Factors Associated With the Recovery of Left Ventricular Ejection Fraction in Patients With Anthracycline-Induced Left Ventricular Dysfunction

Abstract

Background

Neurohormonal blocking drugs, like beta-blockers, angiotensin-converting enzyme inhibitors (ACEIs), and angiotensin receptor blockers (ARBs), are recommended for treating anthracycline-induced left ventricular dysfunction (AILVD). However, there is limited evidence supporting their benefit. Therefore, this study evaluated associations of neurohormonal blockers and other clinical factors with recovery of left ventricular ejection fraction (LVEF) in patients with AILVD.

Methods

This retrospective chart review assessed patients treated with at least one dose of anthracycline, then had ≥10% LVEF reduction or post-anthracycline LVEF value <50%, and then had a follow-up LVEF measurement ≥90 days later. The primary endpoint was LVEF recovery (highest follow-up LVEF−lowest LVEF post-anthracycline). Variables from univariable tests with P < .1 were incorporated in a multiple linear regression model for independent factors significantly associated with LVEF recovery (P < .05).

Results

Out of 104 patients, 83% were female, 86% self-reported white race, 53% had breast cancer, median (IQR) age was 52 (22) years, and LVEF recovery was 14% (16%). The final multivariable model included 2 significant variables: beta-blocker dose after anthracycline exposure (every 25 mg increase in beta-blocker dose was associated with 5.0% increase in LVEF recovery; P = .0005) and the time between the start of the anthracycline and the lowest LVEF post-anthracycline (every 5-year increase in time was associated with 1.8% decrease in LVEF recovery; P = .0379).

Conclusions

In patients with AILVD, a higher beta-blocker dose and earlier detection of LVEF reduction post-anthracycline were significantly and independently associated with improved LVEF recovery. These findings need to be validated in a larger, independent cohort.

Keywords

anthracycline-induced cardiomyopathy anthracycline-induced left ventricular dysfunction‌left ventricular ejection fraction beta-blocker angiotensin-converting enzyme inhibitors angiotensin receptor blockers aldosterone antagonist

Introduction

Since the 1970s, 5-year cancer survival rates have continued to rise and are currently approaching 70%.¹ With more cancer patients surviving, long-term complications of common chemotherapy regimens have come to light. One class of commonly used antineoplastic medications of particular concern are anthracyclines (eg, doxorubicin, daunorubicin, epirubicin, and idarubicin). Anthracyclines are potent chemotherapeutic agents, and a central component of treatment regimens for multiple cancers, including breast cancer, sarcoma, and hematological malignancies.^2–4 Unfortunately, these compounds exhibit cumulative dose-dependent cardiotoxicity associated with left ventricular dysfunction and heart failure (HF).^5,6 Indeed, clinical HF rates secondary to anthracycline exposure have been shown to range from 5% at a lifetime dose of 400 mg/m² to 48% at 700 mg/m².⁶ More than 2 million patients are at risk of developing anthracycline-induced left ventricular dysfunction (AILVD) in the United States alone.⁷

Left ventricular ejection fraction (LVEF) is the most commonly used parameter to monitor systolic dysfunction caused by AILVD. AILVD is defined by a drop in LVEF >10% from baseline to a value <50%.⁸ Historically, a reduction in LVEF following anthracycline exposure was thought to be irreversible.^9,10 However, more recent studies have shown that, with early treatment, it is possible for left ventricular function to return to pre-AILVD levels.^11,12 Unfortunately, even with early treatment, a large proportion of patients either have incomplete or no LVEF recovery.^8,12 Several clinical factors influencing AILVD recovery have been identified. The most prominent include time-to-HF-treatment and the New York Heart Association (NYHA) functional class.¹²

The optimal pharmacologic therapies for AILVD prevention and treatment have not been established. Clinical trials assessing the efficacy of neurohormonal blockers, including beta-adrenergic receptor blockers, angiotensin-converting enzyme inhibitors (ACEIs), and angiotensin receptor blockers (ARBs), in preventing AILVD have produced inconsistent results.^13–19 Limited clinical trials suggest that beta-blockers and ACEIs/ARBs may improve LVEF in the AILVD treatment setting, but these trials were relatively small and observational in nature.^8,12 Less is known about the efficacy of neurohormonal blockers for AILVD prevention and treatment in the real-world clinical setting. Accordingly, the objectives of this study were to (1) evaluate the real-world associations of beta-blockers and ACEIs/ARBs in promoting LVEF recovery in patients with AILVD and (2) identify demographic and clinical factors associated with LVEF recovery during neurohormonal blocker therapy.

Methods

Patient Eligibility

This study was a retrospective chart review of eligible patients within the University of Michigan Health System (UMHS, or “Michigan Medicine”) located in Ann Arbor, Michigan, USA. The patient sample met the following criteria: (1) had been treated with at least one dose of anthracycline for any type of breast cancer, leukemia, lymphoma, or sarcoma; (2) had a documented decrease in LVEF by ≥ 10% after anthracycline exposure compared to the pre-treatment LVEF, or, in the absence of a pre-treatment LVEF, any documented LVEF < 50% post-anthracycline; (3) had a follow-up LVEF assessment ≥ 90 days since the lowest LVEF using the same method of measurement (eg, echocardiography). Patients were excluded if they had a diagnosis of cardiomyopathy prior to administration of their first anthracycline dose, or if they received any treatment with dexrazoxane, the only FDA-approved drug to prevent anthracycline-induced cardiotoxicity.²⁰ This study was approved by, and performed in accordance with, the University of Michigan's investigation review board, IRBMED. As this was secondary analysis of existing clinical data, the IRB granted a waiver of informed consent.

