Renal and metabolic effects of semaglutide plus canagliflozin vs canagliflozin alone in type 2 diabetic nephropathy

Abstract

BACKGROUND

Diabetic nephropathy (DN) is a major microvascular complication of type 2 diabetes and a leading cause of end-stage renal disease. Both canagliflozin and semaglutide are recommended for diabetic nephropathy, yet direct evidence comparing their combination versus monotherapy remains lacking.

AIM

To investigate the efficacy of combination therapy with semaglutide and canagliflozin compared with canagliflozin alone in patients with type 2 DN.

METHODS

A retrospective study was conducted using data sourced from the electronic medical record system of Henan Provincial People’s Hospital. The study included patients with DN treated from October 2022 to March 2024. Patients were divided into two groups on the basis of their treatment: Canagliflozin monotherapy (CM) and canagliflozin and semaglutide combination therapy (CSCT). Renal function, glucose metabolism, lipid profiles, pancreatic function, oxidative stress, and inflammatory markers were assessed at baseline and 6 months after treatment. Adverse events were monitored throughout the study period.

RESULTS

Among 211 patients (107 CM, and 104 CSCT), the CSCT group demonstrated superior outcomes. The albumin-to-creatinine ratio decreased more significantly (145.87 mg/g vs 158.11 mg/g, P = 0.002), and more improvements were found in the glycated hemoglobin A1c (7.08% vs 7.42%, P = 0.005), low-density lipoprotein (86.74 mg/dL vs 94.86 mg/dL, P = 0.032), serum free fatty acids (0.46 mmol/L vs 0.52 mmol/L, P = 0.002), and insulin resistance index (3.94 vs 4.08, P = 0.011). Meanwhile, the islet β-cell function increased (51.22 vs 49.36, P = 0.022). The total adverse event rates were comparable between the two groups, and no significant increase was observed in the gastrointestinal adverse events (P = 0.360).

CONCLUSION

CSCT provided significant improvements in multiple metabolic and renal parameters without increasing adverse events, thus highlighting its potential benefits for patients with DN.

Key Words: Diabetic nephropathy; Semaglutide; Canagliflozin; Combination therapy; Glucose-lipid metabolism; Renoprotection

Core Tip: This retrospective study demonstrates that, in individuals with type 2 diabetes and nephropathy, the addition of semaglutide to canagliflozin monotherapy provides better renal protection and metabolic improvements compared to canagliflozin alone. The combination significantly reduced proteinuria, improved glycemic and lipid control, enhanced β-cell function, alleviated systemic inflammation and oxidative stress, without increasing adverse events. It offers an effective and safe intensification therapy option, particularly for those who have not achieved optimal levels of proteinuria or metabolic parameters under sodium-glucose cotransporter-2 inhibitor monotherapy.

INTRODUCTION

Diabetic nephropathy (DN) is a common and severe microvascular complication of type 2 diabetes, characterized by persistent proteinuria, gradual decline in glomerular filtration rate (GFR), and hypertension[1]. It is the leading cause of end-stage renal disease (ESRD) globally[2]. Clinically, DN manifests symptoms such as proteinuria, edema, and hypertension[3]. The pathogenesis of DN involves complex interactions among metabolic disorders like chronic hyperglycemia and dyslipidemia and hemodynamic factors such as intraglomerular hypertension, oxidative stress, chronic low-grade inflammation, and activation of fibrotic pathways[4,5]. These factors collectively contribute to glomerulosclerosis, tubulointerstitial fibrosis, and podocyte injury.

