In pilot studies the investigators have shown that subcutaneous adipose tissue (SAT) from patients with reduced ejection fraction heart failure (HFrEF) and type 2 diabetes mellitus (T2DM) is dysfunctional. Endothelial cells from the adipose tissue from these patients are senescent and have deleterious effects on healthy human subcutaneous adipocytes, including increasing expression of IL-6 (gene and protein) and reducing glucose uptake. Digoxin, a well-established treatment for HFrEF, selectively clears these senescent endothelial cells and prevents adipocyte dysfunction. This study will examine the effect of digoxin on adipose tissue on the burden of senescent cells.

Type 2 diabetes mellitus and advanced heart failure: a lethal combination. T2DM is a progressive disorder affecting over 400 million people worldwide, a figure which will reach 600 million by 2035. Globally, T2DM causes one cardiovascular death every 12 seconds and accounts for 8.4% of all deaths in people between the age of 20-79. A common cardiovascular sequela of T2DM is chronic heart failure (CHF) due to reduced left ventricular ejection fraction (HFrEF). Despite 'state of the art' therapies, 1 in 7 patients with the combination of T2DM and HFrEF will be dead or admitted to hospital within 2 years. More than 30% of patients with HFrEF suffer from T2DM, with incidence of T2DM rising to half in those hospitalised for heart failure.

Multiple mechanisms have been proposed to explain the increased incidence of HFrEF in patients with T2DM, including the hypertrophic influence of insulin, the adverse effects of hyperglycaemia, increased oxidative stress, and neurohumoral activation. More recently, increasing attention has been paid to insulin resistance as a driver of HFrEF. Resistance to insulin-mediated glucose uptake in its canonical target tissues (liver, muscle and fat) is well-established as a critical perturbation contributing to the progression of T2DM, and as a result, reversing or overcoming this has become a well-accepted therapeutic paradigm.

A study of Swedish individuals found that insulin resistance predicted the development of HFrEF independently of established risk factors. In another study, plasma levels of proinsulin (a marker of insulin resistance) were higher in people who developed HFrEF 20 years before diagnosis. There is also accumulating evidence that HFrEF augments the risk of developing T2DM and drives its progression, with a stronger relationship between the two in those with more severe HFrEF. The investigators have shown that the deleterious effect of T2DM on HFrEF outcomes is similar in patients with and without ischaemic aetiology, supporting the hypothesis that metabolic dysfunction may at least in part drive the progression of HFrEF in patients with the additional burden of T2DM. These studies raise the possibility of a vicious cycle, whereby HFrEF induces metabolic dysfunction which in turn accelerates cardiac dysfunction.

Subcutaneous adipose tissue microvascular endothelial cell senescence in patients with type 2 diabetes mellitus and HFrEF. Subcutaneous adipose tissue (SAT) is the largest AT depot in humans and the most important in terms of its contribution to glucose and lipid homeostasis. Evidence from large scale human genetic, clinical, and preclinical studies, supports the hypothesis that in obese insulin resistant humans dysfunctional SAT is less efficient at glucose uptake and lipid storage, which as a result leads to impaired glucose tolerance and deposition of lipids in tissues ill-equipped to deal with this challenge. While Shimizu et. al. showed that AT dysfunction in murine models is mechanistically linked to the development and progression of HFrEF, very little is known about SAT function in humans with HFrEF. The investigators have performed a detailed examination of SAT from humans with HFrEF with and without T2DM. SAT from patients with HFrEF and T2DM (despite similar body mass index to those without T2DM) had larger adipocytes, increased fibrosis and reduced vascular sprouting in vitro. They took this work further showing that SAT MVEC had increased production of the free radical superoxide and reduced population doubling rate. Stimulated by these findings, which raised the intriguing possibility that SAT MVEC may be adopting a senescent phenotype, the investigators quantified hallmarks of cellular senescence. SAT MVEC from patients with HFrEF and T2DM had: 1) Reduced proliferation in an EdU assay a sensitive method to detect proliferation in live mammalian cells; 2) Increased expression of senescence associated β-Galactosidase; 3) Reduced mitochondrial ATP generation. A sine qua non of senescent cells is secretion of a complex combination of factors collectively referred to as the senescence associated secretory phenotype (SASP). To assess this, the investigators took conditioned media from MVEC and using a cytokine multiplex assay, demonstrated higher concentrations of IL-6 relative to total cell protein (3.4 [0.8] vs. 1.6 [0.4]; P<0.05) in MVEC conditioned media from patients with HFrEF and T2DM versus HFrEF without T2DM.