Study Design and Data Collection

Demographic, clinical, and medication prescription data were collected as follows. Electronic health records (EHRs) for each patient were manually reviewed by 2 independent clinical study team members. Study data were collected and managed using REDCap (Research Electronic Data Capture) hosted at the University of Michigan.^21,22 REDCap is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources. In addition to manual chart review, UMHS’ Electronic Medical Record Search Engine (EMERSE)²³ was utilized to perform term-based queries of clinicians’ notes. Chart review was supplemented by the University of Michigan Data Office for Clinical and Translational Research (DOCTR) which provided data directly from the EHR regarding medication orders, vital and laboratory values, and LVEF measurements. All discrepancies within the collected data sets were analyzed, and disagreements were adjudicated by a third study team member as needed. Collected data consisted of age, sex, race, smoking status (ie, current & former smokers = “yes” and never smokers = “no”), the oncologic diagnosis for which an anthracycline was ordered, relevant comorbid disease states (including coronary artery disease, HF, cardiac valvular dysfunction, diabetes mellitus, hypercholesterolemia, chronic kidney disease, and hypertension), and radiation exposure to the heart. Diagnosis for any of these comorbidities was determined by the presence or absence in patients’ problem lists. Any condition for which there was no documentation in the patients’ medical records was considered not to be present. All baseline characteristics and relevant vital and laboratory values (eg, body mass index, systolic and diastolic blood pressure, mean arterial pressure, pulse, vital status, QRS duration, serum sodium, troponin levels, estimated glomerular filtration rate, and B-type natriuretic peptide) were collected within 6 months of the date of lowest LVEF measurement. Data was available in the EHR starting in 2012 and was collected through May 2021.

Baseline LVEF was defined as the most recent LVEF collected prior to the initiation of anthracycline, if available. The lowest documented LVEF any time after exposure to anthracycline was defined as the lowest LVEF. The subsequent highest documented LVEF ≥90 days from the one defined as the lowest, and measured using the same method, was selected as the highest LVEF. In patients without a documented baseline LVEF (31 out of 104 patients [28.9%]), the baseline LVEF was assumed to be normal (ie, 60%) since the patients did not have a previous diagnosis of cardiomyopathy.

All anthracycline doses were converted to doxorubicin equivalents based on the following conversion: 50 mg/m² doxorubicin = 60 mg/m² daunorubicin = 75 mg/m² epirubicin = 100 mg/m² zorubicin = 10 mg/m² idarubicin = 12.5 mg/m² mitoxantrone.²⁴ Duration of anthracycline therapy was calculated by subtracting the end date available in the EHR from the start date. All beta-blocker doses were converted to carvedilol equivalents based on the following conversion: 1 mg carvedilol = 4 mg metoprolol = 2 mg atenolol = 0.2 mg bisoprolol = 1.6 mg propranolol. Angiotensin inhibitors (ACEI, ARB, or ARNI) were standardized into dose equivalents by the percentage of the target dose used in HFrEF clinical trials, or for angiotensin inhibitors not tested in HFrEF clinical trials, by the maximum daily dose (Supplemental Table 1).²⁵ Target doses were defined as those recommended in HF treatment guidelines.²⁶

The heart exposure to radiation therapy was categorized into 4 groups: no exposure, scatter exposure (est <0.5 Gy), indirect exposure near heart exposure (est 0.5-5 Gy), and partial direct exposure (est >5 Gy).

Statistical Analysis

Categorical variables are described as counts and percentages, and continuous variables are described by median (IQR). The primary outcome was LVEF recovery (highest LVEF−lowest LVEF; see definitions above). Based on Kolmogorov–Smirnov test with P > .05 and visual inspection of distribution plots, LVEF recovery was normally distributed. Therefore, univariable associations of collected variables with the primary outcome were assessed using simple linear regression for continuous variables, t-test for categorical variables with 2 groups, and ANOVA for categorical variables with > 2 groups. The parameter estimate for LVEF recovery from the linear regression model is displayed for continuous variables, and the mean +/− SD LVEF recovery is displayed for each subgroup of categorical variables. Values that met a significance threshold of P < .1 in univariable analyses were selected for entry into a multiple linear regression model, except for serum sodium because it was missing in 32% of patients. Stepwise selection was then employed to remove the variables with P ≥ .05 from the multivariable model. The final model was achieved once all variables remaining in the model were statistically significant at P < .05. Multivariable analysis was then performed using all significant and independent variables. The univariable analyses had approximately 80% power to detect 3.5% recovery in LVEF. All statistical analyses were performed using SAS version 9.4 (Cary, NC).

Results

Descriptive Statistics

A total of 104 patients were eligible for analysis. Tables 1 to 6 present the descriptive statistics for the overall patient sample and the univariable analyses for association with the primary outcome of LVEF recovery (highest LVEF−lowest LVEF). Table 1 shows the baseline characteristics of all patients and univariable associations with LVEF recovery. Of the 104 eligible patients, 83% were female and 86% self-reported white race with a median age of 52 years old. By the time of anthracycline administration, the most common comorbidities patients presented with were hypertension (37%), hypercholesterolemia (23%), diabetes (18%), and a positive smoking status (31%). Out of all patients, 48% had no heart exposure to radiation therapy, 3% had scatter exposure (est < 0.5 Gy), 39% had low dose near heart exposure (est 0.5-5 Gy), and 10% had partial direct exposure (est >5 Gy). Table 2 presents LVEF characteristics and clinical outcomes among the patient sample, including time between anthracycline initiation to a decline in LVEF (median [IQR] 432 [2793]) days and time between the lowest LVEF to the highest LVEF (median [IQR] 522 [986]) days. The follow-up period for assessment of LVEF recovery ranges between 90 and 4703 days. All patients’ highest and lowest LVEF were measured by echocardiography, except for one patient's was measured by Multiple-Gated Acquisition (MUGA). The clinical outcomes include HF (47%) and death (7%). Of the 7 patients who died, 4 died of cardiovascular causes, 2 died of breast cancer, and 1 the cause of death was unknown. Table 3 describes the patients’ cancer therapy with respect to anthracycline and trastuzumab. All 104 patients received anthracyclines, and 40 patients (38.5%) also received trastuzumab. Most of the 40 patients that received trastuzumab started the trastuzumab after the end date of the anthracycline (35; 87.5%), whereas 5 (12.5%) patients started trastuzumab before the end date of the anthracycline. The most common indication for anthracycline administration was breast cancer (53%). On average, patients had taken anthracycline for 61 days. Tables 4 to 6 describe HF medication utilization, including beta-blockers, ACEI/ARBs, and aldosterone antagonists, respectively. The majority of patients started a beta-blocker after receiving anthracycline therapy (73%), and carvedilol (39%) was the most common type of beta-blocker prescribed. Out of the 76 patients who were on beta-blocker therapy after anthracycline administration, 15 patients (14%) reached the HF target dose. The majority of patients also received an ACEI or ARB after anthracycline administration (63%). The most prescribed was lisinopril (45%). A total of 32 patients (31%) were prescribed an aldosterone antagonist after anthracycline administration. The most prescribed aldosterone antagonist was spironolactone (28%), and 24 patients (23%) were prescribed the HF target dose.