Recent advances in DN management focused on intensive glucose and blood pressure control, particularly through the use of renin-angiotensin-aldosterone system inhibitors[6]. However, the introduction of sodium-glucose cotransporter-2 inhibitors (SGLT2is) and glucagon-like peptide-1 receptor agonists (GLP-1 RAs) marks significant progress[7]. Large cardiovascular outcome trials and dedicated renal studies have shown that these two classes of drugs not only reduce blood glucose but also provide robust cardio-renal protection, establishing them as foundational therapies for type 2 DN[8]. Modern treatment strategies for DN have evolved from traditional monotherapy with renin-angiotensin system inhibitors (RASis) to a multitargeted approach centered around SGLT2i, GLP-1 RA, and finerenone[9]. However, despite clear guideline recommendations, a significant under-prescription of these key drugs, particularly RASi and SGLT2i, in clinical practice remains, especially in primary care settings and among specific racial/ethnic groups, leading to substantial treatment gaps. Research indicates that this gap partly stems from clinicians’ insufficient recognition that these medications are indicated for DN as an independent condition, and their cardio-renal protective effects remain crucial even when patients achieve optimal glycemic and blood pressure control[10]. Despite these efforts, achieving optimal outcomes remains challenging due to the multifactorial nature of DN and its associated complications. Therefore, the present study aimed to investigate the effects of canagliflozin and semaglutide combination therapy (CSCT) compared with canagliflozin monotherapy (CM) on renal function and glucose-lipid metabolism in patients with confirmed DN.

The potential mechanisms linking hyperglycemia, dyslipidemia, and the progression of DN involve multiple intertwined pathways. Oxidative stress caused by excessive reactive oxygen species (ROS), along with chronic inflammation characterized by increased levels of tumor necrosis factor-alpha (TNF-α) and C-reactive protein (CRP), play crucial roles in endothelial dysfunction, podocyte injury, and renal fibrosis[11,12]. Additionally, changes in adipokine signaling, such as increased levels of aspartate and free fatty acids (FFAs), can lead to insulin resistance, β-cell dysfunction, and even direct kidney damage[13]. SGLT2is represent a notable pathway. These drugs reduce glucose reabsorption in the kidneys, lowering blood glucose levels[14]. They also improve renal hemodynamics and decrease intraglomerular pressure while showing anti-inflammatory and antifibrotic properties. GLP-1 RAs effectively control blood glucose and body weight and provide additional cardiovascular benefits[15]. New evidence indicates that GLP-1 RAs offer direct renal protection through anti-inflammatory and antioxidative pathways, possibly mediated by regulating adipokines and inflammatory cytokines[16]. The FLOW study, which used renal outcomes as the primary endpoint for the first time, confirmed that the GLP-1 RA semaglutide provides significant renal protection for patients with type 2 diabetes and chronic kidney disease, further solidifying the therapeutic role of this class of drugs[17]. Combining SGLT2is with GLP-1 RAs can produce synergistic effects, enhancing renal protection and metabolic control. Understanding the interactions among these mechanisms is essential for developing more effective DN treatment strategies. This combination therapy may offer new approaches to slowing down DN progression and improving overall patient health.

Although the combination therapy of SGLT2i and GLP-1 RA shows synergistic potential, direct comparative studies specifically evaluating their combined efficacy in treating DN remain limited, representing a significant research gap. This retrospective cohort study focuses on evaluating the efficacy and safety of CSCT compared with CM in patients with DN. The primary goal is to determine if the combination therapy offers better outcomes than monotherapy. The effect of this treatment on key renal parameters, glycemic control, lipid profiles, pancreatic function, oxidative stress markers, and inflammation was thoroughly assessed. The novelty of this study lies in its specific focus on the combined effects of these two drugs on cardio-renal risk factors in the patient with DN population over a 6-month period. By understanding the potential benefits of this combination therapy, valuable insights can be provided for optimizing strategies aimed at protecting renal function and improving metabolic health in patients with a high risk of diabetes.

MATERIALS AND METHODS

Case selection

This retrospective study included 211 patients with DN admitted to Henan Provincial People’s Hospital from October 2022 to March 2024, and demographic information of the patients was collected through the case system. This retrospective study poses no potential harm to patients because it uses de-identified patient data. Thus, informed consent was waived. This waiver and the study have been approved by the Ethics Review Committee of Henan Provincial People’s Hospital, meeting relevant regulatory and ethical standards.

Inclusion and exclusion criteria

Inclusion criteria: (1) Age > 18 years; (2) In accordance with the American Diabetes Association guidelines[18], DN is diagnosed in patients with diabetic who, after excluding other potential causes of kidney damage, exhibit persistent albuminuria [such as urinary albumin-to-creatinine ratio (ACR) ≥ 30 mg/g or urinary albumin excretion rate ≥ 30 µg/minute and/or estimated GFR (eGFR) < 60 mL/minute/1.73 m²]; (3) Therapy with canagliflozin or canagliflozin plus semaglutide; (4) No history of allergies to the study drugs; and (5) Complete medical records.

