Chronic obstructive pulmonary disease and comorbidities in heart failure: the next frontier of sodium–glucose co‐transporter 2 inhibitors?

This article refers to ‘Effects of dapagliflozin in heart failure with reduced ejection fraction, and chronic obstructive pulmonary disease: an analysis of DAPA-HF’ by P. Dewan et al., published in this issue on pages 632–643. The term ‘comorbidity’ is used to identify cardiovascular and non-cardiovascular diseases that are present in patients with heart failure (HF), interfere with care and impair outcomes. It is currently debated to which extent these comorbidities should be treated as part of HF management. In HF with preserved ejection fraction (HFpEF), for which effective therapeutic options are lacking, treatment of comorbidities is in the spotlight.1 By contrast, a number of drugs are indicated for HF with reduced ejection fraction (HFrEF).1 Historically, these medications targeted neurohormonal imbalances; in the last years, however, sodium–glucose co-transporter 2 inhibitors (SGLT2-i) have widened the spectrum of HFrEF medical therapy. In the Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure (DAPA-HF) trial, the first randomized controlled trial testing a SGLT2-i in HFrEF, dapagliflozin on top of established background neurohormonal regulation significantly reduced the risk of worsening HF or cardiovascular death as compared with placebo.2 These results have opened new horizons in HF, with realistic perspective for a new pillar in HFrEF pharmacotherapy.3 Nonetheless, the comorbidity burden in HF remains an unmet need.4 Chronic obstructive pulmonary disease (COPD) is present in up to 30% of patients with HF,5 and is independently associated with a significant short-term increase in hospitalizations6 and long-term increase in all-cause mortality.7 The mechanisms of these associations are not completely understood, but a significant underuse of guideline-recommend HF medications in patients with concomitant COPD has been recognized.5 This is likely multifactorial, with cardiologists being satisfied with individual drug prescription but incomplete titration, poor patient tolerance and/or adherence to HF drugs and only symptomatic benefits of main COPD-specific therapies. Clinical trials may serve as excellent source of clinically relevant information to better understand and tackle the burden of COPD in HFrEF. An article published in this issue of the Journal presents the work by Dewan and colleagues,8 who investigated the interaction between dapagliflozin and COPD on clinical outcomes in patients enrolled in DAPA-HF. This indeed is the first trial specifically reporting about the effects of a SGLT2-i in COPD, although only in a study subgroup. Of the 4744 HFrEF patients included in DAPA-HF, 585 (12.3%) had a history of COPD. Compared to those without COPD, they had worse symptoms and greater functional limitation at study enrolment, and a higher risk of HF hospitalization and death from any cause over a median follow-up of about 2 years. Notably, only chronic kidney disease and diabetes were comorbidities conferring an excess risk comparable to that of COPD. The benefit of dapagliflozin on all pre-specified outcomes was consistent in patients with and without COPD8 (Figure 1). The Authors should be congratulated for driving the attention of the cardiology community once again on this insidious comorbidity, and for demonstrating the efficacy and tolerability of dapagliflozin in HFrEF patients with COPD. Along with their work, several important issues about comorbidities and COPD in HFrEF, as well as about SGLT2-i effects beyond those on the cardiovascular system, are worth being discussed. The analysis by Dewan and colleagues suffers from an issue that is shared with most HFrEF clinical trials, i.e. a detailed definition and description of comorbidities – including COPD – was not planned ahead as part of study design.7 Unfortunately, in trials but also in clinical practice this happens too often. This holds true especially for diagnosis such as COPD that requires validation using a pulmonary function test,9 which is not easily accessible to cardiologists and is currently performed in only about a third of ambulatory HF patients with COPD.5 Like in other trials, the DAPA-HF investigators used a simple tick-box to qualify patients into the COPD subgroup (Figure 1). Reassuring, however, is the rate of former/current smokers that was significantly higher in patients with COPD (82.0% vs. 55.5%). In several previous HF trials, COPD was assigned only to those patients treated with drugs for COPD,5, 7 and about half of the patients (46.3%) with COPD enrolled in the DAPA-HF used at least one form of inhaled therapy.8 However, adherence to and persistence on COPD medications are known to be low, with about half of spirometry-confirmed COPD patients not being able to maintain their initial COPD therapy at 1 year.10 Clearly, the diagnostic inaccuracy and the lack of comorbidity characterization in HFrEF research is driven by trial feasibility and straightforward patient inclusion that can be limited by too complex inclusion criteria. However, this approach increases the toll of indirect and incomplete information, which often cannot easily translate to clinical practice guidance. Even a comorbidity that one would consider as clearly identifiable, such as cancer, has invariably been defined by vague criteria, or has not been defined at all.11, 12 Strikingly, limited improvement in the reporting of