Table 1:

Baseline characteristics of all patients and univariable associations with LVEF recovery

	N	^a Overall	^b Parameter Estimate or Mean ± SD	P*
Age (years)	104	52 (21)	0.080	0.323
Sex	104
Female		86 (82.7%)	16 ± 12.5	0.631
Male		18 (17.3%)	15 ± 11.3	0.631
Race	104
White		89 (85.6%)	15 ± 12.6	0.146
African American		12 (11.5%)	24 ± 8.0
Asian		2 (1.9%)	15 ± 1.1
Other		1 (1.0%)	10 ± N/A
Body Mass Index	104	28 (9.4)	0.101	0.537
Smoking	104
Yes		32 (30.8%)	18 ± 12.5	0.281
No		72 (69.2%)	15 ± 12.2	0.281
Diabetes	104
Yes		19 (18.3%)	14 ± 12.9	0.488
No		85 (81.7%)	16 ± 12.2	0.488
eGFR (mL/min)	57	60 (31.9)	-0.060	0.355
Coronary artery disease	104
Yes		11 (10.6%)	14 ± 14.9	0.621
No		93 (89.4%)	16.1 ± 12.0	0.621
Hypercholesterolemia	104
Yes		24 (23.1%)	16 ± 13.3	0.959
No		80 (76.9%)	16 ± 12.1
Hypertension	104
Yes		38 (36.5%)	19 ± 10.0	0.045
No		66 (63.5%)	14 ± 13.2	0.045
Systolic blood pressure (mmHg)	104	112 (16.7)	0.050	0.561
Diastolic blood pressure (mmHg)	104	64 (9.8)	-0.097	0.572
Mean arterial pressure (mmHg)	104	80 (11.2)	0.001	0.995
Pulse (BPM)	104	81 (19.2)	-0.121	0.218
QRS duration (ms)	21	87 (14)	-0.149	0.430
Troponin (ng/ml)	10	0.18 (0.19)	-28.8	0.265
B-type natriuretic peptide (pg/ml)	27	232 (872)	0.001	0.691
Serum sodium	71	140 (2.6)	1.054	0.082
Heart exposure to radiation Therapy	104
No External Beam Radiation Exposure		50 (48.1%)	17 ± 13.4	0.349
Distant Out of Field Scatter Exposure (est < 0.5 Gy)		3 (2.9%)	14 ± 9.0
Indirect Near Heart Exposure (est 0.5-5 Gy)		40 (38.5%)	17 ± 11.0
Partial Direct Heart Exposure (est >5 Gy)		11 (10.6%)	9.6 ± 12.0

For continuous variables, the median (IQR) is displayed. For categorical variables, the count (%) is displayed.

For continuous variables, the parameter estimate from the linear regression model is displayed. For categorical variables the mean ± SD for each subgroup is displayed.

*p-values < 0.1 are bolded.

BPM, beats per minute; eGFR, estimated glomerular filtration rate; Gy, gray.

Table 2.

LVEF and Clinical Outcomes and Univariable Associations With LVEF Recovery.

	N	Overall^a	Parameter estimate or Mean ± SD^b	P*
Baseline LVEF (%)	73	60 (7.5)	−0.304	.290
Lowest LVEF (%)	104	35 (21.0)	−0.391	< .001
Recovery LVEF (%)(= highest LVEF−lowest LVEF)	104	14 (16.3)	N/A	N/A
Time from anthracycline initiation to a decline in LVEF (days)	104	432 (2793)	−0.001	.054
Time from lowest LVEF to highest LVEF (days)	104	522 (986)	<0.001	.975
Heart failure	104
Yes		49 (47.1%)	16 ± 14.0	.786
No		55 (52.9%)	16 ± 10.6	.786
Death—any cause	104
Yes		7 (6.7%)	13 ± 7.7	.523.564
NoCardiovascular deathYesNo	103	97 (93.3%)4 (3.9%)99 (96.1%)	16 ± 12.612 ± 8.116 ± 12.5	.523.564

For continuous variables, the median (IQR) is displayed. For categorical variables, the count (%) is displayed.

For continuous variables, the parameter estimate from the linear regression model is displayed. For categorical variables, the mean ± SD for each subgroup is displayed.

P-values < .1 are bolded.

LVEF, left ventricular ejection fraction; SD, standard deviation.

Table 3.

Cancer Therapy Characteristics for Patients and Univariable Associations With LVEF Recovery.

	N	Overall^a	Parameter estimate or Mean ± SD^b	P
Anthracycline indication	104
Breast cancer		55 (52.9%)	18 ± 11.2	.240
Leukemia		12 (11.5%)	15 ± 15.1
Lymphoma		25 (24.0%)	16 ± 13.4
Sarcoma		12 (11.5%)	10 ± 10.7
Anthracycline cumulative dose (mg/m²)	101	240 (75)	−0.011	.414
Anthracycline duration (days)	104	61 (75.5)	−0.003	.173
Trastuzumab	104
Yes		40 (38.5%)	17 ± 10.2	.584
No		64 (61.5%)	15 ± 13.5	.584

For continuous variables, the median (IQR) is displayed. For categorical variables, the count (%) is displayed.

For continuous variables, the parameter estimate from the linear regression model is displayed. For categorical variables, the mean ± SD for each subgroup is displayed.

Table 4.

Beta-Blocker Therapy for Patients and Univariable Associations With LVEF Recovery.