Exclusion criteria: (1) History of pancreatitis; (2) History of diabetic ketoacidosis; (3) History of myocardial infarction; (4) History of stroke; (5) Hospital admission for unstable angina; (6) Transient ischemic attack within 180 days before screening; and (7) Heart failure.

Grouping standards

Patient information was collected from the medical record system, and the patients were grouped on the basis of their medication regimen. The CM group (n = 107) was treated with canagliflozin tablets (H20193392, Jiangsu Hengrui Medicine Co., Ltd., China), 0.1 g per time, once daily, orally before breakfast. The CSCT group (n = 104) received semaglutide subcutaneous injection (SJ20210015, Novo Nordisk A/S, Denmark) on the basis of canagliflozin treatment, with an initial dose of 0.25 mg, once weekly. After 4 weeks, the dose was increased to 0.5 mg per time, once weekly. Both groups were treated continuously for 6 months, with the treatment and follow-up period being 6 months in duration.

Evaluation methods for renal function and glucose-lipid metabolism

Methods for evaluating renal function: All operations were carried out between 08:30 and 13:00 on each day. The patients had their normal breakfast and morning dose of insulin before the investigations, during which they rested in a supine position and stood up only to pass urine. They drank 150-200 mL of tap water per hour during the study period.

At 09:00, a single intravenous injection of 3.7 MBq of edetic acid (H20059953, Hebei Zhitong Chemical Co., Ltd., China) labeled with sodium chromate (H10960238, Atom Hi-Tech Co., Ltd., China) was administered, and venous blood samples were collected from the other arm of the patients at 180, 200, 220, and 240 minutes post-injection to measure the GFR by determining the radioactivity within the samples. The small underestimation (10%) of chromium (III)-ethylenediaminetetraacetate complex (Cr-EDTA) clearance vs inulin clearance was corrected for by multiplying the Cr-EDTA clearance by 1.10. The mean day-to-day coefficient of variation in the GFR of each patient was 4%. Serum creatinine was measured using a time reaction technique.

Methods for measuring blood glucose indicators: The participants were asked to fast for 14 hours prior to the test, and venous blood samples were obtained from the arm. Whole blood was collected in vacuum tubes containing potassium oxalate (CAS 6487-48-5, Sigma-Aldrich, Germany) and sodium fluoride (CAS 7681-49-4, Sigma-Aldrich, Germany) and centrifuged (KL05A, Kaida, China) immediately at 1500 g for 10 minutes. Plasma was frozen at -70 °C until analysis. Fasting plasma glucose (FPG) was measured using a hexokinase reference method (Roche Diagnostics, Montclair, NJ, United States). The coefficient of variation for this method was 1.6% at a glucose concentration of 5.5 mmol/L (100 mg/dL) and 1.8% at a glucose concentration of 27.8 mmol/L (500 mg/dL). The hemoglobin A1c (HbA1c) content in the collected whole blood samples was detected using reverse-phase cation exchange chromatography (HA 8160 analyzer; Menarini, Florence, Italy).

Methods for measuring lipid levels: The patients needed to fast for 12 hours in advance. Afterwards, 5 mL of venous blood was collected and immediately centrifuged at 3000 rpm for 10 minutes (KL05A, Kaida, China) to separate the serum. Once the centrifugation was complete, the inspector promptly aliquoted the serum samples into tubes designated for testing. Using an automatic biochemical analyzer (7600; Hitachi, Tokyo, Japan), the concentrations of tot-cholesterol, triglycerides, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) were measured using cholesterol Gen.2 (Roche, Switzerland), Triglycerides Gen.2 (Roche, Switzerland), HDL-C Gen.3 (Roche, Switzerland), and LDL-C Gen.3 (Roche, Switzerland), respectively.

