Diagnosis and treatment of left ventricular dysfunction and heart failure in cancer patients

Introduction

Early diagnosis and effective new treatments for many cancers have resulted in a growing population of survivors either cured of cancer or living with cancer as a stable disease, chronically controlled by long-term treatment. Some therapies which are required to cure cancer or to control it chronically are cardiotoxic. This has resulted in a growing population of cancer patients with cardiovascular complications, with associated morbidity and mortality [1]. Left ventricular dysfunction (LVD) and heart failure (HF) are two of the most serious complications of cancer treatment, particularly if they occur during treatment, leading to treatment interruption and interfering with optimal cancer care. Late presenting cases, years after treatment, are also a concern as they often have a more advanced HF syndrome refractory to HF therapy. Both historical and more contemporary studies report that patients with HF secondary to cancer therapies presenting late (>5 years following treatment) with symptomatic HF have a worse prognosis than patients with HF resulting from other causes [2,3].

Various cancer therapies including anthracycline chemotherapy (AC) (doxorubicin, epirubicin, daunorubicin and idarubicin), molecular targeted therapies such as trastuzumab (Herceptin™) and various tyrosine kinase inhibitors (TKIs), proteasome inhibitors (PIs) for multiple myeloma, immune checkpoint inhibitors (ICIs) and radiation therapy involving the heart in the treatment field, can all cause LVD and HF. They are summarised in Table 1.

Table 1. Cancer therapies known to cause left ventricular dysfunction and heart failure.

* Higher-risk drugs in each category.

Given that the timing and prescribing of potentially cardiotoxic cancer treatments is known, there is an opportunity to risk assess oncology patients prior to treatment and employ surveillance strategies during cardiotoxic cancer treatments and throughout follow-up, with the aim of early detection and implementation of CV risk factor management and cardioprotective treatment strategies where appropriate to support the best cancer treatment.

In modern healthcare this has led to the development of specialist cardio-oncology services to coordinate care between cardiology and oncology with the mission of improving cardiovascular and oncology outcomes for these patients [4,5]. In this article we focus on the detection, management and prevention of LVD and HF caused by cancer therapies, including treatment in patients with LVD or HF at baseline pre-treatment.

Pre-existing HF or LVD at cancer diagnosis pre-treatment

The number of patients with HF is increasing with the ageing population and increasing survival from cardiovascular diseases (CVD) such as acute myocardial infarction, hypertension and valvular heart disease, and cancer survivors [6]. The population of patients with pre-existing CVD presenting with cancer is increasing, including those with pre-existing HF or LVD at the time of cancer diagnosis [7].  This reflects many shared risk factors for CVD and cancer, particularly increasing age, smoking, obesity, diabetes mellitus and sedentary lifestyle. There is also the growing number of cancer survivors who present with a second malignancy, or recurrence of their original cancer, and who have received previous cardiotoxic treatment for their first cancer years earlier. Intriguingly, there is also emerging evidence that HF itself is a risk factor for the development of cancer [8].

Patients with pre-existing HF or LVD represent one of the highest risk patient cohorts for cardiotoxicity, with cardiotoxic cancer treatments associated with direct myocardial toxicity [7]. A detailed baseline assessment is strongly recommended to identify the severity of HF, aetiology, symptom status (including fluid congestion), treatment of any reversible factors and implementation and optimisation of guideline-based medical therapy. As coronary artery disease (CAD) is one of the most common causes of heart failure, the patient should be asked about any history of CAD or any ischaemic chest pain. If the index of suspicion increases (especially in patients over 50 years of age), non-invasive perfusion imaging should be considered to assess for the presence and extent of myocardial ischaemia.