	N	Overall^a	Parameter estimate or Mean ± SD^b	P*
Beta-blocker initiated prior to anthracycline administration	104
Yes		5 (4.8%)	27.3 ± 8.6	.033
No		99 (95.2%)	15.3 ± 12.2	.033
Beta-blocker type prior to anthracycline administration	104
Metoprolol tartrate		3 (2.9%)	25.5 ± 10.5	.171
Carvedilol		1 (1.0%)	35 ± 0
Bisoprolol		1 (1.0%)	25 ± 0
None		99 (95.2%)	15.3 ± 12.2
Normalized beta-blocker daily dose (carvedilol equivalents) before anthracycline administration (mg)		0 (0)	0.829	.028
Beta-blocker initiated after anthracycline administration	104
Yes		76 (73.1%)	17.2 ± 12.2	.080
No		28 (26.9%)	12.4 ± 12.1	.080
Beta-blocker type after anthracycline administration	104
Carvedilol		41 (39.4%)	19.9 ± 12.4	.089
Metoprolol succinate		25 (24.0%)	12.4 ± 11.6
Metoprolol tartrate		4 (3.9%)	26.5 ± 8.7
Atenolol		2 (1.9%)	12.5 ± 10.6
Bisoprolol		2 (1.9%)	13.8 ± 5.3
Carvedilol phosphate		1 (1.0%)	8.0 ± 0.0
Propranolol		1 (1.0%)	15.0 ± 0.0
None		28 (26.9%)	12.4 ± 12.1
Normalized beta-blocker daily dose after anthracycline administration (mg)	104	8.7 (23.9)	0.200	<.001
Reached target beta-blocker dose after anthracycline administration	104
Yes		15 (14.4%)	24.9 ± 13.2	.002
No		89 (85.6%)	14.4 ± 11.5	.002
Beta-blocker treatment duration after anthracycline administration (days)	104	408.5 (941.5)	0.00173	.480
Time between LVEF drop and beta-blocker initiation (days)	104	28.5 (210.5)	−0.002	.445
Time between beta-blocker initiation and LVEF recovery (days)	104	1029 (1022)	−0.0009	.657

For continuous variables, the median (IQR) is displayed. For categorical variables, the count (%) is displayed.

For continuous variables, the parameter estimate from the linear regression model is displayed. For categorical variables, the mean ± SD for each subgroup is displayed.

P-values < .1 are bolded.

ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; LVEF, left ventricular ejection fraction; SD, standard deviation.

Table 5.

ACEI/ARB Therapy for Patients at Baseline and Univariable Associations With LVEF Recovery.

	N	Overall^a	Parameter estimate or Mean ± SD^b	P*
ACEI/ARB initiated before anthracycline administration	104
Yes		7 (6.7%)	24.8 ± 8.7	. 047
No		97 (93.3%)	15.2 ± 12.3	. 047
ACEI/ARB type initiated before anthracycline administration	104	−
Lisinopril		4 (3.9%)	24.0 ± 8.0	.228
Enalapril		1 (1.0%)	37.5 ± 0
Ramipril		1 (1.0%)	25.0 ± 0
Losartan		1 (1.0%)	15.0 ± 0
None		97 (93.3%)	15.2 ± 12.3
% of Target ACEI/ARB dose before anthracycline		0 (0)	0.257	. 075
Reached target ACEI/ARB dose before anthracycline administration	104	−
Yes		0 (0%)	N/A	N/A
No		104 (100%)	15.9 ± 12.3	N/A
ACEI/ARB initiated after anthracycline administration	104
Yes		65 (62.5%)	18.1 ± 12.9	.015
No		39 (37.5%)	12.1 ± 10.2	.015
ACEI/ARB type initiated after anthracycline administration	104
Lisinopril		47 (45.2%)	19.6 ± 13.2	.166
Lisinopril/losartan		6 (5.8%)	12.3 ± 12.4
Losartan		4 (3.9%)	15.1 ± 14.7
Benazepril		2 (1.9%)	3.5 ± 9.2
Enalapril		2 (1.9%)	19.0 ± 12.0
Lisinopril/candesartan		1 (1.0%)	10.0 ± 0.0
Lisinopril/irbesartan		1 (1.0%)	14.5 ± 0.0
Lisinopril/valsartan		1 (1.0%)	32.0 ± 0.0
Valsartan		1 (1.0%)	22.0 ± 0.0
None		39 (37.5%)	12.1 ± 10.2
% of Target ACEI/ARB dose after anthracycline	104	12.5 (25.0)	0.113	. 006
Reached target ACEI/ARB dose after anthracycline administration	104	−
Yes		6 (5.8%)	28.6 ± 9.2	. 009
No		98 (94.2%)	15.1 ± 12.1	. 009
ACEI/ARB treatment duration after anthracycline administration (days)	104	314.5 (1050.5)	0.00114	.431
Time between LVEF drop and ACEI/ARB initiation (days)	104	2 (709)	0.000830	.616
Time between ACEI/ARB initiation and LVEF recovery (days)	104	1177 (1565)	−0.000493	.701

For continuous variables, the median (IQR) is displayed. For categorical variables, the count (%) is displayed.

For continuous variables, the parameter estimate from the linear regression model is displayed. For categorical variables, the mean ± SD for each subgroup is displayed.

P-values < .1 are bolded. ACEI = angiotensin converting enzyme inhibitor; ARB = angiotensin receptor blocker; LVEF = left ventricular ejection fraction; SD = standard deviation

Table 6.

Aldosterone Antagonist Therapy for Patients at Baseline and Univariable Associations With LVEF Recovery.

	N	Overall^a	Parameter estimate or Mean ± SD^b	P
Aldosterone antagonist initiated prior to anthracycline administration	104
Yes		0 (0%)	N/A	N/A
No		104 (100%)	15.9 ± 12.3	N/A
Aldosterone antagonist initiated after anthracycline administration	104	−
Yes		32 (30.8%)	18.4 ± 14.8	.164
No		72 (69.2%)	14.8 ± 11.0	.164
Aldosterone antagonist type after anthracycline administration	104	−	−	−
Spironolactone		29 (27.9%)	17.9 ± 15.2	.282
Eplerenone		3 (2.9%)	23.7 ± 11.0
None		72 (69.2%)	14.8 ± 11.0
% of Target aldosterone antagonist dose after anthracycline	104	0 (50)	0.0205	.379
Reached target aldosterone antagonist dose after anthracycline administration	104
Yes		24 (23.1%)	15.1 ± 14.0	.732
No		80 (76.9%)	16.1 ± 11.8	.732
Aldosterone antagonist treatment duration after anthracycline administration (days)	104	0 (60.0)	0.0008	.664
Time between LVEF drop and aldosterone antagonist initiation (days)	104	61 (476)	−0.00005	.987
Time between aldosterone antagonist initiation and LVEF recovery (days)	104	341 (1030)	0.0009	.681

For continuous variables, the median (IQR) is displayed. For categorical variables, the count (%) is displayed.