Methods for measuring islet function and serum asprosin and FFA levels: The levels of asprosin, FFAs, and fasting insulin (FINS) were detected using the human asprosin ELISA kit (EH4176, Wuhan Fine Biotech Co., Ltd., China), the FFA content (enzymatic method) kit (MC927 L48, Shanghai Enzyme-linked Biotechnology Co., Ltd., China), and the human insulin ELISA kit (ml064302, Shanghai Enzyme-linked Biotechnology Co., Ltd., China), respectively. Islet β-cell function (HOMA-β) and insulin resistance index (HOMA-IR) were then calculated using the following formulas: HOMA-β = 20 × FINS/(FPG-3.5) and HOMA-IR = FINS × FPG/22.5, respectively.

Methods for measuring oxidative stress and inflammatory factor levels: The levels of malondialdehyde (MDA), superoxide dismutase (SOD), CRP, and TNF-α were detected using the MDA content detection kit (NM-W-0104, Nuomin Keda Biotechnology Co., Ltd., China), the SOD activity detection kit (NM-W-0101, Nuomin Keda Biotechnology Co., Ltd., China), the CRP detection kit (YBEA821Hu, Shanghai Yubo Biotechnology Co., Ltd., China), and the human TNF-α ELISA kit (CSB-E04740h, Wuhan Huamei Bioengineering Co., Ltd., China), respectively.

Adverse reactions

The incidence of adverse reactions was determined during the 6-month treatment period. The adverse reactions observed in this study included gastrointestinal (GI) adverse events, urinary tract infections, hypoglycemia, nervous system disorders, and diabetic retinopathy. GI adverse events encompasses symptoms such as abdominal pain, bloating, nausea, vomiting, retching, and acid regurgitation. Urinary tract infections refer to the presence of symptoms and/or signs referable to the urinary tract, accompanied by significant bacteriuria. Biochemical hypoglycemia is defined as plasma glucose ≤ 3.9 mmol/L (70 mg/dL) in diabetes[19]. Disorders of the nervous system comprise diseases of the central nervous system (CNS), peripheral nervous system, and autonomic nervous system, resulting from infectious, degenerative, vascular, or traumatic causes. These disorders disrupt normal neural function, manifesting as neurological or psychiatric symptoms. Diabetic retinopathy is a progressive, potentially blinding disorder caused by chronic hyperglycemia-induced damage to retinal capillaries and neurons.

Efficacy evaluation criteria

“Markedly effective” indicates that FPG and HbA1c are within the normal range, and the body mass index (BMI) decreases. “Effective” indicates that FPG decreases or is between 6.0 mmol/L and 7.8 mmol/L, and BMI decreases. “Ineffective” is defined as not meeting the above criteria[20].

Statistical analysis

Data analysis was conducted using SPSS (version 29.0) statistical software (SPSS Inc., Chicago, IL, United States). Continuous variables were assessed for normality by using Shapiro-Wilk test. Normally distributed data are presented as mean ± SD and compared using t-tests. Categorical data are presented as n (%). χ² test was used when expected frequency ≥ 5. Yates’ correction was applied if the expected frequency was 1-5. Statistical significance was set at P < 0.05.

For continuous outcome measures, independent sample t-test was used to directly compare the endpoint measurements between the CM group and the CSCT group at 6 months post-treatment.

A post-hoc power analysis was performed using G*Power software (version 3.1.9.7) to evaluate whether the sample size in this study was adequate. The effect size used in this analysis was calculated based on the observed values of the primary renal endpoint (urinary ACR) between the two groups at 6 months post-treatment, with a Cohen's d value of 0.42 and a 95% confidence interval (CI): 0.15-0.69. The effect size (Cohen's d = 0.42) used in this analysis was calculated based on the mean difference and pooled standard deviation of the primary renal endpoint (urinary ACR) between the two groups after 6 months of treatment, with a 95%CI: 0.15-0.69. Based on this, with α = 0.05 (two-tailed test), the theoretical sample size required to achieve 80% statistical power is approximately 90 cases per group. The actual sample size in this study exceeded this theoretical requirement, demonstrating that the current sample size was sufficient to support the reliability and validity of the research findings.

RESULTS

General characteristics

In the comparison of general characteristics between the CM group and the CSCT group, several parameters showed no significant differences (all P > 0.05; Table 1). These characteristics included age, gender distribution, BMI, weight, educational level, current smoking status, pulse rate, duration of diabetes mellitus, systolic blood pressure, diastolic blood pressure, left ventricular ejection fraction categories, and DN staging. This finding indicated that the two groups were well-matched in terms of baseline characteristics.

