Dapagliflozin, peptide YY, and weight loss in heart failure with preserved ejection fraction

Abstract

Although sodium–glucose cotransporter 2 inhibitors (SGLT2i) cause glycosuria and natriuresis, the diuretic effect upon initiation is mild, transient, and cannot explain the magnitude of long-term weight loss observed.1 In patients with heart failure with preserved ejection fraction (HFpEF), there is greater weight loss with dapagliflozin in obese compared to lean individuals,2 and most of the weight reduced is fat, rather than water or lean tissue.3 These observations remain poorly understood and unattributable to urine calorie loss alone. The CAMEO-DAPA trial showed that treatment with dapagliflozin for 24 weeks improved rest and exercise haemodynamics in patients with HFpEF compared to placebo.4 To explore potential mechanisms underlying these benefits, we performed a proteomic analysis (Olink Explore 3072 platform) from paired arterial samples in the fasted state among patients in CAMEO-DAPA.4 The Olink Explore 3072 platform measures levels of unique proteins by proximity extension immunoassay, including multiple proteins not evaluable through earlier platforms.5 Normalized protein expression values are generated and expressed in arbitrary units, with higher values corresponding to higher relative protein concentration on a log2 scale. Body composition before and after treatment was assessed by dual X-ray absorptiometry.3 The change in plasma levels of measured proteins in log2 relative units before and after treatment was obtained by subtracting baseline values from 24-week levels, with changes compared between the dapagliflozin and placebo groups using an unpaired t-test. To correct for potential false discovery, we applied the Benjamini–Hochberg adjustment to the P-values to generate adjusted P-values. Next, we applied the .05 threshold against the adjusted P-values to determine statistical significance. Linear regression and associated Pearson r values were used to assess correlation between variables of interest. A total of 34 participants with HFpEF had paired arterial blood samples available for proteomic analysis (14 placebo, 20 dapagliflozin) [mean age 67.1 ± 8.6 years, 62% female, mean body mass index 35.2 ± 6.8 kg/m2], with 2874 proteins measurable. One hundred ninety proteins demonstrated nominally significant changes at an uncorrected P < .05 (Figure 1A), but after correction for false discovery, there was a significant change in only one protein: peptide YY (PYY), a hormone released by cells in the distal intestine in response to feeding that promotes satiety.6 As compared to placebo, dapagliflozin increased PYY at 24 weeks (placebo-corrected mean log2 fold change +1.74 [95% confidence interval (CI) 1.06–2.43], P < .0001, false discovery corrected P = .034) (Figure 1B). To confirm the proteomic findings, we then measured PYY by ELISA (Merck KGaA, Darmstadt, Germany) in independent paired arterial samples in a blinded fashion. The median PYY level at baseline was 102.4 pg/mL [interquartile range (IQR) 69.8–129.2] with good correlation between ELISA and Olink PYY measurement (r = .73, P < .0001). After 24 weeks, dapagliflozin increased PYY by +31.9 pg/mL [IQR −0.4–52.5] compared to a change of −6.2 pg/mL [IQR −15.8 to +21.5] with placebo (P = .02). Dapagliflozin and change in peptide YY (PYY) in heart failure with preserved ejection fraction: dapagliflozin treatment increased PYY compared to placebo over 24 weeks (A, B) even after correction for false discovery [false discovery rate (FDR) corrected P < .05]. The increase in PYY was correlated with weight loss (C), particularly decreases in trunk fat (D), and was also associated with reduction in exercise pulmonary artery wedge pressure (PAWP) (E) and improvement in pulmonary artery compliance (PAC) (F) Dapagliflozin resulted in a placebo-corrected weight loss of 3.4 kg [95% CI 0.8–5.9 kg] (P = .01) in the 34 patients with paired proteomic analyses. Increases in PYY on treatment were correlated with greater weight loss (r = −.50, P = .003), including greater reduction in trunk fat (r = −.63, P < .0001) and android fat (r = −.47, P = .005) (Figure 1C and D). Increases in PYY were also modestly correlated with reduction in pulmonary artery wedge pressure (PAWP) at 20 W exercise (Figure 1E), with similar directional correlations for treatment effects at rest and peak exercise (rest PAWP r = −.30, P = .08, peak PAWP r = −.32, P = .07). We have also found that dapagliflozin enhances pulmonary vascular reserve with exercise in HFpEF,7 and notably, increases in PYY were correlated with greater improvements in pulmonary artery (PA) compliance (r = +.58, P = .0009) and decreases in PA elastance (r = −.49, P = .007) during 20 W exercise (Figure 1F). These correlations were directionally similar with peak exercise (peak PA compliance r = +.32, P = .06, peak PA elastance r = −.31, P = .09). The gut hormone PYY acts on the hypothalamus to decrease appetite and food intake as blood levels rise acutely after a meal.6,8,9 The significant increase in PYY we observed with dapagliflozin compared to placebo suggests that part of the reduction in body weight and fat observed on treatment3 may be mediated by decreased energy intake rather than energy loss through glycosuria, a novel, heretofore undescribed mechanism of SGLT2i in HFpEF. With chronic weight loss, there are counterregulatory responses in gut hormones that promote rebound weight gain, including decreases in PYY, which increases appetite and caloric intake.8 Thus, the weight reduction on treatment in CAMEO-DAPA might have been expected to decrease PYY as a counteradaptation, yet levels remained ∼30% higher after 24 weeks. Importantly, increases in PYY were also associated with favourable reductions in body weight and upper body fat, with signals for favourable haemodynamic correlations and improved pulmonary vascular loading during exercise. Intriguingly, blockade of the SGLT1 receptor in the intestinal tract has been shown to increase PYY levels.10 Dapagliflozin has a negligible effect on SGLT1, so the observed increase in PYY cannot be explained by altered intestinal glucose delivery, but instead suggests possible systemic effects inducing a state of starvation mimicry.5 The present data connecting fat reduction and haemodynamic benefits with PYY also support future efforts at developing PYY agonists as novel treatments for patients with HFpEF, particularly those with obesity. A larger proteomic analysis of the EMPEROR programme using an older version Olink panel demonstrated changes in markers of autophagy and nutrient deprivation signalling with empagliflozin.5 Our study was underpowered to detect these associations, so the lack of other associated effects of dapagliflozin on these pathways in the present analysis should not be interpreted as evidence against the nutrient deprivation-autophagy hypothesis. Importantly, PYY was not part of the tested proteins in the older Olink Explore 1536 panel used in the EMPEROR analysis, which presumably accounts for the newly discovered finding reported here. In summary, we show that treatment with dapagliflozin for 24 weeks increased plasma PYY levels, and the extent of this increase was correlated with weight loss, particularly adipose tissue loss, and favourable haemodynamic effects during exercise. These data raise the possibility that SGLT2i-mediated increases in PYY may contribute to weight loss in obese HFpEF through increased satiety, with secondary haemodynamic benefits. None. B.A.B. receives research support from the National Institutes of Health (NIH) and the United States Department of Defense, as well as research grant funding from AstraZeneca, Axon, GlaxoSmithKline, Medtronic, Mesoblast, Novo Nordisk, and Tenax Therapeutics. B.A.B. has served as a consultant for Actelion, Amgen, Aria, Axon Therapies, BD, Boehringer Ingelheim, Cytokinetics, Edwards Lifesciences, Eli Lilly, Imbria, Janssen, Merck, Novo Nordisk, NGM, NXT, and VADovations, and is named inventor (US patent no. 10,307,1790307179) for the tools and approach for a minimally invasive pericardial modification procedure to treat heart failure. V.M. receives support from Ministry of Health of the Czech Republic, AZV grant NU22-02-00161, and Y.N.V.R. is supported by NIH grant K23HL164901-01, research grants from Sleep Number, Bayer Accelerated Pulmonary Hypertension Award, United Jenesis Award, and the Earl Wood Career development award from Mayo Clinic. M.H. received lecture fees from Sanofi, Novo Nordisk, Eli Lilly, Novartis, Merck and Berlin Chemie, Astra Zeneca, and Boehringer Ingelheim, participated in clinical trials with Sanofi, Eli Lilly, and Novo Nordisk, and received research support from Sanofi. The data necessary for generating the primary results will be provided upon reasonable request to the corresponding author. This investigator-initiated trial was conducted with support from AstraZeneca. B.A.B. is also supported in part by research grants from the National Heart, Lung and Blood Institute (R01 HL128526 and U01 HL160226) and the United States Department of Defense (W81XWH2210245). V.M., M.H., and B.A.B. are also supported by Czech National Institute for Research of Metabolic and Cardiovascular Diseases (Programme EXCELES, ID project no. LX22NPO5104)—funded by the European Union—Next Generation EU. Y.N.V.R. is supported by National Heart, Lung, And Blood Institute award number K23HL164901, grants from Bayer, United pharmaceuticals, Merck, and the Earl Wood Career Development Award from Mayo Clinic. The study was approved the Mayo Clinic Institutional Review Board, and all participants provided written informed consent. URL: https://www.clinicaltrials.gov; unique identifier: NCT04730947.

Topics & Concepts