For continuous variables, the parameter estimate from the linear regression model is displayed. For categorical variables, the mean ± SD for each subgroup is displayed.

Univariable and Multivariable Association Analyses

Tables 1 to 6 also show the results of the univariable analyses. Variables demonstrating an association with P < .1 with improved LVEF recovery included the following: absence of diabetes (P = .067), presence of hypertension (P = .099), decreased time between anthracycline start to decline in LVEF (P = .054), beta-blocker use before (P = .033) and after (P = .080) anthracycline, beta-blocker type after chemotherapy (P = .089), if the beta-blocker prescription reached HF target dose after anthracycline (P = .002), increased beta-blocker dose before (P = .028) and after (P < .001) anthracycline, ACEI/ARB use before (P = .047) and after (P = .015) anthracycline, ACEI/ARB percentage of target dose achieved before (P = .075) and after (P = .006) anthracycline, and reached ACEI/ARB target dose after anthracycline (P = .009). Table 7 presents the final multivariable model of independent variables with P < .05 in the entire patient sample (n = 104) and the subgroup of patients that had baseline LVEF available (n = 73). In the overall patient sample, only 2 variables remained after stepwise selection: average normalized beta-blocker dose after anthracycline exposure (parameter estimate scaled to every 25 mg = 5.04; P = .0005) and time between anthracycline start to lowest LVEF (parameter estimate scaled to every 5-year increment = −1.77; P = .0379). These findings can be interpreted as follows: every 25 mg increase in beta-blocker dose after anthracycline exposure is associated with an average 5% increase in LVEF recovery, and every 5-year increase in the time between start of anthracycline to lowest LVEF is associated with an average 1.8% decrease in LVEF recovery. The results were similar in the subgroup of patients that had baseline LVEF available. Only 2 variables remained after stepwise selection: average normalized beta-blocker dose after anthracycline exposure (parameter estimate scaled to every 25 mg = 3.80; P = .0065) and time between anthracycline start to lowest LVEF (parameter estimate scaled to every 5-year increment = −3.26; P = .0008).

Table 7.

Final Multivariable Model of Statistically Significant and Independent Associations With LVEF Recovery in All Patients (n = 104) and the Subgroup With Baseline LVEF Available (n = 73).

All patients (n = 104)
Parameter	Estimate	Standarderror	t-value	Pr > \|t\|
BB dose after anthracycline exposure (scaled to 25 mg increments)	5.04	1.40	3.59	.0005
Time between anthracycline start to first decline in LVEF (scaled to 5-year increments)	−1.77	0.83	−2.10	.0379

Subgroup with baseline LVEF available (n = 73)
Parameter	Estimate	Standarderror	t-value	Pr > \|t\|
BB dose after anthracycline exposure (scaled to 25 mg increments)	3.80	1.35	2.81	.0065
Time between anthracycline start to first decline in LVEF (scaled to 5-year increments)	−3.26	0.93	−3.52	.0008

Discussion

This retrospective analysis of real-world observational data provides evidence supporting neurohormonal blockers in treating AILVD, with average LVEF recovery being 4.8% and 6.0% higher in patients prescribed beta-blockers and ACEIs/ARBs, respectively, relative to those not prescribed each therapy. Our findings identify 2 important factors that are independently associated with LVEF recovery in AILVD patients. First, we found that for every 25 mg increase in beta-blocker dose after anthracycline exposure, LVEF recovery increased by 5%. This finding supports the initiation and titration of beta-blocker therapy to target HF doses. Next, every 5-year increase between the time of anthracycline start and the lowest LVEF was associated with a 1.8% decrease in LVEF recovery. This finding underscores the importance of early and regular monitoring for cardiac dysfunction, as is recommended by current cancer organization guidance statements.^27–30 This study adds to a very limited body of evidence supporting the real-world efficacy of neurohormonal blockers for treating AILVD. A previous observational study demonstrated an association of beta-blocker therapy with reducing symptomatic HF after anthracycline treatment.³¹ Another study demonstrated that patients with prescriptions for ACEIs or beta-blockers, particularly those started before or within 6 months of cardiotoxic cancer therapy and lasting for at least 6 months, had a reduced cumulative incidence of cardiotoxicity-related diagnoses.³² The study did not find an association between ACEI/beta-blocker dose and the development of cardiotoxicity; however, it must be noted that the study included patients receiving anthracyclines, trastuzumab, or both and did not perform stratification to specifically assess the association of neurohormonal therapies in preventing AILVD. Our study also enrolled a limited number of patients who were taking beta-blockers (n = 5) or ACEIs/ARBs (n = 7) prior to initiation of their anthracycline treatment and experienced breakthrough AILVD.

Pharmacologic strategies to treat AILVD are limited by a lack of evidence regarding optimal selection and timing of therapy. In patients who develop symptomatic HF during or after anthracycline therapy, professional organizations consistently recommend treatment according to current HF guidelines,^27,29,30 which include first-line recommendations for beta-blocker and/or ACEI/ARB therapy.^26,33 With the 2022 update, the American Heart Association (AHA)/American College of Cardiology (ACC)/Heart Failure Society of America (HFSA) recommend neurohormonal blockers only to patients with reduced LVEF or symptoms.²⁶ Similarly, the European Society of Cardiology (ESC), the European Society for Medical Oncology (ESMO), and National Comprehensive Cancer Network (NCCN) make low strength recommendations for neurohormonal blockers for AILVD treatment but not for prevention.^27,29,30 Guidelines from the American Society of Clinical Oncology (ASCO) make no specific recommendations for pharmacologic treatment for AILVD, noting that there is a lack of knowledge regarding the timing and choice of optimal AILVD therapies.²⁸