Table 1 General characteristics, n (%).

Parameter	CM group (n = 107)	CSCT group (n = 104)	t/χ²	P value
Age (years)	63.12 ± 6.14	62.86 ± 6.26	0.305	0.761
Gender			0.004	0.949
Male	55 (51.40)	53 (50.96)
Female	52 (48.60)	51 (49.04)
BMI (kg/m2)	28.43 ± 3.46	28.51 ± 3.52	0.163	0.870
Weight (kg)	79.63 ± 14.11	80.12 ± 14.23	0.251	0.802
Educational level			0.024	0.988
Primary	18 (16.82)	17 (16.35)
Secondary	54 (50.47)	52 (50.00)
Tertiary	35 (32.71)	35 (33.65)
Current smoking	23 (21.50)	21 (20.19)	0.054	0.816
Pulse rate (beats per min)	74.13 ± 5.11	74.22 ± 5.21	0.128	0.899
Duration of diabetes mellitus (years)	10.52 ± 4.10	11.12 ± 3.97	1.068	0.287
Systolic blood pressure (mmHg)	138.21 ± 4.12	138.37 ± 4.06	0.274	0.784
Diastolic blood pressure (mmHg)	78.34 ± 3.26	77.89 ± 3.45	0.983	0.327
LVEF			0.213	0.899
< 40%	18 (16.82)	20 (19.23)
40% to < 50%	26 (24.30)	25 (24.04)
≥ 50%	63 (58.88)	59 (56.73)
Diabetic nephropathy staging			0.011	0.916
G3a	85 (79.44)	82 (78.85)
G3b	22 (20.56)	22 (21.15)

BMI: Body mass index; LVEF: Left ventricular ejection fraction; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Comparison of renal function between two groups before and after treatment

In the comparison of renal function between the two groups before and after treatment, the serum creatinine levels and GFR did not differ significantly between the CM and CSCT groups at both time points (all P > 0.05; Table 2). However, a significant difference was observed in the ACR between the two groups at 6 months post-treatment (P = 0.002), indicating that the CSCT group had better outcomes compared to the CM group. This finding suggests that the intervention in the CSCT group may have had a beneficial effect on reducing proteinuria, which is an important indicator of renal health.

Table 2 Comparison of renal function between two groups of patients before and after treatment.

Parameter	Time	CM group (n = 107)	CSCT group (n = 104)	t	P value
Serum creatinine (mg/dL)	Baseline	1.04 ± 0.32	1.04 ± 0.29	0.122	0.903
	6 months	1.05 ± 0.35	1.06 ± 0.31	0.265	0.791
GFR (mL/minute/1.73 m2)	Baseline	50.83 ± 5.01	50.92 ± 5.19	0.126	0.900
	6 months	50.13 ± 4.61	50.52 ± 4.87	0.603	0.547
ACR (mg/g)	Baseline	211.51 ± 30.11	207.54 ± 30.15	0.956	0.340
	6 months	158.11 ± 28.85	145.87 ± 29.23	3.062	0.002

GFR: Glomerular filtration rate; ACR: Albumin/creatinine ratio; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Comparison of blood glucose levels between two groups before and after treatment

In the comparison of blood glucose levels between the two groups before and after treatment, the baseline measurements for FPG and HbA1c showed no significant differences (all P > 0.05; Figure 1). However, at 6 months post-treatment, the CSCT group had significantly lower FPG (P = 0.004) and HbA1c (P = 0.005) compared to the CM group. This finding suggests that, compared to CM, CSCT may offer greater benefits in improving glycemic control among patients with diabetes.

Figure 1 Comparison of blood glucose levels between two groups of patients before and after treatment. A: Fasting plasma glucose (mg/dL); B: Hemoglobin A1c (%). FPG: Fasting plasma glucose; HbA1c: Hemoglobin A1c; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Comparison of blood lipid levels between two groups before and after treatment

In the comparison of blood lipid levels between the two groups before and after treatment, the baseline measurements for total cholesterol, HDL, LDL, and triglycerides showed no significant differences (all P > 0.05; Table 3). At 6 months post-treatment, significant differences were observed between the two groups. The total cholesterol levels in the CSCT group were significantly lower compared to the CM group (P = 0.043). Similarly, the HDL levels were significantly higher in the CSCT group compared to the CM group at this time point (P = 0.045), indicating better management of "good" cholesterol. Additionally, the LDL levels were significantly lower in the CSCT group compared to the CM group at 6 months (P = 0.032), suggesting improved control of "bad" cholesterol. Furthermore, the triglyceride levels in the CSCT group were significantly lower compared to those in the CM group at 6 months (P = 0.040). These results highlight that CSCT was associated with more favorable lipid profiles compared to CM.

Table 3 Comparison of blood lipid levels between two groups of patients before and after treatment.

Parameter	Time	CM group (n = 107)	CSCT group (n = 104)	t	P value
Tot-cholesterol (mg/dL)	Baseline	177.23 ± 34.56	180.25 ± 32.56	0.653	0.515
	6 months	171.45 ± 32.69	162.41 ± 31.79	2.037	0.043
HDL (mg/dL)	Baseline	45.26 ± 9.56	45.15 ± 9.89	0.079	0.937
	6 months	45.34 ± 8.66	47.69 ± 8.28	2.017	0.045
LDL (mg/dL)	Baseline	101.49 ± 35.26	105.21 ± 35.46	0.764	0.446
	6 months	94.86 ± 27.64	86.74 ± 27.12	2.154	0.032
Triglycerides (mg/dL)	Baseline	145.32 ± 35.64	142.89 ± 35.39	0.496	0.621
	6 months	135.21 ± 28.22	127.56 ± 25.26	2.072	0.040

HDL: High density lipoprotein; LDL: Low density lipoprotein; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Comparison of islet function and serum asprosin and FFA levels between two groups before and after treatment

In the comparison of islet function and serum asprosin and FFA levels between the two groups before and after treatment, the baseline measurements for FFA, HOMA-IR, HOMA-β, and asprosin showed no significant differences between the CM group and the CSCT group (all P > 0.05; Table 4). However, at 6 months post-treatment, significant differences were observed. The CSCT group had significantly lower FFA levels than the CM group at 6 months (P = 0.002), indicating better control of serum FFAs. The CSCT group also showed significantly lower values of HOMA-IR, which measures insulin resistance, than the CM group at 6 months (P = 0.011), suggesting improved insulin sensitivity. In terms of HOMA-β, an indicator of islet β-cell function, the CSCT group demonstrated significantly higher values than the CM group at 6 months (P = 0.022), indicating enhanced β-cell function. Additionally, the serum asprosin levels in the CSCT group were significantly lower compared to the CM group at this time point (P = 0.012), further supporting the beneficial effects of the CSCT intervention.

Table 4 Comparison of islet function and serum asprosin and free fatty acids levels between two groups of patients before and after treatment.

Parameter	Time	CM group (n = 107)	CSCT group (n = 104)	t	P value
FFA (mmol/L)	Baseline	0.55 ± 0.17	0.56 ± 0.18	0.334	0.739
	6 months	0.52 ± 0.16	0.46 ± 0.12	3.172	0.002
HOMA-IR	Baseline	5.19 ± 0.45	5.22 ± 0.42	0.379	0.705
	6 months	4.08 ± 0.36	3.94 ± 0.45	2.582	0.011
HOMA-β	Baseline	45.99 ± 5.64	45.98 ± 5.56	0.007	0.994
	6 months	49.36 ± 5.87	51.22 ± 5.79	2.310	0.022
Asprosin (ng/mL)	Baseline	7.52 ± 1.19	7.57 ± 1.21	0.298	0.766
	6 months	6.33 ± 0.52	6.16 ± 0.48	2.528	0.012

FFA: Free fatty acid; HOMA-IR: Insulin resistance index; HOMA-β: Islet β-cell function; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Comparison of oxidative stress and inflammatory factor levels between two groups before and after treatment

The baseline measurements of MDA, SOD, TNF-α, and CRP did not reveal significant differences between the CM group and the CSCT group (all P > 0.05; Table 5). However, at 6 months post-treatment, notable distinctions emerged. The MDA levels in the CSCT group were significantly lower compared to the CM group (P = 0.020), suggesting reduced oxidative stress. The SOD activity was significantly higher in the CSCT group compared to the CM group (P = 0.010), indicating enhanced antioxidant defense mechanisms. The levels of TNF-α were significantly lower in the CSCT group compared to the CM group (P = 0.021), suggesting a reduction in systemic inflammation. Additionally, the CRP concentrations in the CSCT group were markedly lower compared to those in the CM group (P < 0.001), further supporting the anti-inflammatory effects of the CSCT intervention.

