Semaglutide in type 2 diabetes with chronic kidney disease at high risk progression-real-world clinical practice.

Background: Semaglutide [glucagon-like peptide-1 receptor-agonist (GLP-1RA)] has shown nephroprotective effects in previous cardiovascular studies. However, its efficacy and safety in patients with chronic kidney disease (CKD) and type 2 diabetes (T2D) have been rarely studied.
Methods: This is a multicenter, retrospective, observational study in patients with T2D and CKD with glycosylated hemoglobin A1c (HbA1c) of 7.5-9.5% treated with subcutaneous semaglutide for 12 months in real-world clinical practice. The main objectives were glycemic control as HbA1c <7% and weight loss >5%.
Results: We studied a total of 122 patients, ages 65.50 ± 11 years, 62% men, duration of T2D 12 years, baseline HbA1c 7.57% ± 1.36% and an estimated glomerular filtration rate (eGFR) 50.32 ± 19.21 mL/min/1.73 m2; 54% had a urinary albumin:creatinine ratio (UACR) of 30-300 mg/g and 20% had a UACR >300 mg/g. After 12 months of follow-up, HbA1c declined -0.73% ± 1.09% (P < .001), with 57% of patients achieving values <7% and weight loss of -6.95 kg (P < .001), with 59% of patients showing a reduction of >5% of their body weight. Systolic and diastolic blood pressure decreased -9.85 mmHg and -5.92 mmHg, respectively (P < .001). The mean UACR decreased 51% in the group with baseline macroalbuminuria (UACR >300 mg/g). The mean eGFR (by the Chronic Kidney Disease Epidemiology Collaboration) remained stable. The need for basal insulin decreased 20% (P < .005). Only 7% of patients on insulin had mild hypoglycemic episodes. Semaglutide was stopped in 5.7% of patients for digestive intolerance. Conclusions: In this real-world study, patients with T2D and CKD treated with subcutaneous semaglutide for 12 months significantly improved glycemic control and decreased weight. Albuminuria decreased by >50% in patients with macroalbuminuria. The administration of GLP-1RA in patients with T2D and CKD was safe and well tolerated.

INTRODUCTION

It is known that ∼23–43% of patients with T2D develop chronic kidney disease (CKD) and currently diabetes is the first cause of end-stage kidney disease (ESKD) that require renal replacement therapy [1]. In addition, albuminuria appears early in the diagnosis of T2D in 17–26% of patients [2]. Patients with T2D and CKD who develop macroalbuminuria have a greater risk of cardiovascular disease mortality than progression to ESKD [3]. The Steno group, among others, have shown that intensified multifactorial intervention in type 2 diabetics with microalbuminuria, including a healthy lifestyle and early control of metabolic hyperglycemia, blood pressure (BP) and weight, slowed progression in nephropathy and renal function loss, reducing the risk of ESKD [4]. Furthermore, in the last decade it has been demonstrated that the use of drugs from classes such as renin–angiotensin–aldosterone system (RAAS) inhibitors, glucagon-like peptide-1 receptor agonists (GLP-1RAs) and sodium-glucose cotransporter-2 inhibitors (SGLT2is) can delay or minimize micro- and macrovascular damage [5, 6].

Mann et al. [7] performed a post hoc analysis of the SUSTAIN 1–7 clinical trials and analyzed the safety and effects on the renal function of subcutaneous (SC) once-weekly semaglutide. They demonstrated that SC semaglutide exerted a protective renal effect as compared with placebo. Likewise, there was a greater reduction in the urinary albumin:creatinine ratio (UACR) in those who presented with microalbuminuria (UACR 30–300 mg/g) and was strongest in patients with macroalbuminuria (UACR >300 mg/g) versus those with no presence of albuminuria [7, 8]. To date, the positive effects in terms of renoprotection exerted by semaglutide have been mainly related to albuminuria reduction.

To our knowledge, no real-world studies have been published on the use of semaglutide in patients with CKD. In the present study, our main objective was to evaluate the glycemic and renal efficacy and safety of semaglutide in patients with T2D and CKD over 12 months of follow-up.

MATERIALS AND METHODS

This is a multicenter, retrospective, observational clinical study that was conducted at the departments of nephrology, internal medicine and endocrinology of three Spanish hospitals: the Costa del Sol Hospital in Marbella, Regional University Hospital of Málaga and the Puerto Real Hospital in Cádiz. The inclusion criteria were patients followed in clinics who are >18 years of age with T2D and an estimated glomerular filtratin rate (eGFR) using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula of >15 mL/min/1.73 m2. Patients ≤18 years of age, with both eGFR (by CKD-EPI) >60 mL/min/1.73 m2 and albuminuria <30 mg/g, kidney transplant, CKD stage 5 or patients taking part in clinical trials were excluded from the study. The inclusion period was from May 2019 to January 2020 and patients were enrolled after signing the informed consent. We collected data at baseline and after 6 and 12 months ± 4 weeks from the beginning of treatment with SC semaglutide. A total of 122 patients with T2D and CKD, defined as an eGFR <60 mL/min/1.73 m2 and/or UACR >30 mg/g were included in the study. All patients had previously received treatment with at least two oral and/or SC glucose-lowering agents. During follow-up visits, patients received guidance regarding a healthy lifestyle, such as eating a 1500-calorie diet and performing 120 min/week of aerobic exercise. In addition, antihypertensive and/or dyslipidemia treatment were intensified during the follow-up period when necessary. Patients had three or four follow-up appointments in the outpatient clinic. In the initial visit, semaglutide was started at a dose of 0.25 mg SC/week for 4 weeks. Then the dose was increased to 0.5 mg SC/week for 4 weeks and then 1 mg SC/week. The continuation of the treatment depended on the investigator’s clinical judgement. Consecutive patients were included in the study after signing an informed consent form. The Ethical Committee of the Hospital Costa del Sol approved the study protocol [Ref FRR/RB (Cod 003_dic_PI-T2DCVRS/2019)].

The following data were obtained: anthropometric parameters [weight, body mass index (BMI), BP], cardiovascular risk factors, risk factors for micro and macrovascular complications and analytical parameters [fasting blood glucose, glycosylated hemoglobin A1c (HbA1c), lipid profile, eGFR (by the CKD-EPI equation); and UACR]. UACR was classified as microalbuminuria when values were 30–300 mg/g and macroalbuminuria when values were >300 mg/g.

Hypoglycemic episodes were defined according to the American Diabetes Association criteria [9]. Cardiovascular events and other possible adverse drug effects were also collected. The main outcome variable was the percentage of patients who achieved good glucose control, defined as an HbA1c <7% and weight loss >5% of the total body weight, by the end of the study period. Secondary outcome variables included any changes in fasting blood glucose levels, BMI, BP, eGFR, UACR, hypoglycemic episodes, treatment withdrawal or a reduction in insulin units at the end of the study.

Statistical analysis

A descriptive analysis was performed using measurements of central tendency, dispersion and position for quantitative variables and distribution of frequency for qualitative variables. A paired sample t-test was used for quantitative variables and McNemar's test for qualitative variables was used for studying differences between pairs. A generalized linear model was used to identify differences in quantitative variables in the baseline values versus values at 12 months. Statistical significance was established as P < .05. All statistical analyses were performed using SPSS version 19.0 software (IBM, Armonk, NY, USA).

RESULTS

The clinical and epidemiological characteristics of the 122 patients included in our study are shown in Table 1. The mean age was 65.56 ± 11 years, 62% were males and the mean duration of T2D was 12.3 ± 3.4 years. The mean BMI was 35.8 ± 4.79 kg/m2. Previous antidiabetic treatment included 50% receiving metformin, 42% SGLT2i, 66.4% on basal insulin and 27.8% on rapid-acting insulin. A total of 64.8% of all patients switched from another GLP-1RA (liraglutide, dulaglutide, exenatide LAR) to semaglutide in the framework of this study. The majority (95.9%) of them were under renin–angiotensin system blockade.

Table 1.

Clinical and epidemiological baseline data

Characteristics	Values
Age (years), mean ± SD	65.50 ± 11
Male, n (%)	75 (62)
eGFR (CKD-EPI; mL/min/1.73 m2), n (%)>6060–4545–3030–15	27 (22)42 (34.4)42 (34.4)11 (9)
UACR (mg/g), n (%)>30030–300<30	24 (20)66 (54)32 (26)
Basal insulin, n (%)	81 (66.4)
Rapid acting insulin, n (%)	34 (27.8)
Metformin, n (%)	61 (50)
DPP4i, n (%)	19 (15.6)
GLP-1RA switched, n (%)LiraglutideDulaglutideExenatide LAR	79 (64.8)48 (39.3)28 (23)3 (2.5)

DPP4i: dipeptidyl peptidase-4 inhibitor.

Data on comorbidities and previous cardiovascular events are presented in Figure 1. Seventy-four percent presented micro- or macroalbumiuria. The mean UACR was 349.5 ± 863.2 mg/g {median 78.5 mg/g [interquartile range (IQR) 26–199.5] [20%, >300 mg/g; 54%, 30–300 mg/g and 26%, <30 mg/g]}. The mean eGFR (by the CKD-EPI equation) was 50.32 ± 19.21 mL/min/1.73 m2. The results after 12 months of treatment with once-weekly SC semaglutide are shown in Table 2. The mean HbA1c declined by −0.73 ± 1.09% (from 7.57 ± 1.36 to 6.83 ± 0.85; P < .001), with 57% of patients achieving the target of an HbA1c level of <7%. Patients had a mean weight loss of −6.95 ± 6.0 kg (from 98.48 ± 16.6 to 91.53 ± 16.42; P < .001), with 59% of patients achieving the target of −5% of total body weight. UACR declined by 53%, an average of −162.21 ± 365.77 mg/g [from 349.49 ± 863.2 to 187.28 ± 897.39; P < .001; median 78.5 mg/g (IQR 26–199.5) to 40 (6–103), P < .001). By subgroups according to baseline albuminuria, we observed that the group with macroalbuminuria (UACR >300 mg/g) presented a more significant albuminuria decrease at 6 and 12 months, 39% (P < .05) and 51% (P < .001), respectively (see Figure 2) and was the group with a higher percentage (58.3%) of T2D patients treated with SGT2i (see Table 3). Surprisingly, the greatest weight loss occurred in the group with microalbuminuria (−8.32 ± 5.59 kg, P < .01 in UACR 30–300 mg/g; −6.74 ± 6.07 kg, P < .01 in UACR >300 mg/g; and −4.35 ± 6.05 kg, P < .01 in UACR <30 mg/g).

FIGURE 1:

Previous patients comorbidity, risk factors and cardiovascular diseases N(%)

Table 2.

Data at baseline and after 12 months of treatment with semaglutide

Characteristics	Baseline (n = 122)	12 months(n = 115)	Change after 12 monthsa	P-value
Glucose (mg/dl), mean ± SD	148 ± 52.44	122.84 ± 25.96	−25.95 ± 47.53	<.001
HbA1c (%), mean ± SDPatients with HbA1c <7, n (%)Mild hypoglycemia, n (%)	7.57 ±1.3644 (36)	6.83 ± 0.8566 (57.4)9 (7.4)	−0.73 ± 1.09	<.001
Basal insulin (IU), mean ± SD	44.19 ± 8.37	32.41 ± 8.98	−11.78 ± 15.17	<.005
Rapid-acting insulin (IU), mean ± SD	22.35 ± 9.35	14.58 ± 9.89	−7.77 ± 11.93	.43
Weight (kg), mean ± SD	98.72±16.80	91.53±16.42	−6.95 ± 6.00	<.001
BMI (kg/m2), mean ± SDWeight loss >5%, n(%)	35.8 ± 4.79	33.33 ± 4.7768 (59)	−2.54 ± 2.22	<.001
UACR (mg/g), mean ± SDUACR (mg/g), median (Q1–Q3)	349.49 ± 863.1678.5 (26−199.5)	187.28 ± 497.3940 (6−103)	−162.21 ± 365.77	<.001<.001
eGFR (CKD-EPI;mL/min/1.73 m2), mean ± SD	50.32 ± 19.21	52.55 ± 19.26	2.2 ± 0.38	.07
Systolic BP (mmHg), mean ± SD	129.95 ± 11.27	120.09 ± 09	−9.85 ± 13.34	<.001
Diastolic BP (mmHg), mean ± SD	77.05 ± 8.95	71.12 ± 7.83	−5.92 ± 7.43	<.001
Hypertensive medications/patient, n (%)RAS blockade, n (%)	1.61 (98.4)117 (95.9)	1.49 (95.9)115 (95)
LDL cholesterol (mg/dl), mean ± SD	80.47 ± 32.50	69.68 ± 23.67	−10.79 ± 28.21	<.001
TG (mg/dl), mean ± SD	190.67 ± 113.20	161.51 ± 94.76	−29.15 ± 91.47	.001
Lipid-lowering medications/patients, nStatins,bn (%)Fibrates	1.0116 (95.1)28 (22.9)	1.06120 (98.4)20 (17.4)
Semaglutide withdrawal, n (%)		7 (5.7)

Q1: quartile 1; Q3: quartile 3; LDL: low-density lipoprotein; TG: triglycerides.

aFinal-initial.

bIncludes ezetimibe.

Table 3.

Baseline and changes in oral antidiabetic, antihypertensive and dyslipidemia medication according to basal albuminuria category

Medication by baseline UACR	Baseline, n (%)	24 months, n (%)
SGLT2i (mg/g)>30030–300<30	52 (42.6) 13 (54.2)26 (39.4)13 (40.6)	56 (45.9) 14 (58.3)29 (43.9)13 (40.6)
Metformin (mg/g)>30030–300<30	61 (50) 10 (41.7)37 (56.1)14 (43.8)	44 (36.1) 7 (29.2)31 (47)6 (18.8)
DPP4i (mg/g)>30030–300<30	19 (15.6) 1 (4.2)11 (16.7)7 (21.9)	0
Pioglitazone (mg/g)>30030–300<30	1 (0.8) 001 (3.1)	2 (1.7) 1 (4.2)01 (3.1)
anti-HTA (mg/g)>30030–300<30	120 (98.4) 23 (95.8)65 (98.5)32 (100)	117 (95.9) 23 (95.8)65 (98.5)29 (90.6)
RASi (mg/g)>30030–300<30	117 (95.9) 23 (95.8)64 (97.5)31 (98.5)	116 (95.1) 23 (95.8)64 (97.5)30 (97.4)
Statinsa (mg/g)>30030–300<30	116 (95.1) 23 (95.8)63 (95.5)30 (93.8)	120 (98.4) 24 (100)65 (98.5)31 (96.9)

anti-HTA: antihipertensive medication (includes β-blockers, calcium channel blockers, α-blockers, mineralocorticoidnŁ agonists and diuretics); DPP4i, dipeptyldipeptidase-4 inhibitor; RASi: renin–angiotensin system inhibitor. aIncludes ezetimibe.

Effect of semaglutide treatment on albuminuria: UACR after 6 and 12 months of treatment with semaglutide as compared with baseline. *P < .05; **P < .006. Q1: quartile 1; Q3: quartile 3.

The mean eGFR (by the CKD-EPI equation) remained stable, with a small nonsignificant increase of 2.2 mL/min/1.73 m2. Systolic and diastolic BP levels decreased by an average of 9.85 ± 13.34 mmHg (from 129.95 ± 11.27 to 120.09 ± 6.09; P < .001) and 5.92 ± 7.43 mmHg (from 77.05 ± 8.95 to 71.12 ± 7.83; P < .001), respectively. Low-density lipoprotein cholesterol and triglyceride levels decreased by 10.79 ± 28.21.1 mg/dl (P < .001) and 29.15 mg/dl ± 91.47 (P < .001), respectively. Seven percent of patients on insulin had mild hypoglycemic episodes. The patients in this study decreased the need for basal and rapid-acting insulin by 20.25% [−11.78 IU (95% CI 8.02–15.74; P < .005)] and 23% [−7.77 IU (95% CI 0.56–14.97; P = ns)], respectively. Semaglutide was suspended in 5.7% of cases; 65% of these suspensions were due to digestive intolerance. Nausea (1.6%) and vomiting

(0.8%) were the most frequent digestive problems, followed by constipation (0.8%) and diarrhea (0.6%). Weight loss failure was the second most common reason to stop semaglutide in 30% of the cases and in 5% it was by patient decision.

DISCUSSION

In recent years, SGLT2i and GLP-1RA have been associated with improved cardiovascular and kidney outcomes. GLP-1RA could be used for adequate glucose control in multiple stages of diabetic kidney chronic disease (DKD) without an increased risk of hypoglycemia and with additional benefits in terms of weight reduction, cardiovascular outcomes and kidney outcomes [10, 11]. On the other hand, sulfonylureas, insulin or repaglinide increase the risk of hypoglycemic episodes and metformin has been associated with lactic acidosis [12-14]. Epidemiological data support the relationship between glucose control and nephropathy in diabetes [15]. However, a major challenge in the management of DKD is the increased risk of hypoglycemia. Therefore, careful individualized glucose control, medication prescription, patient education, therapeutic planning and vigilance for hypoglycemia are all important components in the management of patients with DKD. The Kidney Disease: Improving Global Outcomes guidelines in DKD recommend individualizing HbA1c targets, ranging from <6.5% to <8.0% in patients not treated with dialysis. Safe achievement of lower HbA1c targets (<6.5% or <7.0%) may be facilitated by selection of antihyperglycemic agents that are not associated with hypoglycemia [15, 16]. The GLP-1RA liraglutide [17, 18], dulaglutide [19] and SC semaglutide have been shown to improve glucose control in DKD patients with an eGFR >15 mL/min/1.73 m2 and oral semaglutide [20, 21] with an eGFR >30 mL/min/1.73 m2.

Liraglutide obtained better glucose control and weight reduction, with lower hypoglycemia risk compared with placebo [22]. In the PIONEER 5 trial, once-daily oral semaglutide 14 mg was superior to placebo in decreasing HbA1c and body weight in DKD patients with an eGFR of 30–59 mL/min/1.73 m2, with few hypoglycemic episodes [23]. In our study, 57% of patients reached an HbA1c <7% with a significant 20% decrease in basal insulin needs, and 7% of patients had mild hypoglycemic episodes. Therefore we safely ameliorated glucose control in our DKD patients by reducing the risk of hypoglycemia.

In the LEADER trial [17-24], patients with an eGFR >15 mL/min/1.73 m2 were included, 23% of them had an eGFR <60 mL/min/1.73 m2, 36% had albuminuria >30 mg/g and 10% had albuminuria >300 mg/g. Liraglutide decreased by 22% the risk of a secondary composite kidney endpoint (new-onset macroalbuminuria, sustained serum creatinine duplication, initiation of RRT or renal death). This benefit was mainly at the expense of macroalbuminuria reduction. Furthermore, in the SUSTAIN 6 trial [20] they also included patients with an eGFR >15 mL/min/1.73 m2, 28.5% of them had an eGFR <60 mL/min/1.73 m2 and 12.7% had baseline macroalbuminuria. SC semaglutide showed a 36% risk reduction in a secondary combined kidney endpoint (new-onset macroalbuminuria, doubling serum creatinine reaching an eGFR <45 mL/min/1.73 m2, initiation of renal replacement therapy or renal death). The renal benefit was mainly related to the treatment impact on new-onset macroalbuminuria (2.7% in the semaglutide group versus 4.9% in the placebo group; P = .001). Post hoc pooled analyses of LEADER and SUSTAIN 6 showed a 30% reduction in albuminuria and that regression to micro- or normoalbuminuria occurred for all degrees of albuminuria in patients treated with semaglutide [25]. In the same way the subanalysis confirmed a significantly milder eGFR decline with 1 mg SC semaglutide in patients with baseline eGFR between 30 and 60 mL/min/1.73 m2 [2-26]. Liraglutide or semaglutide are small-sized GLP-1RAs that are able to cross the blood–brain barrier and reach satiety centers. Therefore a greater effect on body weight has been seen in clinical trials with them than with large-sized GLP-1RAs such as dulaglutide and albiglutide [27]. In our study, 56% of patients were switched from another GLP-1RA to semaglutide (64.8% from liraglutide, 31% dulaglutide and 3.2% exenatide-LAR). The most frequent indication for the change was failure to lose weight with the previous GLP-1RA. In mice, it has been shown that semaglutide accesses and activates regions of the brain involved in reward behavior and food intake [28]. In SUSTAIN 10, SUSTAIN 7 and SUSTAIN 3, semaglutide has been the more powerful molecule to lower body weight compared with liraglutide, dulaglutide and exenatide, respectively [29-31]. In our study, patients decreased total body weight by >5% in 59% of the population. However, by subgroups, it is not the patients with greater reductions in albuminuria where we find the greatest weight loss. We hypothesize that the effect of semaglutide in the decrease of albuminuria could be more related to direct renal mechanisms than to indirect ones such weight loss, glucose or BP control. Moreover, the GLP1-ARs exert their nephroprotective effect by indirect and direct mechanisms. It seems that the local or receptor-mediated mechanisms on the kidney are responsible for >60% of the kidney benefit. In the LEADER study, the difference in kidney outcome was not altered by adjustment for a change in glucose control, body weight or systolic BP (indirect mechanisms) [17]. Experimental data demonstrated that the improvements in renal outcomes may be ascribed in part to the anti-inflammatory antioxidative and hemodynamic properties of the GLP1-RAs [32].

In our study, mean systolic and diastolic BP decreased by an average of 9.85 ± 13.34 mmHg (from 129.95 ± 11.27 to 120.09 ± 6.09; P < .001) and 5.92 ± 7.43 mmHg (from 77.05 ± 8.95 to 71.12 ± 7.83; P < .001), respectively. These decreases are greater than in SUSTAIN 6 [20], where the mean systolic BP decreased 3.4 and 5.4 mmHg among those receiving 0.5 and 1.0 mg of SC semaglutide, respectively. In other SUSTAIN trails, SC semaglutide was generally associated with reductions in systolic BP that achieved statistical significance (except for SUSTAIN 1, 7 and 10) [31, 33–36]. Diastolic BP was generally reduced, but differences were not significant versus comparators. In our study, weight loss may have contributed to better BP control. Interestingly, 72% of our patients were at high risk of CKD progression with micro- or macroalbuminuria [mean for all the patients 349 ± 843 mg/g; median 78.5 (IQR 26–199.5)]. In addition, between 30 and 40% already have an underlying cardiovascular disease. The positive effect of semaglutide in terms of decreasing albuminuria may suggest a delay in progression toward ESKD; however, we were unable to demonstrate a positive effect in terms of eGFR delay. The lack of this positive effect may be related to the small sample size and the short-term follow-up. In post hoc pooled analysis from SUSTAIN 6 and PIONEER 6 of SC and oral semaglutide, respectively [37], in 6480 subjects, researchers observed an annual rate of eGFR change of 0.60 mL/min/1.73 m2 (95% CI 0.31–0.90; P < .0001) lower with semaglutide compared with placebo in the first year estimated to treatment difference (ETD). Specifically in the subgroup with an eGFR ≥30–<60 mL/min/1.73 m2 at baseline, the ETD for semaglutide versus placebo was 1.07 mL/min/1.73 m2/year (95% CI 0.46–1.68; P < .01). The authors conclude that although benefits were observed in the overall population, the findings indicate that the primary benefit may be observed in those with established CKD. To date, there are no published GLP-1RA trials with a primary endpoint of kidney events or enrolling only DKD patients. Although real-life studies tend to obtain poorer results than clinical trials, the positive results achieved in this study may be explained in part by good therapeutic compliance, which was possibly related to a higher frequency of visits.

One of the main limitations of our study is that the design was uncontrolled and there were multiple interventions concomitant with semaglutide treatment. Also, the existence of modifications in the individual behavior of patients in response to their awareness of being observed cannot be ruled out. In addition, this was an exploratory analysis and its results must be interpreted with caution. Currently there is an ongoing randomized clinical trial, the FLOW trial [A Research Study to See How Semaglutide Works Compared to Placebo in People With Type 2 Diabetes and Chronic Kidney Disease (FLOW) (NCT03819153)], is assessing the impact of semaglutide on primary kidney outcomes in DKD.

In conclusion, in real-life clinical practice, 12 months of treatment with SC semaglutide in patients with T2D and CKD at high risk of progression ameliorates metabolic control and weight. Furthermore, we also demonstrated a reduction in albuminuria of >50% in the population with macroalbuminuria, improved BP and lipid control with a decrease in insulin needs.