The pathophysiology of AILVD is generally thought to involve free radical mediated DNA damage and the downstream consequences. It should be noted that the beta-blocker and ACEI/ARB treatment is probably treating the consequences of the anthracycline-induced damage. Beta-blockers or ACEIs/ARBs demonstrated no cardioprotective or preventive effect on LV dysfunction.³⁴ Instead, they primarily work to treat the damage caused by anthracyclines in the early stage.³⁵ Several mechanisms of beta-blockers for treating AILVD have been proposed. These include modulating the neurohormonal sympathetic system to reduce heart rate and myocardial oxygen demand, thereby relieving stress on the heart;³⁶ reducing reactive oxygen species to mediate oxidative stress;³⁷ improving systolic function to enhance cardiac output;³⁸ inhibiting the apoptosis of cardiac cells by blocking specific pathways;³⁹ and reducing cardiac fibrosis to maintain heart function.⁴⁰

Few clinical trials have assessed the efficacy of beta-blockers or ACEIs/ARBs in treating AILVD once it has manifested (ie, recovering LVEF and/or reducing HF symptoms). Two prospective clinical trials initiated enalapril and/or a beta-blocker (carvedilol or bisoprolol) in 221⁸ and 201¹² patients who developed AILVD. While neither trial included an untreated AILVD cohort for comparison, both found that more than half of patients had at least partial LVEF recovery with neurohormonal blockade: one study found that 55% of patients had a recovery >10% from LVEF nadir¹² and the other found that 82% had a recovery >5% from LVEF nadir.⁸ In both trials, LVEF recovery was significantly greater in patients treated with both an ACEI/ARB and a beta-blocker relative to those treated with either as monotherapy.

In addition to considerations for the selection of therapy, we found that attainment of target HF doses was associated with LVEF recovery for both beta-blockers and ACEIs/ARBs in our univariable analyses. This finding provides a potential explanation for discordant results seen in past clinical trials assessing AILVD prevention with neurohormonal blockers. All 4 clinical trials assessing ACEI/ARB monotherapy for AILVD prevention wherein doses were titrated (if tolerated) to achieve target HF doses found protective effects.^15,41–43 Similarly, one combination trial with carvedilol and enalapril, where doses were titrated to target HF doses, also demonstrated a protective effect.¹⁷ While the only beta-blocker monotherapy trial for AILVD prevention that attempted to titrate to target HF doses was negative, this may have been due to the average actual carvedilol dose in the trial being 22 mg/day (less than half of the target dose of 50 mg/day).¹⁶ It is also noteworthy that cumulative anthracycline dose and combination therapy with an anthracycline plus trastuzumab, which have both been associated with AILVD development,^8,31 were not associated with reduced LVEF recovery in our study. This finding suggests that neurohormonal blockers appear effective for LVEF recovery in these high-risk populations, though it is worth pointing out that most patients in our study were not on high-dose anthracyclines (average cumulative dose was 240 mg/m² doxorubicin equivalents).

This study has some limitations. It had a small sample size and was performed at a single site. This small sample also has limited diversity; the majority of the patients are female with self-reported White race. Moreover, its observational design only allowed for collection of LVEF measurements obtained during normal clinical care. Our findings should be considered preliminary and require replication using larger real-world clinical datasets. Additionally, cancer staging and metastatic status were not collected. Future research investigating the differences in staging among patients and the invasiveness of their cancers could be beneficial. Lastly, the improvement of LVEF collected in the study was determined by the single highest LVEF collected. The stability of LVEF over time and whether the continuation of HF therapy increases stability should be further investigated.

Conclusion

Two variables were significantly and independently associated with LVEF recovery in this small sample of patients with AILVD: beta-blocker dose after anthracycline exposure and the time between anthracycline start to lowest LVEF. Every 25 mg increase beta-blocker dose after anthracycline exposure was associated with an average 5% increase in LVEF recovery, and every 5-year increase in the time between the start of anthracycline to lowest LVEF was associated with an average 1.8% decrease in LVEF recovery. Our findings from this real-world dataset corroborate those from past clinical trials^8,12 and suggest that beta-blocker therapy should be titrated to target HF doses, as tolerated, to achieve maximum LVEF recovery. Additionally, our findings suggest the benefits of early monitoring of cardiac function, which would facilitate early detection and therapy. Future work is needed to validate these findings in larger real-world clinical cohorts.

Supplemental Material

sj-docx-1-cpt-10.1177_10742484241304304 - Supplemental material for Factors Associated With the Recovery of Left Ventricular Ejection Fraction in Patients With Anthracycline-Induced Left Ventricular Dysfunction

Supplemental material, sj-docx-1-cpt-10.1177_10742484241304304 for Factors Associated With the Recovery of Left Ventricular Ejection Fraction in Patients With Anthracycline-Induced Left Ventricular Dysfunction by Tyler Shugg, PharmD, PhD, Tk Nguyen, PharmD, Xuesi Hua, BS, Blair Richards, MPH, James Rae, PhD, Robert Dess, MD, Daniel Perry, MD, Bradley Kay, MD, Salim S. Hayek, MD, Monika Leja, MD, and Jasmine A. Luzum, PharmD, PhD in Journal of Cardiovascular Pharmacology and Therapeutics

Footnotes

Author Contributions

Data collection: TS, TN, JE, BR, BD, DP, and BK; interpretation: TS, TN, JE, XH, BR, JR, BD, DP, BK, SSH, and MJ; writing: TS, TN, JE, XH, SSH, and JAL; revision: TS, TN, JE, XH, JR, BD, DP, BK, SSH, ML, and JAL; data analysis: TN, BR, and JAL; critical review: TS, TN, JE, XH, BR, JR, BD, DP, BK, SSH, ML, and JAL; funding acquisition: ML and JAL; conceptualization: JR, ML, and JAL.

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: J.A.L. is a consultant for Ariel Precision Medicine. The other authors have no conflicts of interest to disclose.

Funding

TS is supported by NIH grant #K23GM147805. JAL was supported by NIH grant #K08HL146990. EMERSE: Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Numbers U24CA204863 and P30CA046592, as well as the National Center for Advancing Translational Sciences of the National Institutes of Health under Award Numbers UL1TR000433 and UL1TR002240. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

ORCID iD

Xuesi Hua

Supplemental Material

Supplemental material for this article is available online.

