Umeå University's logo

Background: Sodium-glucose cotransporter 2 inhibitors (SGLT2i) are a class of drugs that has been used in the treatment of type 2 diabetes mellitus (T2DM) since 2012. Later trials showed that SGLT2i-treatment was also beneficial for patients with heart failure (HF) with reduced ejection fraction (HFrEF), mildly reduced ejection fraction (HFmrEF) and preserved ejection fraction (HFpEF) regardless of diabetes status. SGLT2is are now considered a cornerstone treatment for HF regardless of left ventricle ejection fraction (LVEF), and usage is increasing year-on-year, including in patient groups that were excluded from the clinical trials that the indication and recommendations are based on, such as patients suffering from HF caused by transthyretin amyloid cardiomyopathy (ATTR-CM).

Aims: This thesis aimed to evaluate the basis for the universal adoption of SGLT2is in HF treatment over four papers which explore different facets of this subject.

I. Evaluate external validity of pivotal SGLT2i trials in patients with HFrEF by assessing eligibility in a real-world, all-comer HF cohort

II. Assess the prevalence of diabetes mellitus (DM) and prediabetes in a real-world, all-comer HF cohort

III. Conduct a pilot trial evaluating tolerability and effect of SGLT2i in patients with ATTR-CM

IV. Plan a multicentre randomized controlled trial (RCT) to further establish the role of SGLT2i-treatment in patients with ATTR-CM

Methods: In papers I and II, a cohort of 2433 patients made up from patients with at least one registered diagnosis of HF from the Heart Centre or Department of Internal Medicine at the University Hospital of Umeå between the dates 2010-01-01 and 2019-12-31 was used (the HF cohort).

In paper I, data regarding 681 patients in the HF cohort with a documented left ventricle ejection fraction (LVEF) of ≤40% (HFrEF) were reviewed with regards to the eligibility criteria of the DAPA-HF and EMPEROR-reduced trials, and the characteristics of this HF cohort were compared with the comparator trials.

In paper II, patients were divided into groups based on glycaemic profiles and HF phenotype. Prevalence of glycaemic status was compared between HF phenotypes, HF treatment compared between glycaemic groups, and T2DM treatment compared between HF phenotypes.

In paper III, 10 patients with ATTR-CM were recruited into a pilot trial set up as a n-of-1 series with 3 periods: a baseline period, an intervention period with dapagliflozin 10 mg o.d. and a withdrawal period. Patients were assessed repeatedly for vital parameters, n-terminal pro-brain natriuretic peptide levels (NT-proBNP), 6-minute walking tests (6MWT) and Kansas City Cardiomyopathy Questionnaires (KCCQ).

In IV, design and rationale for a future multicentre RCT is presented, with sample size calculations.

Results: Paper I: Eligibility rates for the HF cohort for DAPA-HF and EMPEROR-reduced was 52% and 39%, respectively. The most common reasons for ineligibility were low levels of NT-proBNP and reduced renal function. Patients in the HF cohort were on average 13 years older and generally more comorbid than comparator trial participants but with a lower prevalence of DM.

Paper II: Of the 2326 patients in the HF cohort with a documented LVEF, 681 had HFrEF, 561 HFmrEF and 1084 HFpEF. The proportion of patients in different glycaemic groups was not significantly different between HF phenotypes, with 26-27% having known DM, 3-6% having probable DM, 18-20% having prediabetes and 22-25% being normoglycemic. Compared to normoglycemic patients, patients with glycaemic disorders were more likely to be treated with beta blockers and loop diuretics. The treatment of DM was not different between HF phenotypesv

Paper III: 54 patients were reviewed for eligibility, of which 10 were enrolled and 7 completed the trial. There was a trend towards better outcome parameters during the intervention phase. Compared to baseline, NT-proBNP was 142 (95% CI -324 to +41) ng/l lower during intervention and 6MWT distance 24.3 (95% CI -2.8 to +51.5) meters longer, while KCCQ was stable over the baseline and intervention phase but declined by 2.9 (95% CI -7.3 to +1.5) points during the withdrawal phase). Systolic blood pressure was lower during the intervention phase, -8.8 (95% CI -13.2 to 4.4) mmHg.

IV: Design for a multicentre RCT to assess the effect of dapagliflozin is considered and a placebo-controlled, double-blinded 2x2 crossover trial with KCCQ total symptom score as primary outcome variable is suggested, which would require 80 participants.

Conclusions: There are significant differences between trial populations and the target population for SGLT2i treatment, which challenged the external validity of the trial results in a real-world setting.

The prevalence of T2DM and prediabetes among HF patients is not different between HF phenotypes. The prevalence in Northern Sweden is comparable to what have been found in other settings.

Treatment with SGLT2i dapagliflozin in patients with ATTR-CM may be associated with beneficial effects on NT-proBNP, walking distance and symptom burden, but could also cause a potentially clinically significant drop in systolic blood pressure. These effects need to be studied in a larger, adequately powered trial, where a multicentre crossover trial with 80 participants could be an effective method to do so.