The treatment of clinically significant HF should be implemented according to the current ESC heart failure guidelines [9]. For patients with heart failure and reduced ejection fraction (HFrEF), medical treatment including angiotensin-converting enzyme inhibitors (ACEi) or angiotensin receptor blockers (ARB) in combination with ß-blockers and mineralocorticoid receptor antagonists (MRA), should be initiated where they are tolerated by the patient and providing there are no other contraindications. In patients with HF, reviewing doses of guideline-directed medical therapy is important, with uptitration to target doses in patients not currently receiving optimal doses [9]. This may need to be performed in an expedited fashion if oncology treatment is being delayed pending rapid uptitration and optimisation of HF therapies, but with appropriate monitoring of blood pressure, heart rate, renal function and serum potassium as appropriate. In some oncology patients where initiation of cancer therapy cannot be delayed, further uptitration of ACEi/ARB or ß-blockers can occur in a timely manner after starting cancer treatment, potentially starting with non-cardiotoxic treatments first if available, and consideration of the introduction of the cardiotoxic treatment once the cardiac status has been optimised.

The decision on whether to proceed with a potentially cardiotoxic cancer treatment in a patient with pre-existing LVD or HF depends on several factors (Table 2). This management plan requires close discussion between the cardiology and oncology teams regarding the options, including availability and efficacy of alternative non-cardiotoxic cancer treatments, dose reduction or modification and, if treatment with a cardiotoxic treatment is initiated, close surveillance with cardiac biomarkers, imaging, electrocardiography (ECG) and clinical review. Assessment of any treatment-emergent cardiac symptoms is strongly recommended. There are some cases where the severity of HF is sufficiently high that cardiotoxic cancer therapy is absolutely contraindicated. However, in the experience of the authors, many patients with pre-existing CVD including LVD and HF can be treated safely with cancer treatments known to cause cardiotoxicity including anthracycline chemotherapy, trastuzumab and TKIs. Treatment decisions made within a multidisciplinary team for complex cases is strongly recommended, ideally in a cardio-oncology service, and with all the information provided to the patient so that they can make an informed decision regarding the risks versus benefits of proceeding with treatment.

Table 2. Factors influencing cancer treatment in patients with pre-existing heart failure.

In the case of patients with pre-existing LVD or HF who require anthracycline chemotherapy in a curative treatment pathway, e.g., lymphoma and sarcoma patients, consideration of dexrazoxane as primary prevention may be considered based on the growing evidence of cardioprotection in paediatric cancer populations [1,10]. Dexrazoxane acts to reduce myocardial toxicity when infused before the anthracycline infusion, so should be considered in a primary prevention strategy. The current licence for dexrazoxane use in adults is restricted to metastatic breast cancer. However, it is the opinion of the authors that it may be considered on a patient-by-patient basis for off-label use for cancer patients with pre-existing HF, particularly HFrEF, where an anthracycline (AC)-containing (AC) chemotherapy protocol is associated with a significantly higher rate of cure compared to non-AC-containing treatment protocols.

Primary prevention in high-risk cases

Baseline cardiovascular risk assessment is recommended for all oncology patients scheduled to receive potentially cardiotoxic cancer therapies. For those patients where the risk of the cancer treatment is LVD or HF, and whose baseline CV risk is high, we recommend referral to a cardio-oncology service to optimise pre-existing CVD and risk factors in a timely manner. Consideration should be given to starting cardioprotective medication when indicated.

Options to reduce the risk of cardiotoxicity depend on the cancer treatment class. These are summarised in Table 3. Most studies have focused on the role of dexrazoxane for AC, and the role of ACEi, BB, MRA or a combination in oncology patients scheduled to receive AC or AC plus trastuzumab.

Table 3. Primary prevention strategies to reduce left ventricular dysfunction and heart failure in cancer patients.

The decision to start primary prevention treatment with ACEi, ß-blocker or both will depend on patient-specific factors (blood pressure, renal function, heart rate and rhythm, comorbidities and potential contraindications) and perceived benefits, and treatment-specific factors. These should be discussed with the patient with an explanation of the rationale, benefits and potential side effects.