References

Siegel

Miller

Jemal

. Cancer statistics, 2019. CA Cancer J Clin. 2019;69(1):7-34. doi:https://doi.org/10.3322/caac.21551

Eichenauer

Aleman

BMP

André

, et al. Hodgkin lymphoma: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol: Off J Eur Soc Med Oncol. 2018;29(Suppl 4):iv19-iv29. doi:https://doi.org/10.1093/annonc/mdy080

Casali

Bielack

Abecassis

, et al. Bone sarcomas: ESMO-PaedCan-EURACAN clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol: Off J Eur Soc Med Oncol. 2018;29(Suppl 4):iv79-iv95. doi:https://doi.org/10.1093/annonc/mdy310

Cardoso

Kyriakides

Ohno

, et al. Early breast cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up†. Ann Oncol: Off J Eur Soc Med Oncol. 2019;30(8):1194-1220. doi:https://doi.org/10.1093/annonc/mdz173

Von Hoff

Layard

Basa

, et al. Risk factors for doxorubicin-induced congestive heart failure. Ann Internal Med. 1979;91(5):710-717. doi:https://doi.org/10.7326/0003-4819-91-5-710

Swain

Whaley

Ewer

. Congestive heart failure in patients treated with doxorubicin: A retrospective analysis of three trials. Cancer. 2003;97(11):2869-2879. doi:https://doi.org/10.1002/cncr.11407

Gianni

Herman

Lipshultz

Minotti

Sarvazyan

Sawyer

. Anthracycline cardiotoxicity: From bench to bedside. J Clin Oncol: Off J Am Soc Clin Oncol. 2008;26(22):3777-3784. doi:https://doi.org/10.1200/jco.2007.14.9401

Cardinale

Colombo

Bacchiani

, et al. Early detection of anthracycline cardiotoxicity and improvement with heart failure therapy. Circulation. 2015;131(22):1981-1988. doi:https://doi.org/10.1161/circulationaha.114.013777

Bristow

Billingham

Mason

Daniels

. Clinical spectrum of anthracycline antibiotic cardiotoxicity. Cancer Treat Rep. Jun 1978;62(6):873-879.

10.

Ewer

. Cardiotoxicity of anticancer treatments: What the cardiologist needs to know. Nature Rev Cardiol. 2010;7(10):564-575. doi:https://doi.org/10.1038/nrcardio.2010.121

11.

Fornaro

Olivotto

Rigacci

, et al. Comparison of long-term outcome in anthracycline-related versus idiopathic dilated cardiomyopathy: A single centre experience. Eur J Heart Fail. 2018;20(5):898-906. doi:https://doi.org/10.1002/ejhf.1049

12.

Cardinale

Colombo

Lamantia

, et al. Anthracycline-induced cardiomyopathy: Clinical relevance and response to pharmacologic therapy. J Am Coll Cardiol. 2010;55(3):213-220. doi:https://doi.org/10.1016/j.jacc.2009.03.095

13.

Dessì

Madeddu

Piras

, et al. Long-term, up to 18 months, protective effects of the angiotensin II receptor blocker telmisartan on epirubin-induced inflammation and oxidative stress assessed by serial strain rate. Springerplus. 2013;2(1):198. doi:https://doi.org/10.1186/2193-1801-2-198

14.

Georgakopoulos

Roussou

Matsakas

, et al. Cardioprotective effect of metoprolol and enalapril in doxorubicin-treated lymphoma patients: A prospective, parallel-group, randomized, controlled study with 36-month follow-up. Am J Hematol. 2010;85(11):894-896. doi:https://doi.org/10.1002/ajh.21840

15.

Gulati

Heck

Ree

, et al. Prevention of cardiac dysfunction during adjuvant breast cancer therapy (PRADA): A 2 x 2 factorial, randomized, placebo-controlled, double-blind clinical trial of candesartan and metoprolol. Eur Heart J. 2016;37(21):1671-1680. doi:https://doi.org/10.1093/eurheartj/ehw022

16.

Avila

Ayub-Ferreira

de Barros Wanderley

, et al. Carvedilol for prevention of chemotherapy-related cardiotoxicity: The CECCY trial. J Am Coll Cardiol. 2018;71(20):2281-2290. doi:https://doi.org/10.1016/j.jacc.2018.02.049

17.

Bosch

Rovira

Sitges

, et al. Enalapril and carvedilol for preventing chemotherapy-induced left ventricular systolic dysfunction in patients with malignant hemopathies: The OVERCOME trial (preventiOn of left Ventricular dysfunction with Enalapril and caRvedilol in patients submitted to intensive ChemOtherapy for the treatment of Malignant hEmopathies). J Am Coll Cardiol. 2013;61(23):2355-2362.

18.

Livi

Barletta

Martella

, et al. Cardioprotective strategy for patients with nonmetastatic breast cancer who are receiving an anthracycline-based chemotherapy: A randomized clinical trial. JAMA Oncol. 2021;7(10):1544-1549. doi:https://doi.org/10.1001/jamaoncol.2021.3395

19.

Cochera

Dinca

Bordejevic

, et al. Nebivolol effect on doxorubicin-induced cardiotoxicity in breast cancer. Cancer Manag Res. 2018;10:2071-2081. doi:https://doi.org/10.2147/cmar.S166481

20.

Costanzo

Ratre

Andretta

Acharya

Bhaskar

LVKS

Verma

. A comprehensive review of cancer drug-induced cardiotoxicity in blood cancer patients: Current perspectives and therapeutic strategies. Curr Treat Opt Oncol. 2024;25(4):465-495. doi:https://doi.org/10.1007/s11864-023-01175-z

21.

Harris

Taylor

Minor

, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inf. 2019;95:103208. doi:https://doi.org/10.1016/j.jbi.2019.103208

22.

Harris

Taylor

Thielke

Payne

Gonzalez

Conde

. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inf. 2009;42(2):377-381. doi:https://doi.org/10.1016/j.jbi.2008.08.010

23.