Senescent MVEC communicate via SASP with healthy adipocytes to induce a pro-inflammatory phenotype. Through SASP, senescent cells are thought to induce an inflammatory state that can provoke local tissue damage leading to a persistent, self-reinforcing, inflammatory microenvironment. The bulk of evidence demonstrating deleterious cell-cell communication facilitated by SASP has been generated using animal models. For example genetically induced EC specific senescence in mice leads to adipocyte and whole-body insulin resistance. The investigators developed a co-culture system to allow them to examine human SAT MVEC to adipocyte communication. When MVEC from patients with HFrEF and T2DM were co-cultured with healthy human subcutaneous adipocytes, IL-6 mRNA and protein were increased, and glucose uptake decreased, compared to adipocytes cultured with MVEC from HFrEF patients without T2DM. IL-6 is a complex pleiotropic cytokine with multiple effects that may influence glucose homeostasis. Consistent with this IL-6 has a negative effect on adipocyte glucose uptake. It is thought that up to 35% of circulating IL-6 derives from AT. With this in mind the investigators measured serum IL-6 concentration in patients with HFrEF and T2DM and found it to be significantly higher compared to patients with HFrEF alone (HF 3.9pg/ml [0.5] c.f. HFDM 8.1pg/ml [1.2]; P<0.01). These data demonstrate the possibility of a deleterious signalling circuit between senescent MVEC and SAT adipocytes, which in turn may contribute to systemic metabolic dysregulation.

Senolytics. Drugs targeting the deleterious impact of senescent cells can be broadly described as: i) senomorphic agents that target pathological SASP signalling and ii) senolytic agents that specifically eliminate senescent cells. The first published study in humans of senolytic agents examined the effect of Dasatinib plus Quercetin in patients with idiopathic pulmonary fibrosis, and demonstrated that Dasatinib plus Quercetin improved a number of markers of physical performance in patients with this disorder. Recently Hickson et. al. showed that 11 days after a short course of Dasatinib plus Quercetin in individuals with advanced T2DM, senescent cell burden in SAT was reduced with a commensurate favourable effect on SASP.

Digoxin induced senolysis: a new role for an old drug. Recent reports raise the intriguing possibility that the cardiac glycoside digoxin may have senolytic actions. Consistent with this, digoxin reduced β-galactosidase, enhanced proliferation and reduced IL-6 secretion by >15%. In SAT MVEC from patients with HFrEF and T2DM. When SAT MVEC from patients with HFrEF and T2DM pre-treated with digoxin were co-cultured with healthy subcutaneous adipocytes, normal levels of glucose uptake were achieved.

Clinical trials of digoxin in patients with HFrEF. It is well established that digoxin safely improves haemodynamics, symptoms and the deleterious neurohumoral profile of patients with advanced HFrEF. In the DIG trial, digoxin safely reduced the risk of HF death and hospitalisation at serum concentrations between 0.5-0.9ng/mL which is usually achieved by a daily dose of 125mcg. The RADIANCE study which compared continuing digoxin therapy with its withdrawal from background ACEI and loop diuretic therapy showed an increased risk of HF within 14 days in those in whom digoxin was withdrawn. The DIG trial had limited biochemical investigations and did not include HbA1c or fasting/non-fasting glucose, but a recent subgroup analysis showed data supporting the safety of digoxin in patients with HFrEF and a clinical diagnosis of diabetes per se. Despite these data, digoxin use has fallen substantially over the last 2 decades; once prescribed to 80% of patients with HFrEF, it is now taken by less than 40% of patients. The investigators' data raise the intriguing possibility that digoxin may be used as a senolytic agent to normalise SAT dysfunction in patients with advanced HFrEF and T2DM.

Hypothesis: Digoxin administered to patients with HFrEF and T2DM reduces microvascular endothelial cell senescence and improves subcutaneous adipose tissue function.

Fundamental aims:

1. To examine the effect of digoxin on SAT senescent MVEC burden in patients with HFrEF and T2DM.
2. To examine the effect of systemic digoxin on phenotypic hallmarks of SAT MVEC senescence and SAT dysfunction.

Inclusion Criteria:

• Aged ≥18yrs,
• HFrEF (LVEF<40%)
• T2DM (taking anti-diabetic medication, fasting plasma glucose ≥7.0 mmol/L and/or a serum HbA1c >48mmol/L
• On optimal medical therapy,
• Able/prepared to give informed written consent.

Exclusion Criteria:

• Significant cognitive impairment,
• Important co-morbidity limiting ability to comply with study procedures,
• Hyperkalemia (>5.5mmol/L)
• eGFR<30ml/min/1.73m2
• Current/previous (<6m) participation in other studies.

Upon reasonable request, anonymised data can be shared securely.

Participants will take either digoxin or placebo.