Cancer Therapy and the Heart; The Necessity to Calibrate Risk

This article refers to ‘Baseline cardiovascular risk assessment in cancer patients scheduled to receive cardiotoxic cancer therapies: a position statement and new risk assessment tools from the Cardio-Oncology Study Group of the Heart Failure Association of the European Society of Cardiology in collaboration with the International Cardio-Oncology Society’ by A.R. Lyon et al., published in this issue on pages 1945–1960. The legacy of potentially deleterious effects on the heart due to the administration of cancer chemotherapies has been known for over 40 years. Anthracycline-based cancer therapy began in the 1960s and evidence of cardiac injury was noted in the 1970s.1, 2 Treatment of certain childhood cancers with anthracycline-based regimens led to a contracted heart, i.e. Grinch syndrome,3 while in adults, anthracycline-based regimens led to significant cardiac remodelling and in the most extreme circumstances, eventual heart failure.4 Anthracyclines inhibit DNA and RNA synthesis and create free radicals that further damage DNA and likely cardiomyocytes as well, but the exact basic mechanisms of harm remain under active investigation.5 The clinical phenotype is more certain — myocardial atrophy and a reduction in left ventricular mass typically years after the initial exposure. The subsequent heart failure can have devastating consequences and the effect of usual guideline-directed medical therapies (GDMT) is largely unproven or ineffective, especially if there is a delay in diagnosis and therefore treatment.6 Heretofore, active surveillance strategies represented best clinical guidance along with modulated dose exposure, particularly in the setting of a priori left ventricular dysfunction.7 Limited data suggest that early introduction of GDMT might mitigate injury but these observations are not yet rooted in incontrovertible evidence.6 Today, our concerns of associated cardiotoxicity after cancer therapy exposure have markedly evolved due to the fortunate increase in effective disease-modifying anti-cancer therapies. Never before has the armamentarium to treat cancer been as robust and expectations been as hopeful.8 Yet, with more agents, more evidence of cardiac toxicity emerges. In response, authors representing the Joint Heart Failure Association of the European Society of Cardiology and the International Cardio-Oncology Society (HFA-ICOS) have developed an important position statement: ‘Baseline cardiovascular risk assessment in cancer patients scheduled to receive cardiotoxic cancer therapies’.9 Beyond the legacy of anthracycline-based regimens, a number of new classes of cancer therapies are now associated with cardiotoxicity: HER2-targeted therapies; vascular endothelial growth factor inhibitors; targeted kinase inhibitors; proteasome inhibitors; immunomodulatory drugs including immune checkpoint inhibitors; RAF and MEK inhibitors; and androgen deprivation therapies. Each of these drug classes has the potential to induce a variety of cardiotoxicities including heart failure and myocarditis, significant hypertension, ischaemia and thromboembolic events, QTc prolongation and associated/other arrhythmias, metabolic syndrome and increased burden of known cardiovascular risk factors.7, 8, 10 Given the plethora of agents and the varied cardiotoxicities, there has been the necessity to aggregate these drug classes in one discussion that aligns each class with known cardiovascular complications and importantly identifies what is known about risk assessment and vulnerability to harm. The authors have adroitly executed this task. The discussion is developed as seven ‘proformas’ rather than guidelines which we applaud. In general terms, a proforma defines those minimum or universal requirements needed to execute certain functionalities. Thus, the seven proformas provided offer guidance in the cardiovascular risk assessment of those patients with indications for use of these several anti-cancer drug classes. Several aspects of this risk assessment scheme are especially laudable: the strident emphasis to not allow cardiovascular assessment to engender any delay in the administration of disease-modifying therapy for certain cancers; the strength of endorsement that all decision-making should occur in a shared decision-making model (between the cardiologist and oncologist) and should be executed by a disease management team with the patient fully informed in the decision-making process; and a very clear path of engagement based on treatment type with the emerging resource of cardio-oncologists — relying both on a unique skillset and a convenient portal of entry for further study.11 These several attributes are gratifying and speak to the excellent intentions of this collage of highly regarded experts in cardio-oncology. Yet, no risk assessment strategy is perfect and the present HFA-ICOS cardiovascular risk assessment tools have limitations. There are some (limited) predicate risk scoring tools at the intersection of cancer and cardiovascular disease (CVD) which are consistent with usual characteristics of risk assessment while others are largely dependent upon observational data (Table 1).12-22 The construct of the HFA-ICOS risk assessment strategy varies from traditional cardiovascular risk assessment tools: there is no derivation cohort and no validation cohort; the tools do not generate a finite ‘score’ that precisely calibrates risk; the event targets vary according to the drug class; and there is no defined suite of tools that allow for measures of risk enhancement or net reclassification. Consequently, as a more traditional risk assessment tool, the HFA-ICOS tools are in variance. However, for the intended purposes of this cardiovascular risk assessment scheme, this variation from the prototypical risk score algorithm for CVD is not unreasonable. The vagaries and complexity of cancer treatments and the full portfolio of cardiovascular effects make it unwieldy if not inconceivable to create a classical risk assessment tool. Heart failure All-cause mortality FRESCO (‘Spanish risk function of coronary and other CV events’) plus NT-proBNP19, 20 Moreover, the field of cardio-oncology would be greatly facilitated by discovery of cardio-protective strategies that allow for administration of today's more targeted anti-cancer molecular therapies while protecting the heart from direct injury. As we await these discoveries to better inform future proformas, the cardiovascular risk assessment tool developed by the HFA-ICOS is a needed and welcome addition to the field. These intuitive proformas will better inform everyday clinical decision-making, thus enabling recent discoveries of effective cancer therapies to be extended to more patients with freedom from cardiovascular consequences for those without pre-existing CVD; and a reduced risk for those with already established CVD. As cardio-oncology data emerge and become more robust, these current and future proformas would benefit from true risk calibration and might be enhanced with a machine learning process to better enable the prediction of cardiotoxicity.25 Certainty of the anti-cancer effect after any dose modification also needs to be validated. Finally, we must point out the most compelling aspect of cardio-oncology: the shared risk factors for both CVD and cancer.26 The incorporation of heart-healthy and ‘cancer-healthy’ lifestyle choices — e.g. no smoking; lean body mass; diets rich in fruits, nuts, vegetables and fish and low in saturated fats; and exercise — remain worthwhile messages that we must continue to promulgate. The best risk assessment tool for CVD or cancer yet developed remains a decrease in risk of disease development by early avoidance of risk. Conflict of interest: G.F. participated in committees of trials sponsored by the industry. The other authors have nothing to disclose.