Hanauer

Mei

Law

Khanna

Zheng

. Supporting information retrieval from electronic health records: A report of university of Michigan's nine-year experience in developing and using the electronic medical record search engine (EMERSE). J Biomed Inf. 2015;55:290-300. doi:https://doi.org/10.1016/j.jbi.2015.05.003

24.

Shankar

Marina

Hudson

, et al. Monitoring for cardiovascular disease in survivors of childhood cancer: Report from the cardiovascular disease task force of the children's oncology group. Pediatrics. 2008;121(2):e387-e396. doi:https://doi.org/10.1542/peds.2007-0575

25.

Luzum

Edokobi

Dorsch

, et al. Survival association of angiotensin inhibitors in heart failure with reduced ejection fraction: Comparisons using self-identified race and genomic ancestry. J Card Fail. 2022;28(2):215-225. doi:https://doi.org/10.1016/j.cardfail.2021.08.007

26.

2022 AHA/ACC/HFSA. Guideline for the management of heart failure. J Cardiac Fail. May 2022;28(5):e1-e167. doi:https://doi.org/10.1016/j.cardfail.2022.02.010

27.

National Comprehensive Cancer Network. 2022. NCCN Clinical Practice Guidelines in Oncology: Survivorship. 2022. https://www.nccn.org/professionals/physician_gls/pdf/survivorship.pdf.

28.

Armenian

Lacchetti

Barac

, et al. Prevention and monitoring of cardiac dysfunction in survivors of adult cancers: American society of clinical oncology clinical practice guideline. J Clin Oncol: Off J Am Soc Clin Oncol. 2017;35(8):893-911. doi:https://doi.org/10.1200/jco.2016.70.5400

29.

Zamorano

Lancellotti

Rodriguez Muñoz

, et al. 2016 ESC position paper on cancer treatments and cardiovascular toxicity developed under the auspices of the ESC committee for practice guidelines: The task force for cancer treatments and cardiovascular toxicity of the European Society of Cardiology (ESC). Eur J Heart Failure. 2017;19(1):9-42. doi:https://doi.org/10.1002/ejhf.654

30.

Curigliano

Lenihan

Fradley

, et al. Management of cardiac disease in cancer patients throughout oncological treatment: ESMO consensus recommendations. Ann Oncol: Off J Eur Soc Med Oncol. 2020;31(2):171-190. doi:https://doi.org/10.1016/j.annonc.2019.10.023

31.

Seicean

Alan

Plana

Budd

Marwick

. Cardioprotective effect of β-adrenoceptor blockade in patients with breast cancer undergoing chemotherapy: Follow-up study of heart failure. Circulation Heart Failure. 2013;6(3):420-426. doi:https://doi.org/10.1161/circheartfailure.112.000055

32.

Wittayanukorn

Qian

Westrick

Billor

Johnson

Hansen

. Prevention of trastuzumab and anthracycline-induced cardiotoxicity using angiotensin-converting enzyme inhibitors or β-blockers in older adults with breast cancer. Am J Clin Oncol. 2018;41(9):909-918. doi:https://doi.org/10.1097/coc.0000000000000389

33.

McDonagh

Metra

Adamo

, et al. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42(36):3599-3726. doi:https://doi.org/10.1093/eurheartj/ehab368

34.

Henriksen

Hall

MacPherson

, et al. Multicenter, prospective, randomized controlled trial of high-sensitivity cardiac troponin I–guided combination angiotensin receptor blockade and Beta-blocker therapy to prevent anthracycline cardiotoxicity: The cardiac CARE trial. Circulation. 2023;148(21):1680-1690. doi:doi:https://doi.org/10.1161/CIRCULATIONAHA.123.064274

35.

Heck

Mecinaj

Ree

, et al. Prevention of cardiac dysfunction during adjuvant breast cancer therapy (PRADA): Extended follow-up of a 2 x 2 factorial, randomized, placebo-controlled, double-blind clinical trial of candesartan and metoprolol. Circulation. 2021;143(25):2431-2440. doi:doi:https://doi.org/10.1161/CIRCULATIONAHA.121.054698

36.

Zeng

. Preventive use of beta-blockers for anthracycline-induced cardiotoxicity: A network meta-analysis. Front Cardiovasc Med. 2022;9:968534. doi:https://doi.org/10.3389/fcvm.2022.968534

37.

Vuong

Stein-Merlob

Cheng

Yang

. Novel therapeutics for anthracycline induced cardiotoxicity. Review. Front Cardiovasc Med. 2022;9:863314. doi:https://doi.org/10.3389/fcvm.2022.863314

38.

Nohria

. β-Adrenergic blockade for anthracycline- and trastuzumab-induced cardiotoxicity. Circulation: Heart Failure. 2013;6(3):358-361. doi:https://doi.org/10.1161/CIRCHEARTFAILURE.113.000267

39.

Osoro

Sharma

Amir

, et al. Prevention and management of anthracycline induced cardiotoxicity: A review. Health Sci Rev. 2022;5:100070. doi:https://doi.org/10.1016/j.hsr.2022.100070

40.

Mir

Badi

Bugazia

, et al. Efficacy and safety of cardioprotective drugs in chemotherapy-induced cardiotoxicity: An updated systematic review & network meta-analysis. Cardio-Oncology. 2023;9(1):10. doi:https://doi.org/10.1186/s40959-023-00159-0

41.

Cardinale

Colombo

Sandri

, et al. Prevention of high-dose chemotherapy-induced cardiotoxicity in high-risk patients by angiotensin-converting enzyme inhibition. Circulation. 2006;114(23):2474-2481. doi:https://doi.org/10.1161/circulationaha.106.635144

42.

Janbabai

Nabati

Faghihinia

Azizi

Borhani

Yazdani

. Effect of enalapril on preventing anthracycline-induced cardiomyopathy. Cardiovasc Toxicol. 2017;17(2):130-139. doi:https://doi.org/10.1007/s12012-016-9365-z

43.

Słowik

Jagielski

Potocki

, et al Anthracycline-induced cardiotoxicity prevention with angiotensin-converting enzyme inhibitor ramipril in women with low-risk breast cancer: Results of a prospective randomized study. Kardiol Pol. 2020;78(2):131-137. doi:https://doi.org/10.33963/kp.15163

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB