Growth Differentiation Factor-15 Reveals the Dark Side of Heart Failure

This article refers to ‘Growth differentiation factor-15 is not modified by sacubitril/valsartan and is an independent marker of risk in patients with heart failure and reduced ejection fraction: the PARADIGM-HF trial’ by N. Bouabdallaoui et al., published in this issue on pages xxx. Over the past three decades, new drug and device therapies have substantially improved the prognosis of patients with heart failure with reduced ejection fraction (HFrEF).1 The Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure Trial (PARADIGM-HF) marks the latest chapter in this extraordinary success story.2 In PARADIGM-HF, the combined angiotensin receptor–neprilysin inhibitor sacubitril/valsartan was shown to be more effective than the angiotensin-converting enzyme inhibitor enalapril in reducing the composite primary endpoint of cardiovascular death or heart failure (HF) hospitalization, its individual components, and all-cause mortality.2 Despite these therapeutic advances, mortality and hospitalization rates in HFrEF remain relatively high: over the course of 27 months, 13.3% of patients in the sacubitril/valsartan group died from cardiovascular causes and 12.8% were hospitalized for worsening HF, outcomes that highlight the need for additional HF-targeted therapies.2 Patients with HFrEF in the community are older and tend to have more cardiovascular events and greater cardiovascular mortality than patients enrolled in clinical trials.3-5 In addition, many of these patients have multiple non-cardiovascular conditions that may trigger hospitalizations and influence overall survival. Further reductions in mortality and hospitalizations in HFrEF will require a better characterization of these adverse non-cardiovascular outcomes and improved treatment strategies to address HF co-morbidities.3-5 There is a continued interest in using biomarkers to obtain pathophysiological insight and improve the management of HF. The key process in HF is cardiac remodelling in response to chronic disease stresses acting on the myocardium. Biomarkers such as the natriuretic peptides that are released from the failing heart reflect the remodelling process and its impact on cardiac function. Because HF is a systemic disease, biomarkers related to extracardiac pathologies may provide incremental information.6 In this issue of the Journal, Bouabdallaoui and colleagues have measured growth differentiation factor-15 (GDF-15) in almost 2000 patients enrolled in PARADIGM-HF. Using a commercially available ELISA, serum GDF-15 measurements were obtained at baseline, 1 month, and 8 months after randomization.7 GDF-15 concentrations determined with this ELISA correlate reasonably well with those measured by a CE marked, fully automated electrochemiluminescence immunoassay that has recently been approved for clinical use.8 Confirming previous studies in HFrEF,9-11 most patients presented with elevated levels of GDF-15 (> 1200 ng/L). Over the course of 8 months, GDF-15 concentrations remained remarkably stable in the overall study population despite quite some inter-individual variability (some patients doubled their GDF-15 and other patients reduced their levels by half). Higher baseline levels and increasing GDF-15 levels over time were associated with greater risks of fatal and non-fatal cardiovascular events during follow-up, even after adjustment for clinical variables, N-terminal pro-B-type natriuretic peptide (NT-proBNP), and high-sensitivity cardiac troponin T. In a reclassification analysis, GDF-15 offered incremental prognostic information on the primary endpoint when added to clinical risk markers and the two cardiac biomarkers, emphasizing its independent association with cardiovascular events. The relation between GDF-15 and adverse cardiovascular outcomes has stimulated a search for therapies that lower GDF-15 and modify the risk associated with high levels of this biomarker. In patients with acute HF, serelaxin has been shown to lower GDF-15 levels indicating that GDF-15 may respond to pharmacological interventions.11 With all the limitations of a single study, this observation may be consistent with the pathobiology of GDF-15 (see below) and the proposed anti-inflammatory and antioxidant effects of serelaxin that may limit end-organ damage in acute HF.12 In the Valsartan Heart Failure Trial (Val-HeFT) in HFrEF, valsartan treatment reduced BNP levels but not GDF-15.13, 14 In PARADIGM-HF, sacubitril/valsartan was more effective than enalapril in reducing NT-proBNP but had no discernible impact on circulating GDF-15.7, 15 Given GDF-15's lack of response to valsartan and sacubitril/valsartan in HFrEF, Bouabdallaoui and colleagues conclude that GDF-15 represents ‘an integrated biomarker of multiple co-morbidities rather than a specific reflection of cardiovascular health’.7 Growth differentiation factor-15 is a stress-responsive cytokine that can be produced by cardiovascular and non-cardiovascular cell types under pathological conditions.16 The GDF-15 promoter contains two p53 transcription factor binding sites that are required and sufficient for the induction of GDF-15 expression.17 Activation of p53 is a fundamental cellular response to inflammation, oxidative stress, cellular senescence, and oncogene activation.18 The circulating levels of GDF-15 reflect these cellular stressors, which are prevalent in many chronic disease states, including HF. In HFrEF, GDF-15 concentrations increase with HF severity as reflected by symptoms and increased cardiac biomarkers. At the same time, GDF-15 levels are higher in HFrEF patients with co-morbidities such as renal dysfunction, cachexia, or anaemia.16 We therefore propose that GDF-15 reflects co-morbidities and cardiovascular disease. Notably, among all investigated outcomes in PARADIGM-HF, GDF-15 and its changes showed the closest independent association with HF-related mortality.7 GDF-15 seems to reflect a pathophysiological process in HFrEF that triggers cardiovascular events but is not addressed by some of the most powerful currently available HF therapies. Wouldn't it be great to identify a therapy that addresses this residual cardiovascular risk, at least some of which appears to be reflected by GDF-15? The current study reports on another interesting observation: GDF-15 and its changes over time were strongly associated with non-cardiovascular mortality.7 The authors did not provide a model adjusting for age and co-morbidities, known confounders of this relationship.19 Obviously, non-cardiovascular death has many different causes and no biomarker will predict them all. In PARADIGM-HF, cancer and infections accounted for over half of the non-cardiovascular deaths,20 but the authors missed the opportunity to explore what specific causes of non-cardiovascular death are predicted by GDF-15.7 In apparently healthy individuals and patients with coronary artery disease, GDF-15 has emerged as a unique and independent predictor of incident cancer.16 Future studies should explore if GDF-15 helps identify HF patients at risk of developing cancer.21 In conclusion, the interesting study by Bouabdallaoui and colleagues indicates that GDF-15 provides insight into HFrEF's residual cardiovascular and non-cardiovascular risks − the less-well explored, ‘dark side’ of the disease. The search for therapies that address the risks associated with high GDF-15 levels in HFrEF continues. With a better understanding of the upstream disease pathways reflected by GDF-15, new treatment targets may emerge. Considering the marker's pathobiology, patients with increased GDF-15 may potentially benefit e.g. from anti-inflammatory, antioxidant, or antiaging therapies. Some of these therapies may ultimately target HF and some of its co-morbidities.22-24 Conflict of interest: K.C.W. is named as co-inventor on patents regarding the use of GDF-15 for cardiovascular applications and has a licensing contract with Roche Diagnostics, the manufacturer of the recently launched Elecsys® GDF-15 assay.