Sodium Glucose Co-Transporter 2 Inhibitors Improve Renal Congestion and Left Ventricular Fibrosis in Rats With Hypertensive Heart Failure

Tomofumi Nakatsukasa; Tomoko Ishizu; Masumi Ouchi; Nobuyuki Murakoshi; Kimi Sato; Masayoshi Yamamoto; Kunio Kawanishi; Yoshihiro Seo; Masaki Ieda

doi:10.1253/circj.CJ-22-0105

Abstract

Background: Elevated central venous pressure (CVP) in heart failure causes renal congestion, which deteriorates prognosis. Sodium glucose co-transporter 2 inhibitor (SGLT2-i) improves kidney function and heart failure prognosis; however, it is unknown whether they affect renal congestion. This study investigated the effect of SGLT2-i on the kidney and left ventricle using model rats with hypertensive heart failure.

Methods and Results: Eight rats were fed a 0.3% low-salt diet (n=7), and 24 rats were fed an 8% high-salt diet, and they were divided into 3 groups of untreated (n=6), SGLT2-i (canagliflozin; n=6), and loop diuretic (furosemide; n=5) groups after 11 weeks of age. At 18 weeks of age, CVP and renal intramedullary pressure (RMP) were monitored directly by catheterization. We performed contrast-enhanced ultrasonography to evaluate intrarenal perfusion. In all high-salt fed groups, systolic blood pressure was elevated (P=0.287). The left ventricular ejection fraction did not differ among high-salt groups. Although CVP decreased in both the furosemide (P=0.032) and the canagliflozin groups (P=0.030), RMP reduction (P=0.003) and preserved renal medulla perfusion were only observed in the canagliflozin group (P=0.001). Histological analysis showed less cast formation in the intrarenal tubule (P=0.032), left ventricle fibrosis (P<0.001), and myocyte thickness (P<0.001) in the canagliflozin group than in the control group.

Conclusions: These results suggest that SGLT2-i causes renal decongestion and prevents left ventricular hypertrophy, fibrosis, and dysfunction

Renal function is a strong prognostic factor of congestive heart failure. Renal dysfunction in patients with heart failure indicates deteriorating long-term prognosis¹ and is associated with continuous central venous pressure (CVP) elevation.²^,³ CVP elevation followed by venous congestion reduces renal medullary perfusion, urine flow, and sodium excretion from renal tubules.⁴^,⁵ Impaired sodium excretion causes additional volume overload to the heart, which results in a vicious cycle of events. Therefore, renal decongestion may be a main therapeutic target in heart failure. Loop diuretics are frequently used for heart failure treatment, but they mainly reduce the intravascular fluid and not the interstitial fluid. To achieve interstitial fluid removal, a significant dose of loop diuretics is required, which results in excessive circulating plasma volume reduction⁶ and underfilling. Sodium glucose co-transporter (SGLT) 2 inhibitor (SGLT2-i) has been reported to improve heart failure prognosis.⁷^–⁹ SGLT2 inhibitors are reported to have a different diuretic effect than loop diuretics and are expected to mainly reduce interstitial congestion.⁶^,¹⁰ However, it is not well known whether SGLT2-i can remove renal decongestion in congestive heart failure. The aim of this study was to investigate the effect of SGLT2-i on renal congestion and myocardial remodeling.

Methods

Animal Procedures

Thirty-two male Dahl-Iwai S rats (DIS/Eis) (Japan SLC Inc. Shizuoka, Japan), which model hypertensive heart failure, were used. Eight rats were fed a 0.3% NaCl low-salt diet (the control group; LS, n=8), and the remaining 24 rats were fed an 8% NaCl high-salt diet after 6 weeks of age. They were divided into 3 groups: untreated (HS, n=8), SGLT2-i (canagliflozin 10 mg/kg/day) (n=8), and loop diuretic (furosemide 20 mg/kg/day) groups (loop-d, n=8) at 11 weeks. We strictly followed all applicable international and institutional guidelines for the care and use of animals. All animals were treated under an experimental protocol approved by the University of Tsukuba’s Institutional Animal Care and Use Committee and in compliance with the Guide for the Care and Use of Laboratory Animals (reference number: 20-146).

Echocardiographic Studies

Echocardiographic studies were performed at 10, 15, and 18 weeks of age. Rats were administered inhalation anesthetic using isoflurane adjusted to a dose to maintain the heart rate at 300–350 beats/min and were confirmed to be completely sedated during the procedure. Cardiac image sequences were acquired with a Vevo2100 (VisualSonic Inc., Toronto, Canada) using a 13- to 24-MHz liner transducer (MS-250). Left ventricular (LV) diastolic diameter (LVDd), systolic diameter, diastolic interventricular septum thickness (IVSTd), diastolic posterior wall thickness (PWTd), and LV global longitudinal strain were obtained from a parasternal long-axis view. LV mass was calculated as 1.053 × [(LVDd + IVSTd + PWTd)³ − LVDd³], and LV ejection fraction was calculated by using the Teichholz method.

Hemodynamic Studies

At 10, 12, 15, and 18 weeks of age, we measured blood pressure using a tail-cuff system (CODA standard system; Hakubatec Life Science Solutions Co., Ltd., Tokyo, Japan). All rats were placed on a heated surgical plate and anesthetized in the same way as is done during echocardiography. A 24-gauge indwelling catheter was cannulated into the femoral vein followed by attachment to a y-connecting catheter. A fiber-optic pressure sensor (FISO-LS-PT9; FISO Technologies, Quebec, Canada) was inserted into the inferior vena cava to measure CVP. Subsequently, to measure RMP, the left kidney was exposed outside, and the fiber-optic pressure sensor was inserted into the renal medulla under echocardiographic guidance.¹¹^,¹²

Contrast-Enhanced Ultrasonography

The contrast-enhanced ultrasonography was performed with an Aplio (Canon Medical Systems Co., Otawara, Japan) using a PLT-1204BT probe to quantify the circulation of renal capillaries in the cortex and medulla, and was conducted in the right kidney, which was not used for RMP measurement. A perflubutane microbubble ultrasound contrast agent (Daiichi-Sankyo Co. Ltd., Tokyo, Japan) was administered through the femoral vein as a bolus of 5 μL/kg followed by an injection of 5 mL saline. Renal microcirculation was analyzed by using a Vitrea Workstation (Canon Medical Systems Co.).¹¹ The region of interest (ROI) was placed in the cortex and medulla, where the contrast enhancement was most homogeneous. A time-intensity curve was generated from an average signal intensity in decibels within ROIs. Further, on a time-intensity curve, we measured time to peak (TTP in seconds), which was defined as the time from initial rise to peak intensity.

Urine and Serum Measurements

For 24-h urine sample collection, rats at 18 weeks were placed in metabolic cages (CT-10 s type II; CLEA Japan, Inc., Tokyo, Japan) with free access to food and water. The collected urine was used for measurement of 24-h urine volume; sodium, potassium, and creatinine levels; and liver-type fatty acid-binding protein (L-FABP). Serum from trunk blood was used for analysis of serum sodium, potassium, urea nitrogen, creatinine, interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α) levels.

Pathology

The right kidney was fixed with 10% formaldehyde in paraffin and sectioned into 4-μm-thick slices. Pathological specimens were stained with Masson’s trichrome to evaluate renal fibrosis and periodic acid-Schiff (PAS) to assess renal tubular cast formation and glomerular lesions. Tissue sections of the kidney and LV were photographed at high resolution by using a NanoZoomer Digital Pathology system (2.0RS; Hamamatsu Photonics, Hamamatsu, Japan). The distribution of the proportion of cast area in the renal tubule that was measured using QuPath software version (0.2.3)¹³ was calculated by dividing the total area of kidney by the cast area to assess the urine flow in the renal tubule. Focal segmental glomerulosclerosis (FSGS), and global sclerosis (GS) were evaluated by dividing the number of glomerular lesions by the total number of glomeruli. The total cortical layer was divided into the outer cortex and the middle juxtamedullary cortex. Glomerular evaluation was also performed for each layer. The LV fibrotic area identified by Masson’s trichrome staining was calculated using Image J software. LV hypertrophy was evaluated by the cross-sectional area and LV myocyte thickness;¹⁴ 50–60 myocytes were randomly selected, and the myocyte area and thickness in the nuclear position were measured.

RNA-Seq Protocol and Data Analysis

Total RNA was extracted from LV using the TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s instructions. The RNA pellets were dissolved in 30 µL of Milli-Q Water, and an Agilent RNA 600 Nano Kit (Cat# 5067-1511; Agilent) on the Bioanalyzer (Agilent) was used to check the RNA solutions of samples (3 specimens/group) for integrity. As 500 ng of the RNAs from each sample was used, libraries were prepared using the NEBNext Ultra II RNA Library Prep Kit for Illumina and the NEBNext rRNA Depletion Kit v2 (Cat# E7770S and E7400L; New England Biolabs, Ipswoch, MA, USA), and 12 polymerase chain reaction cycles were performed. Concentrations and size distributions of the libraries were evaluated with an Agilent DNA 7500 kit (Cat#5067-1506; Agilent) on the Bioanalyzer and all samples were passed for analyses on next generation sequencing equipment. The libraries were pooled for denaturation and neutralization, and the concentrations were adjusted to 1 nM. The libraries were diluted to 1.8 pM and applied for a next-generation sequencing run using NextSeq500/550 v2.5 (75 Cycles) Kits (Cat#20024906; Illumina) in the NextSeq 500 System (Illumina). The sequencing was performed with paired reads of 36 bases. FASTQ files were exported for analysis.

Data Analysis

The results are presented as the mean±standard error, and the number of replicates was at least 5 per group for each data set. The significance among groups was determined by one-way analysis of variance. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test. Correlations between parameters were evaluated by using Pearson’s correlation coefficient. All P values were 2-sided, and P<0.05 was considered statistically significant. All data were analyzed using JMP 11.0 (SAS Institute Inc, Cary, NC, USA).

Results

Two rats in the HS group, 2 in the SGLT2-i group, and 3 in the loop-d group died by 18 weeks. One rat in the LS group died during the experiment.

Body Weight (BW), Organ Weight, Blood Pressure, and Heart Rate

At the age of 18 weeks, the BW of high-salt diet rats was higher than that of low-salt diet rats, whereas the LV weight to BW ratio and the kidney weight to BW ratio in high-salt diet rats were higher than those in the LS group. The lung weight to BW ratio was not significantly different among the 4 groups (Table). Systolic and diastolic blood pressure in all rats on a high-salt diet were constantly higher than those in the LS group after 12 weeks, and the SGLT2-i group did not show a significantly lower blood pressure than that of the HS group (Supplementary Figure 1). Heart rate did not differ between the 4 groups.

Table. Characteristics of Hypertensive Rats at 18 Weeks of Age

	LS (n=7)	HS (n=6)	SGLT2-I (n=6)	Loop-d (n=5)	P value
Weight, g	361.0±5.7	325.2±10.1*	323.1±8.7*	330.3±10.1	0.011
LV/BW, g/kg	2.6±0.1	4.0±0.3*	3.4±0.1*	4.0±0.3*	<0.001
Lung/BW, g/kg	4.4±0.1	5.1±0.4	4.4±0.1	4.7±0.2	0.074
Kidney/BW, g/kg	7.4±0.2	11.6±0.7*	11.7±0.3*	10.9±0.2*	<0.001
Hemodynamics
sBP, mmHg	152.0±3.2	210.5±9.0*	196.7±10.1*	217.3±6.5*	<0.001
dBP, mmHg	103.3±3.1	152.2±8.4*	137.9±10.3*	154.9±7.5*	<0.001
HR, beats/min	336.8±10.2	398.4±26.5	384.0±10.9	396.2±13.8	0.4978
Serum parameters
Sodium, mEq/L	147.0±0.6	139.5±4.2	144.2±0.5	145.4±2.0	0.139
Cre, mg/dL	0.43±0.08	0.51±0.05	0.44±0.05	0.61±0.07	0.062
IL-6, pg/mL	32.4±1.7	30.0±1.5	38.5±5.7	37.6±4.7	0.330
TNF-α, pg/mL	5.71±2.14	6.71±0.87	4.82±1.04	7.68±1.58	0.640
Urine parameters
UV, mL/day	15.9±2.1	76.8±9.2*	104.3±16.2*	59.4±9.9*	<0.001
Alb·Cre, mg/g	498.3±89.1	6,057.2±1,403.2*	3,819.4±537.3*	3,252.6±390.4	<0.001
Sodium, mEq/day	0.4±1.8	20.5±1.9*	22.7±1.9*	12.4±2.1*	<0.001
CCR, mg/min/kg	6.2±0.4	6.8±0.9	8.7±0.9	5.5±0.6	0.0356
L-FABP, ng/mgCr	1.69±0.17	51.2±15.3*	35.7±17.1	28.6±9.8	0.0463
Echocardiography
Dd, mm	7.2±0.1	7.5±0.4	7.8±0.1	8.0±0.3	0.1882
Ds, mm	4.8±0.1	5.6±0.3*	5.6±0.2*	5.8±0.2*	0.0129
IVSTd, mm	1.8±0.1	2.6±0.1*	2.1±0.1**	2.4±0.1*	0.0002
PWTd, mm	2.0±0.1	2.5±0.1*	2.1±0.1**	2.4±0.1*	0.0014
LVEF, %	61.1±1.1	46.2±2.6*	52.8±3.9	52.1±1.1	<0.001
LV mass, mg	1,024.3±47.9	1,609.5±45.1*	1,326.2±42.4^,*	1,658.8±43.6*	<0.001
GLS, %	−22.4±1.3	−16.2±0.7*	−22.3±1.6**	−16.2±0.7*	0.0006

Data are presented as mean±standard error. Alb, albumin; BW, body weight; CCR, creatinine clearance; Cre, creatinine; dBP, diastolic blood pressure; Dd, diastolic diameter; Ds, systolic diameter; GLS, global longitudinal strain; HR, heart rate; HS, high-salt diet; IL-6, interleukin-6; IVSTd, interventricular septum thickness; L-FABP, liver-type fatty acid-binding protein; Loop-d, loop diuretic; LS, low-salt diet; LV, left ventricular; LVEF, left ventricular ejection fraction; PWTd, diastolic posterior wall thickness; sBP, systolic blood pressure; SGLT2-i, sodium glucose co-transporter 2 inhibitor; TNF-α, tumor necrosis factor-α; UV, urine volume. Corrected L-FABP (ng/mgCr) equals L-FABP (ng/mL) divided by urine creatinine (mg/dL)×100. *P<0.05 vs. the LS group, **P<0.05 vs. the HS group. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test.

Serum and Urinary Parameters

Serum sodium value did not differ between the groups. Regarding urine parameters, all rats on a high-salt diet showed higher total urine volume, sodium excretion, and elevated urinary albumin than those in the LS group. The SGLT2-i and loop-d groups did not show significantly increased urine volume compared to the HS group. There were no differences in serum creatinine and creatinine clearance among the 4 groups. Urine L-FABP was significantly elevated in the HS group. In the SGLT2-i and loop-d groups, urinary L-FABP value was also more than 15-fold higher than that in the LS group, although it was not statistically significant. Inflammatory biomarkers showed no difference between the 4 groups.

Echocardiography

Table shows the echocardiographic data of the rats at 18 weeks of age. Only the HS group showed an impaired LV ejection fraction (LVEF) (Figure 1A). LV global longitudinal strain (LV-GLS) was equivalent in the LS and SGLT2-i groups, whereas that in the loop-d and HS groups was clearly exacerbated compared to these 2 groups. LV end-diastolic diameter (LVDd) was not significantly different among the 4 groups. In the SGLT2-i group, LV wall thickness and mass were less than those in the HS and loop-d groups. At 10 weeks, there was no difference in the echocardiographic parameters between the 4 groups (Figure 1B). LVEF in the HS, SGLT2-i, and loop-d groups was lower than that in the LS group, and it decreased further at 18 weeks in the HS group. Although the LV mass in the HS, SGLT2-i, and loop-d groups was significantly greater than that in the LS group at 15 weeks, the SGLT2-i group did not show progressive LV mass increase from 15 to 18 weeks. At 15 weeks, a similar trend was also shown in the LV-GLS. LV wall thickness in the HS and loop-d groups was higher than that in the LS group. There was no difference in LV wall thickness between the LS and SGLT2-i groups throughout the study period. LVDd was not significantly different among the 4 groups.

Figure 1.

Standard echocardiographic parameters. (A) Representative cases of left ventricular (LV) contraction evaluated with B-mode. LS, low-salt diet; HS, high-salt diet; SGLT2-i, sodium glucose co-transporter 2 inhibitor; Loop-d, loop diuretic. (B) Data are presented as mean±standard error in each group including LVEF, LV-GLS, Dd, LV mass, IVSTd, and PWTd. Dd, diastolic diameter; EF, ejection fraction; GLS, global longitudinal strain; IVSTd, diastolic intraventricular septum thickness; PWTd, diastolic posterior wall thickness. *P<0.05, **P<0.01 vs. LS group, ^†P<0.05, ^††P<0.01 vs. HS group. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test.

CVP and Renal Congestion

The relationship between RMP and CVP was significant but modest (R=0.44, P=0.030) (Figure 2A). In the SGLT2-i group, both CVP and RMP were lower than that in the HS group; the loop-d group did not show reduced RMP, although CVP reduced to the same degree as that in the LS group (Figure 2B). In contrast-enhanced ultrasonography, although TTP in the cortex did not show a significant difference among the 4 groups, TTP in the medulla showed differences (Figure 3B). TTP in the medulla was significantly higher in the HS and loop-d groups; the SGLT2-i group showed a similar TTP to that in the LS group. CVP did not relate to TTP in either the cortex or medulla (Figure 3C). The TTP in the medulla correlated with RMP (R=0.62, P=0.001).

Figure 2.

Central volume pressure and renal medullary pressure. (A) Shows the correlation between CVP and RMP. (B) CVP and RMP value bars are the mean±standard error in each group. CVP, central venous pressure; RMP, renal medullary pressure. *P<0.05, **P<0.01, ***P<0.001 vs. the LS group; ^#P<0.05, ^##P<0.01 vs. the HS group. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test.

Figure 3.

Time to peak intensity. (A) Shows representative images of the intensity curve in the medulla. White arrows represent peak intensity, and yellow arrows describe the time from inflection point to the peak of the intensity curve. (B) TTP intensity was quantified in the cortex and medulla. Comparisons of TTP in the cortex and medulla are expressed as the mean±standard error in each group. (C) Correlation between hemodynamics evaluation, including CVP and RMP, and TTP in the cortex and medulla. TTP, time to peak intensity. **P<0.01, ***P<0.001 vs. the LS, group; ^##P<0.01 vs. the HS group. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test (B).

Renal Pathology and Correlation With Renal Medulla Pressure and Perfusion

Renal fibrosis area evaluated by pathological specimens stained with Masson’s trichrome hardly developed in all 4 groups at 18 weeks of age. The percentage of the tubular cast area in the LS group was the smallest among the 4 groups (Figure 4A). In glomerular evaluation, the percentage of GS and FSGS in the SGLT2-i and loop-d groups did not differ from that in the HS group (Figure 4B). FSGS in the SGLT2-i group was decreased in the juxtamedullary cortex compared to that in the HS group, whereas no significant difference was observed in the outer cortex. The SGLT2-i group showed a lower percentage cast area than the HS group, whereas in the loop-d group, there was no difference compared with the HS group. Both RMP and TTP in the renal medulla were significantly correlated with the percentage cast area (Figure 4C).

Figure 4.

The renal tubular and the glomerular lesion. (A) Left panel shows representative images of tubular casts stained by PAS. Insets in each panel are higher magnification images of the boxed area. The right panel shows comparisons of the percentage of tubular cast area that are expressed as the mean±standard error in each group. (B) Shows the results of the percentage of FSGS and GS. (C) Illustrates correlations between the percentage of tubular cast area and the RMP and TTP in the medulla. FSGS, focal segmental glomerulosclerosis; GS, global sclerosis; PAS, periodic acid-Schiff. *P<0.05, **P<0.01 vs. the LS, group; ^#P<0.05, ^##P<0.01 vs. the HS group. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test (A,B).

Myocardial Pathology and Correlation With Renal Medulla Pressure and Perfusion

LV fibrosis in the HS group was significantly worse than that in the LS group. The SGLT2-i group showed suppressed fibrosis compared to the loop-d group (Figure 5A). Moreover, LV-myocyte cross-sectional area and thickness in the HS and loop-d groups progressed, whereas the SGLT2-i group showed ameliorated LV hypertrophy (Figure 5B,C). CVP was only slightly correlated with LV fibrosis (R=0.42, P=0.036) but not with LV-myocyte cross-sectional area (P=0.101) and thickness (P=0.109) (Supplementary Figure 2). In contrast, Figure 6 indicates the significant positive correlation between renal medulla congestion, including TTP in medulla and RMP, and LV fibrosis and hypertrophy.

Figure 5.

LV fibrosis and hypertrophy. (A–C) Left panels show representative images of LV fibrosis (A), LV myocyte cross-sectional area (B), and myocyte thickness (C). Figures on the right show comparisons among the 4 groups, corresponding to those shown in the left panels. ***P<0.001 vs. the LS group; ^###P<0.001 vs. the HS group. Differences between the control or untreated group and the treatment groups were evaluated using Dunnett’s test (A–C).

Figure 6.

Relationship between renal function and myocardial fibrosis and hypertrophy. These panels show correlations between renal congestion parameters, and the percentage of the left ventricular (LV) fibrosis area, cross-sectional area, and myocyte thickness.

RNA Expressions in the Left Ventricle

The heat map of genes that showed an expression difference among the 4 groups showed a similar pattern between the LS and SGLT2-i groups (Figure 7). In the SGLT2-i group, a total of 109 genes were identified as being downregulated, and had a RNA expression fold change of <−1.5 and a FDR P value <0.05 compared to the HS group. Nine of these genes associated with LV fibrosis and dysfunction, including ALAS2,¹⁵ NCOA4,¹⁶ CILP, LTBP2, Comp,¹⁷ GPNMB,¹⁸ and NCAM1,¹⁹ were identified. Further, NPPA (atrial natriuretic peptide) and NPPB (B-type natriuretic peptide) had an inhibited expression in the SGLT2-i group (Supplementary Table).

Figure 7.

Heat map of left ventricular genes in a rat with hypertension. Heat map of left ventricular genes that are differentially expressed among the 4 groups identified from RNA sequencing. HS, high-salt diet; Loop-d, loop diuretic; LS, low-salt diet; SGLT2-i, sodium glucose co-transporter 2 inhibitor.

Discussion

This study revealed that SGLT2-i reduced RMP and preserved renal medullary perfusion in the hypertensive heart-failure model rat. Moreover, renal decongestion caused by SGLT2-i was associated with LV structural and functional preservation. Furosemide reduced CVP but not renal medullary pressure accompanied by LV fibrosis and hypertrophy. To the best of our knowledge, this study is the first to show the favorable relationship between renal decongestion caused by SGLT2-i and LV remodeling.

Renal Decongestion Caused by Diuretics

Canagliflozin removed fluids from both intravascular and renal medullary interstitial compartments, whereas furosemide reduced only CVP and not renal congestion. Furosemide has strong diuretic effects by inhibiting Na⁺ reabsorption in the ascending Henle’s loop, whereas SGLT2-i decreases reabsorption of glucose and sodium in the proximal tubular cells, which leads to osmotic diuretic and natriuretic effects. This type of diuresis effect may be expected to reduce water and Na⁺ reabsorption and ameliorate extracellular fluid expansion.⁶^,¹⁰^,²⁰^,²¹ This may decrease the osmotic gradient between the renal tubular and interstitial fluids. Canagliflozin may reduce renal medullary interstitial fluid and release the compression of the renal tubules and vasa recta, thus resulting in an improvement of microcirculation in the renal medulla. In addition, SGLT2-i significantly decreased FSGS in the juxtamedullary cortex. A previous study indicated that glomeruli in the juxtamedullary cortex were mainly injured due to a disorder in renal medullary interstitial perfusion, and that the renin-angiotensin-aldosterone system (RAAS) is associated with glomerular lesions in the cortex.²² This result also suggests that canagliflozin improved renal medullary perfusion. However, as there was no difference in the overall number of glomerular lesions among rats fed with a high-salt diet, we considered that the administration of canagliflozin did not reduce urinary albumin. Moreover, canagliflozin significantly suppressed renal intra-tubular cast formation, whereas furosemide did not. Interstitial injury leads to renal tubule dysfunction and, as a result, intratubular urinary flow decreases. We have used cast formation as an evaluation of renal decongestion because amelioration of intratubular urinary flow is one of the factors that leads to a reduction of cast formation. These results suggest that it is possible to improve urine flow by renal decongestion without affecting blood pressure, thereby contributing to renal protection. Loop diuretics increase Na⁺ excerption in the ascending limb of the loop of Henle; however, chronic administration of loop diuretics induces an increase of sodium reabsorption in distal nephrons, and loop diuretics may gradually be unable to produce sufficient osmotic gradient.²³ Chronic heart failure causes venous congestion followed by fluid accumulation in the interstitial space of the body. In addition, vasa recta that has poor autoregulation of renal medullary circulation is affected easily by interstitial pressure, which may further promote renal congestion.²⁴ Reduced renal medullary perfusion causes increased sodium reabsorption,⁶ which may further promote fluid volume retention including that in the interstitial compartment. This study showed that urinary Na⁺ was not significantly increased by furosemide administration. Low urinary Na⁺ in patients with heart failure is known as a predictive factor of loop diuretic resistance, which is supportive of our results.²⁵ It has been reported that the severity of renal congestion is associated with prognosis in patients with heart failure;²⁶ however, most patients who were discharged did not show sufficient renal decongestion;²⁷ therefore, some of these patients could have had persisting renal congestion, even if heart failure was well-controlled after treatment. In this study, urinary sodium excretion was found to be lower, and urinary volume tended to decrease in the loop-d group compared to that in the other high-salt fed groups. However, it has been reported that Dahl rats had an increase in urine volume when treated with loop diuretics.²⁸ Intravascular dehydration could occur when renal congestion is alleviated with loop diuretics, which causes mainly intravascular fluid removal,⁶^,¹⁰ leading to further deterioration in renal function. In contrast, although the urine volume in the SGLT2-i group has not increased significantly compared with that of the HS group, CVP decreased and excessive fluid was sufficiently removed, thus relieving renal congestion. In patients with chronic heart failure, the reno-protective effect of SGLT2-i may be obtained from renal medulla decongestion.

Renal Decongestion Effects on the LV

Canagliflozin and furosemide reduced venous pressure. Although cardiac preload was mitigated by both treatments, canagliflozin significantly suppressed the progression of LV fibrosis and hypertrophy, unlike the loop diuretic. Many clinical trials have also shown improvement in LV hypertrophy by SGLT2-i, with similar results to our study,²⁹^,³⁰ and in the clinical trial, the use of SGLT2-i reduced the risk of cardiovascular death or heart failure causing hospitalization in patients with both heart failure with preserved and reduced LVEF.⁷^–⁹ The present study showed that there is a positive correlation between renal medullary decongestion and LV fibrosis and hypertrophy improvement. Moreover, canagliflozin did not significantly decrease blood pressure; therefore, the cardioprotective effects of canagliflozin could not be explained by the effect on blood pressure reduction. Worsening renal congestion was correlated with progression of LV hypertrophy, which suggests that SLGT2-i could exert a cardioprotective effect by way of renal decongestion. Several cardiovascular benefits of SGLT2-i are suggested, including an increase in ketone body production and utilization, improvement of cardiac energy metabolism,³¹^,³² and inhabitation of oxidative stress and inflammation. However, in this study, mRNA expression associated with inflammation as well as serum IL-6, TNF-α, and ketone oxidation, including Bdh1 and Oxct1,³³ were not significantly suppressed. Moreover, although canagliflozin can bind to SGLT1 of the heart, it was suggested that canagliflozin does not influence SGLT1 of the heart in consideration of the maximal concentration of plasma-unbound canagliflozin.³⁴ Therefore, these may suggest that renal decongestion is mainly related to a cardio-protective effect. Furosemide activates the RAAS, which was not assessed in this study; however, it was reported that SGLT2-i could suppress renal RAAS in animal studies, and the effects of cardiovascular benefit through inhabitation of RAAS remains unclear.³⁵ In addition, this study did not intend to conclude whether cardiovascular benefits are obtained from renal decongestion or directly by SGLT2-i.

Study Limitations

This study had several limitations. First, because there was no significant difference in renal function among the four groups, we could not assess whether SGLT2-i has an impact on renal function. In the loop-d group, in which natriuresis would be expected, the urine volume tended to decrease compared to the other groups, and tubular disorders, which cannot be evaluated by creatinine clearance alone, could have occurred. Moreover, in the loop-d group, renal medullary perfusion was low and a reduction in renal medullary perfusion could have easily increased the reabsorption of sodium from the renal tubules.²⁴ As a result, urinary volume and sodium excretion at the age of 18 weeks may have been lower than those in the HS group. In this aspect, SGLT2-i may have a long-term renoprotective effect. Second, we used Dahl-sensitive rats, which caused them hypertension followed by nephrosclerosis when they were fed with a high-salt diet. Therefore, causes of renal congestion could not only be heart failure, but also nephrosclerosis caused by hypertension. However, SGLT2-i did not decrease blood pressure sufficiently to be statistically significant compared to that in the control group, and it is thought that SGLT2-i may ameliorate renal congestion caused by heart failure. Third, in the SGLT2-i group, some genes associated with LV remodeling were suppressed compared to the control and loop-d groups. However, when the hearts were dissected, left ventricular remodeling had already progressed in all rats. It could be said that this change of RNA expression was not caused by renal decongestion caused by SGLT2-i, but that it occurred as a result of the progression of heart failure. Fourth, the rats in the SGLT2-i group were all given the same dose. Therefore, it is unclear whether the effect of SGLT2-i is dose-dependent. The mechanisms by which renal decongestion leads to the beneficial effects on the left ventricle could not be investigated in this study. Further investigations, including molecular biology experiments, are needed to clarify whether renal decongestion by SGLT2-i contributes to cardiorenal protective effects in patients with renal dysfunction.

Conclusions

Canagliflozin induced renal decongestion in a rat model of hypertensive heart failure and inhibited the progression of LV fibrosis and hypertrophy without deteriorating blood pressure. SGLT2-i may have a promising effect on renal decongestion, which is the major therapeutic target of hypertensive heart disease in preventing LV structural and functional remodeling.

Acknowledgment

We thank Yuko Tashiro, Yumi Isaka, and Takehito Sugasawa for their technical assistance. We would also like to thank Editage (www.editage.com) for English-language editing.

Sources of Funding

This work was partly supported by JSPS KAKENHI (Grant Number JP 19K17730 [to K.K.]), and partly supported by grants from the Canon Medical Systems Co, Daiichi-Sankyo Co, Otsuka Pharmaceutical Co., and Teijin Pharma (to T.I.). All sponsors had no role in the study design; data collection and analysis; result interpretation; or in preparation, review, and approval of the article.

Disclosures

Mitsubishi Tanabe Pharma Corporation, Japan, donated canagliflozin for this study.

IRB Information

All animals were treated under an experimental protocol approved by the University of Tsukuba’s Institutional Animal Care and Use Committee and in compliance with the Guide for the Care and Use of Laboratory Animals (reference number: 20-146).

Data Availability

The deidentified participant data will not be shared.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-22-0105

References

1. Smilde TDJ, Hillege HL, Navis G, Boomsma F, de Zeeuw D, van Veldhuisen DJ. Impaired renal function in patients with ischemic and nonischemic chronic heart failure: Association with neurohormonal activation and survival. Am Heart J 2004; 148: 165–172.
2. Damman K, van Deursen VM, Navis G, Voors AA, van Veldhuisen DJ, Hillege HL. Increased central venous pressure is associated with impaired renal function and mortality in a broad spectrum of patients with cardiovascular disease. J Am Coll Cardiol 2009; 53: 582–588.
3. Mullens W, Abrahams Z, Francis GS, Sokos G, Taylor DO, Starling RC, et al. Importance of venous congestion for worsening of renal function in advanced decompensated heart failure. J Am Coll Cardiol 2009; 53: 589–596, doi:10.1016/j.jacc.2008.05.068.
4. Shimada S, Hirose T, Takahashi C, Sato E, Kinugasa S, Ohsaki Y, et al. Pathophysiological and molecular mechanisms involved in renal congestion in a novel rat model. Sci Rep 2018; 8: 16808, doi:10.1038/s41598-018-35162-4.
5. Firth JD, Raine AE, Ledingham JG. Raised venous pressure: A direct cause of renal sodium retention in oedema? Lancet 1988; 1: 1033–1035.
6. Hallow KM, Helmlinger G, Greasley PJ, McMurray JJV, Boulton DW. Why do SGLT2 inhibitors reduce heart failure hospitalization? A differential volume regulation hypothesis. Diabetes Obes Metab 2018; 20: 479–487.
7. Anker SD, Butler J, Filippatos G, Ferreira JP, Bocchi E, Böhm M, et al. Empagliflozin in heart failure with a preserved ejection fraction. N Engl J Med 2021; 385: 1451–1461.
8. Packer M, Anker SD, Butler J, Filippatos G, Pocock SJ, Carson P, et al. Cardiovascular and renal outcomes with empagliflozin in heart failure. N Engl J Med 2020; 383: 1413–1424.
9. Bhatt DL, Szarek M, Steg PG, Cannon CP, Leiter LA, McGuire DK, et al. Sotagliflozin in patients with diabetes and recent worsening heart failure. N Engl J Med 2021; 384: 117–128.
10. Ohara K, Masuda T, Murakami T, Imai T, Yoshizawa H, Nakagawa S, et al. Effects of the sodium-glucose cotransporter 2 inhibitor dapagliflozin on fluid distribution: A comparison study with furosemide and tolvaptan. Nephrology 2019; 24: 904–911.
11. Komuro K, Seo Y, Yamamoto M, Sai S, Ishizu T, Shimazu K, et al. Assessment of renal perfusion impairment in a rat model of acute renal congestion using contrast-enhanced ultrasonography. Heart Vessels 2018; 33: 434–440.
12. Chiba H, Seo Y, Sai S, Namekawa M, Ishizu T, Aonuma K. Renoprotective effects of tolvaptan in hypertensive heart failure rats depend on renal decongestion. Hypertens Res 2019; 42: 319–328.
13. Bankhead P, Loughrey MB, Fernández JA, Dombrowski Y, McArt DG, Dunne PD, et al. QuPath: Open source software for digital pathology image analysis. Sci Rep 2017; 7: 16878.
14. Civitarese RA, Talior-Volodarsky I, Desjardins JF, Kabir G, Switzer J, Mitchell M, et al. The α11 integrin mediates fibroblast-extracellular matrix-cardiomyocyte interactions in health and disease. Am J Physiol Heart Circ Physiol 2016; 311: H96–H106.
15. Sawicki KT, Shang M, Wu R, Chang HC, Khechaduri A, Sato T, et al. Increased heme levels in the heart lead to exacerbated ischemic injury. J Am Heart Assoc 2015; 4: e002272.
16. Tang M, Huang Z, Luo X, Liu M, Wang L, Qi Z, et al. Ferritinophagy activation and sideroflexin1-dependent mitochondria iron overload is involved in apelin-13-induced cardiomyocytes hypertrophy. Free Radic Biol Med 2019; 134: 445–457.
17. Park S, Ranjbarvaziri S, Zhao P, Ardehali R. Cardiac fibrosis is associated with decreased circulating levels of full-length CILP in heart failure. JACC Basic Transl Sci 2020; 5: 432–443.
18. Järve A, Mühlstedt S, Qadri F, Nickl B, Schulz H, Hübner N, et al. Adverse left ventricular remodeling by glycoprotein nonmetastatic melanoma protein B in myocardial infarction. FASEB J 2017; 31: 556–568.
19. Ackermann MA, Petrosino JM, Manring HR, Wright P, Shettigar V, Kilic A, et al. TGF-β1 affects cell-cell adhesion in the heart in an NCAM1-dependent mechanism. J Mol Cell Cardiol 2017; 112: 49–57.
20. Shimizu K, Kurosawa T, Sanjo T, Hoshino M, Nonaka T. Solute-free versus electrolyte-free water clearance in the analysis of osmoregulation. Nephron 2002; 91: 51–57.
21. Lindner G, Schwarz C, Funk GC. Osmotic diuresis due to urea as the cause of hypernatraemia in critically ill patients. Nephrol Dial Transplant 2012; 27: 962–967.
22. Mori T, Cowley AW Jr. Role of pressure in angiotensin II-induced renal injury: Chronic servo-control of renal perfusion pressure in rats. Hypertension 2004; 43: 752–759.
23. Abdallah JG, Schrier RW, Edelstein C, Jennings SD, Wyse B, Ellison DH. Loop diuretic infusion increases thiazide-sensitive Na(+)/Cl(−)-cotransporter abundance: Role of aldosterone. J Am Soc Nephrol 2001; 12: 1335–1341.
24. Cowley AW, Roman RJ, Fenoy FJ, Mattson DL. Effect of renal medullary circulation on arterial pressure. J Hypertens Suppl 1992; 10: S187–S193.
25. Biegus J, Zymliński R, Sokolski M, Todd J, Cotter G, Metra M, et al. Serial assessment of spot urine sodium predicts effectiveness of decongestion and outcome in patients with acute heart failure. Eur J Heart Fail 2019; 21: 624–633.
26. Iida N, Seo Y, Sai S, Machino-Ohtsuka T, Yamamoto M, Ishizu T, et al. Clinical implications of intrarenal hemodynamic evaluation by Doppler ultrasonography in heart failure. JACC Heart Fail 2016; 4: 674–682.
27. Komuro K, Shimazu K, Koizumi T, Imagawa S, Anzai T, Yonezawa K. Demonstration of improved renal congestion after heart failure treatment on renal perfusion imaging with contrast-enhanced ultrasonography. Circ Rep 2019; 1: 593–600.
28. Mouginot D, Laforest S, Drolet G. Challenged sodium balance and expression of angiotensin type 1A receptor mRNA in the hypothalamus of Wister and Dahl rat strains. Regul Pept 2007; 142: 44–51.
29. Verma S, Mazer CD, Yan AT, Mason T, Garg V, Teoh H, et al. Effect of empagliflozin on left ventricular mass in patients with type 2 diabetes mellitus and coronary artery disease: The EMPA-HEART cardiolink-6 randomized clinical trial. Circulation 2019; 140: 1693–1702.
30. Lee MMY, Brooksbank KJM, Wetherall K, Mangion K, Roditi G, Campbell RT, et al. Effect of empagliflozin on left ventricular volumes in patients with type 2 diabetes, or prediabetes, and heart failure with reduced ejection fraction (SUGAR-DM-HF). Circulation 2021; 143: 516–525.
31. Ferrannini E, Mark M, Mayoux E. CV protection in the EMPA-REG OUTCOME trial: A “Thrifty Substrate” hypothesis. Diabetes Care 2016; 39: 1108–1114.
32. Santos-Gallego CG, Requena-Ibanez JA, San Antonio R, Ishikawa K, Watanabe S, Picatoste B, et al. Empagliflozin ameliorates adverse left ventricular remodeling in nondiabetic heart failure by enhancing myocardial energetics. J Am Coll Cardiol 2019; 73: 1931–1944.
33. Kimura T, Nakamura K, Miyoshi T, Yoshida M, Akazawa K, Saito Y, et al. Inhibitory effects of tofogliflozin on cardiac hypertrophy in Dahl salt-sensitive and salt-resistant rats fed a high-fat diet. Int Heart J 2019; 60: 728–735.
34. Ohgaki R, Wei L, Yamada K, Hara T, Kuriyama C, Okuda S, et al. Interaction of the Sodium/Glucose Cotransporter (SGLT) 2 inhibitor Canagliflozin with SGLT1 and SGLT2: Inhibition kinetics, sidedness of action, and transporter-associated incorporation accounting for its pharmacodynamic and pharmacokinetic features. J Pharmacol Exp Ther 2016; 358: 94–102.
35. Satou R, Cypress MW, Woods TC, Katsurada A, Dugas CM, Fonseca VA, et al. Blockade of sodium-glucose cotransporter 2 suppresses high glucose-induced angiotensinogen augmentation in renal proximal tubular cells. Am J Physiol Renal Physiol 2020; 318: F67–F75.

責任著者(Corresponding author)

J-STAGEへの登録はこちら（無料）