Table 5 Comparison of oxidative stress and inflammatory factor levels between two groups of patients before and after treatment.

Parameter	Time	CM group (n = 107)	CSCT group (n = 104)	t	P value
MDA (μmol/L)	Baseline	4.65 ± 1.09	4.68 ± 1.11	0.158	0.875
	6 months	4.12 ± 0.99	3.78 ± 1.13	2.336	0.020
SOD (U/mg prot)	Baseline	42.05 ± 5.27	42.69 ± 5.35	0.882	0.379
	6 months	45.49 ± 5.55	47.46 ± 5.46	2.597	0.010
TNF-α (ng/L)	Baseline	15.19 ± 3.12	15.26 ± 3.26	0.162	0.871
	6 months	10.43 ± 2.36	9.68 ± 2.29	2.325	0.021
CRP (mg/L)	Baseline	2.45 ± 0.42	2.47 ± 0.45	0.334	0.739
	6 months	1.74 ± 0.26	1.39 ± 0.24	10.144	< 0.001

MDA: Malondialdehyde; SOD: Superoxide dismutase; TNF-α: Tumor necrosis factor-α; CRP: C-reactive protein; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Adverse events occurring in two groups during treatment

The adverse events during the treatment process were compared between the CM and CSCT groups (Table 6). The GI adverse events showed no significant difference between the groups (P = 0.360). Similarly, the urinary tract infections did not differ significantly between the two groups (P = 0.519). The hypoglycemia rates were comparable, with no significant difference observed (P = 0.667). Nervous system disorders occurred without a significant difference between the groups (P = 0.688). Diabetic retinopathy incidence was likewise similar, with no significant difference detected (P = 0.980).

Table 6 Adverse events occurring in two groups of patients during the treatment process, n (%).

Parameter	CM group (n = 107)	CSCT group (n = 104)	χ²	P value
GI adverse events	11 (10.28)	15 (14.42)	0.838	0.360
Urinary tract infections	5 (4.67)	7 (6.73)	0.416	0.519
Hypoglycaemia	10 (9.35)	8 (7.69)	0.185	0.667
Nervous system disorders	3 (2.80)	5 (4.81)	0.161	0.688
Diabetic retinopathy	2 (1.87)	1 (0.96)	0.001	0.980

GI: Gastrointestinal; CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

Comparison of clinical efficacy between two groups

In terms of clinical efficacy comparison between the two groups, the distribution of outcomes was evaluated (Figure 2). The CSCT group showed a higher proportion of markedly effective results, while the CM group exhibited a higher proportion of effective results. Ineffective results were less common in the CSCT group (P = 0.041). This indicates that the effectiveness rate was better in the CSCT group compared to the CM group.

Figure 2 Comparison of clinical efficacy between two groups of patients. CM: Canagliflozin monotherapy; CSCT: Canagliflozin and semaglutide combination therapy.

DISCUSSION

This study compared the efficacy and safety of two treatment strategies in patients with DN. The results show that the combination therapy excelled in improving proteinuria, enhancing glucose and lipid metabolism, boosting β-cell function, reducing insulin resistance, and lowering oxidative stress and systemic inflammation. These multifaceted improvements provide notable clinical benefits to patients. First, the reduction in urinary protein helps slow down the progression of DN and reduces the risk of ESRD. Second, better control of blood glucose and lipids effectively prevents macrovascular and microvascular complications. Additionally, improvements in insulin sensitivity and reductions in inflammation contribute to a solid foundation for long-term metabolic health. These enhancements not only improve patient health in the short term but also provide a basis for sustained health over the long term. Notably, these benefits did not come with a significant increase in adverse events. This finding was crucial for understanding how to better manage diabetes because these improvements directly affect cardiovascular risk and other long-term health issues.