Detection and management of LV dysfunction during cancer treatment

Baseline cardiovascular risk

The absolute risk of developing LVD or HF during cancer treatment depends on the specific anti-cancer therapy, the doses prescribed, previous cardiotoxic cancer therapies, and pre-existing CVD. These all contribute to the baseline CV risk of the cancer patient. Most information is available for AC. The risk of LVD from anthracyclines depends on the total cumulative dose and the interaction with pre-existing risk factors and other oncology therapies. Overall, the estimated range is between 2 and 48%, with a clear dose/risk relation, although pre-existing CV disease and risk factors also determine absolute risk [11,12]. For contemporary ABVD or CHOP protocols delivering 300 mg/m2 doxorubicin, in younger lymphoma patients with low baseline CV risk the risk of short-term cardiotoxicity is ~5%, whereas in older patients with higher baseline CV risk it is 30% or higher. Co-administration of other chemotherapy agents such as cyclophosphamide is also associated with an increased risk. Any dose above 200 mg/m2 may cause cardiotoxicity, even in individuals without pre-existing CVD.

There is growing use of targeted molecular therapies (Table 1) such as trastuzumab, an HER2 receptor antagonist, TKIs against vascular endothelial growth factor (VEGF) for a range of solid tumours (including metastatic renal cancer, thyroid cancer, hepatocellular carcinoma, gastrointestinal stromal tumour [GIST] and sarcoma), BCr-Abl TKIs for chronic myeloid leukaemia and PIs for multiple myeloma, which all cause direct myocardial toxicity and are associated with a higher risk of developing LVD and HF, particularly in individuals with pre-existing CV disease.

The novel cancer immunotherapies (ICIs) can also cause significant LV dysfunction. This is an emerging problem [13]. ICIs cause an immune-mediated myocarditis, with severe cases presenting with fulminant myocarditis, cardiogenic shock VT or VF, and death from cardiogenic shock or refractory arrhythmias has been reported in up to 50% of cases [13,14]. A growing number of cases of non-inflammatory HF due to ICIs is also being recognised [13]. The incidence of HF from ICIs is currently not known. With the wider use of ICIs in oncology patients, the true incidence and risk of cardiotoxicity from these agents will become established. The risk factor profile for these immunotherapy agents is also poorly understood, but may include dual ICI therapy, combination with a known cardiotoxic TKI, pre-existing CV disease including coronary artery disease, premorbid autoimmune disorders or cardiac antigen expression in the tumour.

Definition of LV dysfunction during cancer treatment

Several different definitions of cardiotoxicity during cancer treatment exist, reflecting different definitions used in cancer trials and in real-world cardiology practice. The current ESC position statement on cardio-oncology describes LVD as a decrease in the left ventricular ejection fraction (LVEF) of ≥10% and to a value below the lower limit of normal. The lower limit of normal for LVEF depends upon the local echocardiography department but is frequently defined as 50% [1]. LVEF >55% is usually normal, and in patients with LVEF 50-54% in a borderline range which is low normal in some patients and reduced in others depending upon other independent markers of LV dysfunction, e.g., symptoms, global longitudinal strain (GLS) and elevation of cardiac biomarkers. Given that the variability of LVEF measurement is at least +/-5%, repeated assessment to evaluate the trend of decline versus stability, other imaging parameters (e.g., GLS) and cardiac biomarkers are essential.

Based on echocardiographic findings and biomarkers (troponin and brain natriuretic peptide [BNP]), Pareek et al suggested a six-step classification of myocardial toxicity from their cardio-oncology service [4]. This classification contains six categories with three mild “subclinical” categories (1-3) and three categories of more severe cardiotoxicity: 1) early biochemical cardiotoxicity, 2) early functional cardiotoxicity, 3) early mixed cardiotoxicity, 4) symptomatic HFpEF (LVEF ≥50%), 5) asymptomatic left ventricular systolic dysfunction (LVSD) (LVEF <50%), and 6) symptomatic LVSD (LVEF <50%). Besides the diagnostic tools, the authors recommend treatment options and give advice on how to conduct the cancer treatment (Table 4) [4].

