findings align with this, as we did not observe a statistically significant long-term benefit of prophylactic CPT in this low-risk cohort. However, given the small sample size and limited number of clinical events, the absence of statistically significant differences should not be interpreted as evidence that prophylactic CPT is ineffective. Rather, these findings should be interpreted as hypothesis-generating or exploratory to support the need for larger powered studies to identify which patients are most likely to benefit from CPT.

Importantly, our study also demonstrated increased rates of CV risk factor development in this patient cohort. Notably, 7 patients (6.4%) experienced breast cancer recurrence during follow-up which may have independently contributed to the development of new CV risk factors, irrespective of CPT status. We observed an incidence of 40% of hypertension, hyperlipidemia, weight gain, new CV-related prescription, and/or diabetes over the median 7-years follow-up. These data are consistent with findings in other breast cancer survivor cohorts. One study of women completing HER2 therapy reported that, over roughly 8 years, 33% of participants developed hypertension and 58% were diagnosed with hypercholesterolemia [15]. In addition, epidemiologic studies indicate that breast cancer survivors face an elevated risk of cardiovascular disease and events compared to the general population, especially when compared longitudinally [16]. Thus, a preventative approach to cardiac health and close surveillance of risk factors are paramount to survivors’ care. Although prophylactic CPT was not associated with improved MACE outcomes in this cohort, optimization of modifiable CV risk factors represents an important area for future investigation in reducing long-term CV morbidity among HER2-positive breast cancer survivors. These findings highlight the need for continued evaluation of CV prevention strategies in patients receiving newer, less cardiotoxic cancer therapies.

This study has several limitations. First, our sample size was small, particularly in the randomized carvedilol treatment arms (Groups B and C), with very few clinical events. Consequently, the study may be substantially underpowered to detect modest treatment effects, and failure to detect statistically significant differences may not translate to absence of treatment benefits. Secondly, the retrospective chart-review design of the post-hoc analysis, introduces the potential for bias and incomplete data, particularly regarding follow-up echocardiographic data. Thirdly, all patients were treated at a single center, Northwestern Memorial Hospital, limiting the variance and generalizability of the sample size. Additionally, survivorship-related factors such as breast cancer recurrence and receipt of additional cardiotoxic therapies (chemotherapy, radiation, endocrine therapy, and/or subsequent HER2-directed agents) were not adjusted for in our analysis given the very small number of events. These may represent unmeasured confounders influencing long-term CV risk independent of CPT. Furthermore, although participants were derived from an RCT, this study represents an observational post hoc follow up after completion of a randomized CPT intervention, thus continuation of carvedilol and subsequent CV management after trial period were not protocolized, thereby limiting causal inference regarding long term effects of prophylactic CPT. Finally, development of some new cardiovascular risk factors may reflect a normal aging process given the long period of follow-up – this must be considered when interpreting findings.

These findings support a risk-tailored approach for breast cancer patients. Patients receiving non-anthracycline regimens for early-stage HER2-positive breast cancer appear to be at low risk for MACE over a 7 year follow-up. In this population, we did not observe a statistically significant long-term benefit of prophylactic CPT despite changes in GLS. These findings align with current guidelines that do not recommend universal cardioprotection in low-risk cases, as well as the findings of the original PROTECT HER2 trial [7, 14]. Rather, clinicians’ focus in low-risk cases should be on survivorship care and addressing the development of modifiable CV risk factors. All breast cancer survivors, regardless of cardiotoxicity risk, should be counseled for the development of cardiovascular risk factors, need for screening and advised on prevention strategies.

Abbreviations

HER2

Human epidermal growth factor receptor2

LVEF

Left ventricular ejection fraction

CPT

Cardioprotection

HFH

Heart failure hospitalization

Authors' contributions

TM & IB were major contributors to writing the manuscript, and TM collected the data. MZ & AB performed the statistical analysis. NA conceived of the study design, analyzed the data, and revised the manuscript. All authors read and approved the final manuscript.

Funding

Not applicable.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due privacy concerns but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Northwestern University IRB approved this study (STU00200675). Consent was waived for the post-hoc analysis but obtained for the PROTECT HER2 cohort.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.