‘Cardiotoxicity’: time to define new targets?

This editorial refers to ‘Classification, prevalence, and outcomes of anticancer therapy-induced cardiotoxicity: the CARDIOTOX registry’†, by J. López Sendón et al., on page 1720. Cardiovascular disease has become an important source of morbidity and mortality in cancer survivors.1 Cardiac imaging and myocardial biomarkers are useful tools for identifying early myocardial injury.2 There has been increasing interest in the use of cardiac imaging to detect cardiac injury as it provides direct assessment of myocardial function.3 In this issue of the European Heart Journal, López Sendón et al. report the results of the CARDIOTOX registry.4 This is a prospective study of 865 patients, scheduled for anticancer therapy related to moderate or high risk of cardiotoxicity, with a median follow-up of 24 months. Cardiotoxicity was defined as new onset or worsening of myocardial damage or ventricular function from baseline during follow-up, and it was present in 37.5% of the sample. In this study, cardiotoxicity was more prevalent in older patients (mean age in the whole sample was 54.7 years old, and 61.8 years old for the group with severe cardiotoxicity, P = 0.001), in line with previous data,5 and there was a higher prevalence in males. The authors also explored the prevalence of abnormal biomarkers and left ventricular disease parameters using a recently published classification, based on the cardio-oncology consult experience at the Royal Brompton Hospital.6 There have been many different definitions of cardiotoxicity7–9 for many years, and there have also been changes in the concept over time. The most widely recognized diagnosis of cardiotoxicity is based on changes in left ventricular ejection fraction (LVEF). A drop below a certain level or an absolute decrease in LVEF can be interpreted as an early signal of possible cardiotoxicity, but thresholds for clinical decisions may vary depending on different authors and guidelines. LVEF measurement is subject to important intra- and interobserver variability.10 The diagnosis of heart failure is clinical, but it also involves assessment and monitoring of LVEF. Any grade of cancer therapy-induced myocardial dysfunction should be taken into consideration in order to minimize adverse remodelling. In the current study, the cut-off point for LVEF to be considered as significant is 50%. However, according to the recommendations of the ASE/EACVI consensus on the role of cardiac imaging in oncology patients,11 this cut-off point should be 53%. Conversely, the cut-off point of 53% is not formally accepted by the ESC community, and instead LVEF <50% is preferred.12 This lower threshold is in line with the definition of cardiotoxicity commonly used in registries and trials in patients with cancer. The ESC Position Paper recommends that, if LVEF decreases 10% to a value below the lower limit of normal (considered as an LVEF >50%), angiotensin-converting enzyme inhibitors (or angiotensin receptor blockers) in combination with beta-blockers are recommended to prevent further left ventricular dysfunction or the development of symptomatic heart failure, unless contraindicated.12 In the study of López Sendón et al., it is considered that asymptomatic patients with LVEF ≥50% with elevated biomarkers or a 10% decrease of LVEF to an LVEF <53% should be considered as having mild cardiotoxicity. However, in our opinion, a 10% decrease of LVEF to an LVEF <53% (or 50%) is an alarm sign. These patients should be followed very carefully, and therefore they might be included in the severe category. This different consideration could lead to changes in the cardiac management of these patients and, possibly, even affect the decision for the discontinuation of cancer therapy. In this study, a decrease in LVEF >10% with a final LVEF <50% or 53% was not a sensitive marker to predict mortality at 2 years follow-up and, in fact, as an isolated criterion, an LVEF >40% showed no relationship with a worse outcome. So, it is controversial whether this should be used as a criterion to discontinue cancer therapy, because cancer treatment interruptions certainly increase cancer mortality. There is not a strong agreement on this cut-off point and a significant drop of LVEF during cancer therapy represents an important warning to initiate cardioprotection. Whether it means that oncological treatment should be stopped remains unclear. This leads to raising the question of whether LVEF should be used as the gold standard to define left ventricular function. It is time to really consider whether ejection fraction should be replaced by analysis of left ventricular mechanics in order to better assess systolic function. Obviously this will impact many aspects in cardiology outside cardio-oncology, such as definitions of heart failure based on ejection fraction or even criteria used to intervene in certain valvulopathies. No doubt this would be a major change in decision-making in cardiology. In a recent article by Santoro et al.,13 116 consecutive female patients with HER2-positive breast cancer undergoing chemotherapy were enrolled. Patients underwent an echo-Doppler exam with determination of LVEF and global longitudinal strain (GLS) at baseline and every 3 months during cancer therapy. All patients developing a GLS drop of >15% or an LVEF reduction of <50%, initiated cardiac treatment (ramipril and carvedilol). In the 99.1% of patients successfully completing the cancer treatment, GLS and LVEF were significantly reduced at the end of follow-up. Combined subclinical and overt cardiac dysfunction was diagnosed in 23.3% of the sample. By cardiac treatment, LVEF recovery was observed in most of them, and so they could finish their cancer therapy. With the definition of cardiotoxicity used in the CARDIOTOX registry, it was found that there was a high incidence (37.5%) of patients with worsening ventricular function during high-risk chemotherapy. However, the prevalence of severe cardiotoxicity was low, and it might be due to the exclusion of patients with previous heart failure and severe left ventricular disease. With the available data, one might think that these patients should have been included in the study, because they have, in fact, a higher risk of developing cardiotoxicity.14 Also the management and close tracking in an integrated cardio-oncology service could have improved the outcomes of the patients in the registry.6 In previous studies, the incidence of cardiotoxicity varied depending on the selection of patients, cancer therapies, and the definition of cardiotoxicity.15 It is likely that early identification and treatment of cardiovascular risk factors as well as asymptomatic forms of cardiotoxicity may delay further deterioration of ventricular function.16 Regarding the mortality, 6.2% of the sample died during follow-up (cardiovascular death occurred in 0.4%). This is a low rate of mortality, in contrast to other similar studies.1 As expected, mortality was higher in the group with severe cardiotoxicity, with a strong prognostic relationship (the mortality rate in the group with severe cardiotoxicity was 22.9 deaths per 100 patients-year vs. 2.3 deaths per 100 patients-year in the other groups), but the presence of severe cardiotoxicity was rare. The study by López Sendón et al. provides a clear vision of the definition and the clinical implications of cardiotoxicity, with its unanswered questions (Figure 1). The next goal is to establish the relationship of different grades of cardiotoxicity and the clinical outcomes. The elaboration of a clinical score to determine the risk of severe cardiotoxicity and, according to this, refine the follow-up and treatment of these patients, will be the key of the cardio-oncology field for the present and near future. The circle of cardiotoxicity. The circle of cardiotoxicity. Conflict of interest: J.L.Z. works in the same city and country as the authors, but he is not affiliated with the institution of the authors, or the same university. There are no relationships between hospitals, joint research investigations, or personal relations. A.P.S. has no conflicts to declare. The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.