Table 4. Myocardial toxicity classification from Royal Brompton Hospital cardio-oncology service (reproduced from Pareek et al [4] with permission).

* continuing cardiotoxic cancer therapy may be suitable in selected cases depending on the risk/benefit ratio, severity of left ventricular impairment, symptoms, cancer stage and response.

 ¹ If LVEF fall is to >50%, then incorporate either biomarker elevation or GLS reduction (<18% if normal baseline, or <15% relative reduction of GLS if reduced at baseline).

 ² If ACEI or BB are not tolerated, or the patient is already taking these agents when cardiotoxicity is diagnosed, consider adding aldosterone antagonist.

³ If LVEF <35% follow the ESC HF guidelines regarding eligibility for cardiac resynchronisation therapy, sacubitril/valsartan and ivabradine.

ACEI: angiotensin-converting enzyme inhibitor; BB: beta-blocker; BNP: brain natriuretic peptide; GLS: global longitudinal strain; HF: heart failure; HFpEF: heart failure with preserved ejection fraction; LVEF: left ventricular ejection fraction; LVSD: left ventricular systolic dysfunction

Surveillance, detection and management of cardiotoxicity during cancer treatment

The two most widely studied strategies for surveillance and early detection are measurement of cardiac biomarkers (cardiac troponin and NPs) and cardiac imaging to measure LVEF and, if available, left ventricular GLS with echocardiography [1,4]. Currently, there are no clear guidelines regarding the specific timing of the investigations (echocardiography and biomarkers) as many variables including baseline CV disease, baseline cardiac and risk, the nature and planned dose of cancer treatment proposed, and previous cardiotoxic cancer treatment (predicted lifetime cumulative dose) influence the decision on frequency of assessment. Many guidelines and position statements support the use of regular cardiac biomarker measurement and cardiac imaging (usually echocardiography) in patients receiving cardiotoxic cancer therapies associated with a risk of developing LVD and HF [1,15].

One algorithm is not appropriate for all patients receiving cardiotoxic treatments. Baseline CV risk is important to personalise the surveillance strategy. An option for patients who receive anthracycline could be to measure cardiac biomarkers such as cardiac troponin serially with each cycle of chemotherapy, repeating echocardiography at a frequency dependent on the baseline risk assessment. The ESC position paper recommends repeat echocardiography after every four cycles of anti-HER2 treatment or after a cumulative doxorubicin dose (or equivalent) of 200 mg/m2 in low-risk patients, whereas in patients with an abnormal baseline assessment the surveillance should be more frequent [1]. Several trials are currently underway, including the Cardiac Care (EudraCT 2017-000896-99) and SUCCOUR (ACTRN12614000341628) trials, assessing the role of high-sensitivity troponin and speckle tracking surveillance, respectively, to guide initiation of cardioprotective treatment in patients receiving anthracycline chemotherapy.

Pareek et al provide guidance about how to proceed with the cancer treatment when abnormalities are detected during cardiotoxic cancer treatment (Table 4). Cancer treatment should continue in patients with “subclinical” cardiotoxicity (categories 1-3), with management options including closer surveillance and/or the use of cardioprotective medications such as ß-Blocker and ACEi. In contrast, for more severe cardiotoxicity (categories 4-6), a temporary interruption is usually necessary for heart recovery.

All of these decisions should be made based on the potential risks and benefits. One category which requires specific consideration is the cohort of breast cancer patients treated with trastuzumab in whom asymptomatic LVSD (category 5 cardiotoxicity) is detected using surveillance echocardiography. This is a relatively common problem: in the experience of the authors, many patients with mild asymptomatic decreases in LVSD (in the range 40-49%) and without high rises in NP or troponin can continue trastuzumab uninterrupted whilst cardioprotective medication is started. In this cohort, closer monitoring is required if continuing trastuzumab in a patient with symptomatic LVSD. This is best managed by a specialist cardio-oncology service.

Whilst no specific β-Blocker or ACEi is specifically recommended, there are some research studies which guide the field to specific examples. Carvedilol and nebivolol show the best cardioprotective effect in patients who received anthracycline therapy, whereas no significant benefit has been found for metoprolol [16-18]. Propranolol has a potentially adverse effect and should be avoided [19]. A retrospective propensity-matched randomised study suggested that women taking β-blockers were protected against trastuzumab-related cardiotoxicity [20]. A preclinical study suggested that carvedilol rather than β1-selective β-blockers would be particularly protective [21].

The most widely used ACEi is enalapril which also has the strongest evidence base for HF treatment. Studies have assessed enalapril in primary prevention or after troponin elevation with evidence of benefit [22,23]. The OVERCOME trial assessed the role of enalapril and carvedilol in combination against anthracycline chemotherapy in haematology patients with evidence of cardioprotection [24]. The ICOS-ONE trial recently reported that a primary prevention strategy with enalapril was equally effective to a biomarker-guided strategy where enalapril was restricted to those who developed a troponin rise during anthracycline treatment [22]. The PROACT trial is currently recruiting patients receiving anthracyclines to evaluate the absolute benefit of enalapril as primary prevention versus an untreated control arm (EudraCT: 2017-001094-16).

In our opinion, primary prevention is suitable for high-risk patients but is not currently recommended as routine for all cancer patients receiving cardiotoxic drugs due to potential side effects. Many low-risk patients may not require cardioprotection.

For patients with more established cardiotoxicity (LVEF <40%), particularly symptomatic patients, the treatment should follow the ESC Heart Failure guidelines and include the use of MRA [9]. Spironolactone increases the levels of progesterone and possibly oestrogen, and spironolactone should be avoided and eplerenone selected in patients with oestrogen-sensitive cancers, e.g., ER+ breast cancer when an MRA is indicated.

LV dysfunction early following cancer treatment (within the first 12 months)

Symptomatic ventricular dysfunction is the last stage of cardiotoxicity. It can be potentially irreversible and every effort should be made to avoid reaching this late stage.

The time point at which cardiotoxicity becomes clinically apparent varies widely and depends upon multiple variables including the cancer therapy, total dose and duration and baseline CV risk, including baseline LV function and pre-existing CV disease.

Anthracycline-induced cardiotoxicity is most likely a progressive phenomenon with declining LVEF on serial echocardiography. One large cohort study systematically assessed LV function with echocardiography in 2,625 cancer patients receiving anthracycline chemotherapy. New LVD following anthracycline chemotherapy was detectable in 9% of patients, with 95% of cases detectable within the first year of treatment [25]. This was the subclinical stage (category 5 cardiotoxicity); hopefully, intervening with appropriate treatment (ACEi and/or β-blockers) will prevent the development of future HFrEF.

Therefore, it is reasonable to offer all patients receiving anthracycline chemotherapy a cardiac assessment including echocardiography at 12 months after completing therapy, with additional 3- and 6-month assessments in higher-risk patients.

Long-term cancer survivors and late HF

All patients receiving cardiotoxic cancer treatments should be informed that they should seek medical attention if they develop possible symptoms of cardiovascular disease. They should highlight their previous cancer treatment to the doctor who reviews them. Medical professionals should specifically enquire about previous cancer diagnoses and treatment in patients presenting with new LVD and HF, particularly in the absence of other causes. The medical community, including primary care, emergency medical departments and cardiologists, should be educated that late effects including heart failure may develop many years after cancer treatment. When patients present late (>12 months) and particularly >10 years after treatment, the clinical HF syndrome is often more resistant and refractory to standard HF treatments, perhaps reflecting a more established and permanent myocardial injury [2].

Treatment should follow HF guidelines with ACEi, β-blockers and MRAs [9]. In HF patients with refractory HF despite triple medical therapy, more advanced guideline-directed treatments may be required including cardiac resynchronisation therapy (CRT), ivabradine and/or sacubitril-valsartan [9]. Finally, advanced mechanical HF therapies including LV assist devices and cardiac transplantation may be considered in survivors, particularly if they are more than 5 years from completion of cancer treatment without evidence of recurrence and designated cured of cancer by their oncologist.

There are no trial data or guidelines regarding the nature and timing of routine long-term follow-up beyond 12 months of adult cancer patients who received potentially cardiotoxic cancer therapies. One important question is whether follow-up is required for all patients or just high-risk individuals or those who have received cancer treatments most likely to result in permanent cardiac injury (anthracycline chemotherapy, radiation to the heart, patients with troponin rises during molecular targeted therapies). Studies in paediatric cancer survivors have shown that cardiac biomarkers and NPs have prognostic and diagnostic value in long-term follow-up, but their value varies in different timeframes from the completion of cancer treatment [26-28]. The single most important risk factor for development of HF in paediatric cancer survivors is the lifelong cumulative dose of anthracycline and radiotherapy. One surveillance model has been recommended by an international expert panel for long-term follow-up for survivors of paediatric cancers who received anthracycline chemotherapy, radiation to the heart field or both (29). Survivors of childhood cancer who have received a total cumulative dose of doxorubicin (or equivalent) of ≥250 mg/m2 or ≥35 Gy radiotherapy to the chest are considered high risk. Survivors with moderate doses of anthracyclines (≥100 mg/m2) and chest radiotherapy (≥15 Gy) are also considered high risk. Lifelong surveillance with 5-yearly cardiac assessment including echocardiography is recommended from 2 years after completing treatment [29]. This may also be considered in patients with moderate doses of anthracycline chemotherapy or chest radiotherapy.

This paediatric model with long-term surveillance could be appropriate for younger adult patients, e.g., Hodgkin’s lymphoma patients who have received ABVD chemotherapy and/or mediastinal radiotherapy. However, it is important that any surveillance has actionable management plans if abnormalities are detected, and that patients are counselled appropriately so that they are not left psychologically “traumatised” by the risk of cardiac problems after being cured of cancer.

One variable which should be considered when developing a surveillance strategy for adult cancer survivors is the date of the cancer treatment and modalities and doses administered. For example, there has been a serial reduction in the dose of radiotherapy used in Hodgkin’s lymphoma and breast cancer over the last 30 years, plus the implementation of heart-sparing strategies in breast cancer patients in the last 15 years. Therefore, cancer survivors treated between 1950 and 1999 are likely to have the highest risk due both to the time since treatment and to the higher doses of anthracycline chemotherapy and radiotherapy used in that era.

There is a growing recognition that a subgroup of cancer survivors develops HF with relatively low doses of anthracyclines and without conventional risk factors, i.e., a low baseline CV risk. This raises the possibility of genetic susceptibility of cardiotoxicity in these patients. In the future, genetic testing may also reclassify a subgroup of low-risk patients into a higher risk category where long-term surveillance may be appropriate. Appropriate trials are required to increase the evidence for long-term cardiac surveillance in cancer patients, including appropriate studies regarding the modality (biomarkers, imaging, or both), frequency of review for different risk categories, and cost-effectiveness [30].

Conclusions

Heart failure caused by cardiotoxic cancer therapies is a growing problem due to the ageing population, increasing baseline CV risk of patients at the time of cancer diagnosis and the growing cardiotoxic profile of several modern targeted cancer therapies and their effectiveness. There is a large population of people cured of cancer or living with their cancer controlled by chronic treatment. This represents an opportunity for modern cardiology with baseline CV risk assessment, surveillance using cardiac biomarkers and modern cardiac imaging, cardioprotection targeted at appropriate patients, and long-term surveillance strategies for high-risk survivors, in order to prevent this growing HF epidemic and increase survival free of cancer and HF.