comorbidities in HF trials has been made over the last 15 years. In fact, it has been suggested that the increase in the prevalence of cardio-metabolic conditions such as hypertension, diabetes and hyperlipidaemia observed over time is the result of a selection bias against enrolling subjects with comorbidities with greater competing risk of non-cardiovascular death, such as COPD and cancer.13 Although with challenges that must be acknowledged, it is advisable that future HF clinical trials consider adopting more standardized methods of reporting comorbidities, in an effort to provide strong evidence for the care of real-world multi-morbid HF patients. While awaiting for the release of new Heart Failure Association guidelines, dapagliflozin and empagliflozin have been recommended to reduce the risk of HF hospitalization and cardiovascular death in symptomatic patients with HFrEF already receiving guideline-directed medical therapy, regardless of the presence of diabetes.2, 14-16 Dewan and colleagues now show that the beneficial effects of dapagliflozin are maintained in the presence of COPD, further strengthening the use of SGLT2-i in HFrEF. In DAPA-HF, the contribution of infections and other causes to death was greater in COPD than in non-COPD patients, as expected (Figure 1). It is reasonable that non-cardiovascular causes, particularly infections, also accounted for many episodes of worsening HF in subjects who had COPD, although the reasons for worsening HF were not collected in the DAPA-HF. Thus, dapagliflozin improved the outcomes in a subset of participants in DAPA-HF especially vulnerable to events driven by non-cardiovascular factors. From the epidemiological (and clinical) standpoint, this observation lends additional support to the claim that SGLT2-i should be prescribed to the broad HFrEF population.17 Concerning the mechanism(s) of action, the favourable effect of dapagliflozin in COPD, as compared with placebo, might be also secondary to the induction of diuresis with improvement of COPD-related pulmonary hypertension and right chamber overload (Figure 1). Indeed, in a recent double-blind, placebo-controlled, randomized trial involving 70 patients with HFrEF and diabetes, empagliflozin reduced pulmonary capillary wedge pressure by 2.4 mmHg on average over 12 weeks of treatment,18 and it can be hypothesized that pulmonary and right heart congestion similarly improve upon SGLT2-i therapy. A diuretic effect was shown in further studies.19-21 Nonetheless, the so-called cor pulmonale develops only in some patients with COPD. Hence, other factors may underlie the benefits of dapagliflozin in the subgroup of HFrEF patients with COPD. The knowledge about the actions of SGLT2-i is continuously evolving, and several pleiotropic, non-cardiovascular activities have been proposed.22 In principle, these could also explain the advantage provided by dapagliflozin in participants in DAPA-HF who had COPD; this speculation is intriguing and deserves being addressed by future investigations. Finally, the results presented in this issue of the Journal raise exciting questions about the potential use of dapagliflozin and other SGLT2-i in HFpEF with COPD and COPD with normal left ventricular function, but cor pulmonale. Partial answers might be obtained from Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure (DELIVER, NCT03619213) and EMPagliflozin outcomE tRial in Patients With chrOnic heaRt Failure With Preserved Ejection Fraction (EMPEROR-Preserved, NCT03057951),23, 24 although – again – patient phenotyping will likely be poor. Dedicated pathophysiological studies are warranted. As of today, cardiologists know that SGLT2-i are a valid option for their patients who have COPD in addition to HFrEF. Importantly, this does not imply that disease-modifying drugs can be omitted and substituted for by SGLT2-i. Neurohormonal modulators are still the backbone of HFrEF medical therapy, and SGLT2-i represent an additional line of treatment to be added with confidence (Figure 1). There is a significant gap in the optimization of HFrEF treatment in patients with concomitant COPD, and beta-blocker underutilization in this population remains an important issue with potential adverse consequences.5 This is not apparent when analysing trial populations such as the one from the DAPA-HF, in which 92.3% of COPD patients were taking beta-blockers, with more than half of them being at >50% of target dose.8 Guidelines state that COPD is not a contraindication to beta-blockers, and that these medications should at least be started in COPD patients with HFrEF.1 Clinicians' inertia rather than patient's intolerance is thought to be the main barrier to medical therapy optimization in these patients in the real world.5, 25 In these high-risk individuals, new therapies such as SGLT2-i should be considered to date as addition to conventional treatment and, only occasionally, as an alternative in patients that are genuinely intolerant to recommended medications (Figure 1). By pursuing neurohormonal modulation and, on top of it, adding SGLT2-i, outcomes will improve for the population with HFrEF and comorbid COPD. The next frontiers could be then HFpEF with COPD and COPD without left ventricular dysfunction, but with right ventricular failure. For these, mostly symptomatic inhalator therapy is available; whether SGLT2-i can confer (more than symptomatic) benefit needs to be addressed and we hope to welcome focused trials anytime soon. Conflict of interest: none declared.