In this study, the combination therapy reduced proteinuria, which was a significant discovery. Proteinuria is a recognized marker of kidney damage and a key predictor of DN progression[21]. The improvement in ACR likely resulted from multiple synergistic renal protective mechanisms. Canagliflozin reduces intraglomerular hypertension by promoting tubuloglomerular feedback, while semaglutide mediates a wide range of intracellular signaling pathways in renal cells through the activation of the GLP-1 receptor[22,23]. At the molecular level, this synergistic effect may arise from the combined action of the GLP-1R/cAMP/PKA pathway activated by semaglutide and the improved cellular energy metabolism state promoted by canagliflozin, more effectively inhibiting the activation of the key pro-fibrotic pathway TGF-β/Smad[24]. PKA can phosphorylate and inhibit the transcriptional activity of Smad3, while canagliflozin reduces the upstream expression of TGF-β by improving mitochondrial function and lowering ROS levels[25]. This convergence at cellular signaling nodes jointly attenuates fibroblast activation and extracellular matrix deposition, thereby providing anti-fibrotic effects that surpass those of single-agent therapies[26]. These findings are consistent with those of existing literature indicating that SGLT2is and GLP-1 RAs target complementary pathways, hemodynamics, and inflammation in DN, thereby enhancing renal protection[27]. The eGFR did not show significant changes over 6 months, consistent with the understanding of SGLT2is. These drugs often cause initial hemodynamic adjustments, which may mask long-term structural benefits[28]. Therefore, longer follow-up periods are needed to assess whether eGFR could be better preserved.

The observed lipid-lowering effect highlights the metabolic diversity of this combination therapy. Previous studies indicated that effective glucose control could lead to positive changes in lipids, such as reductions in total cholesterol and LDL cholesterol[29]. The findings of the present study are consistent with these observations, showing an improvement trend in the lipid profiles of the CSCT group, suggesting additional benefits beyond glucose control. Canagliflozin moderately increased HDL, but when used in combination with semaglutide, the lipid improvements were more pronounced. Semaglutide activates GLP-1 receptors on hepatocytes and adipocytes, triggering the cAMP/PKA signaling cascade, which inhibits SREBP-1c maturation and nuclear translocation, reducing hepatic fatty acid and triglyceride synthesis[30]. It also enhances PPAR-γ activity, promoting normal adipocyte differentiation and fatty acid storage. Canagliflozin uses its diuretic-glucosuric effect to induce mild energy deficiency, activating AMPK. Activated AMPK phosphorylates and inhibits SREBP-1c and, together with PPAR-γ, promotes fatty acid oxidation[31]. The two signaling pathways intersect and complement each other at the key transcription factors SREBP-1c and PPAR-γ, ultimately leading to decreased hepatic lipid synthesis and enhanced systemic lipid oxidation[32]. The reduction in FFA further supports this because increased FFAs accelerate liver very low-density lipoprotein production and cause ectopic lipid deposition. Mechanistically, semaglutide likely inhibits adipose tissue lipolysis via CNS regulation[33]. Previous studies showed that GLP-1 based therapies outperformed SGLT2is alone in lipid regulation[34].

Apart from renal parameters, the combination therapy outperformed monotherapy in glucose control. This advantage likely stems from the complementary mechanisms of the two drugs: Canagliflozin reduced blood glucose by promoting urinary glucose excretion independently of insulin, and semaglutide enhanced glucose-dependent insulin secretion, inhibited glucagon, and delayed gastric emptying to help control blood glucose[35]. Notably, improvements in HOMA-β and reductions in HOMA-IR indicate that this combination not only effectively lowers blood glucose but also improves underlying β-cell dysfunction and insulin resistance. These findings align with previous reports showing that GLP-1 RAs restored β-cell glucose sensitivity and SGLT2is alleviated glucotoxicity[36]. Additionally, a decrease in serum aspartate levels further confirms the metabolic advantages of combination therapy. Aspartate is related to hepatic glucose production and insulin resistance, and its reduction correlates with improved insulin sensitivity, making it a novel biomarker for metabolic health in diabetes interventions[37].

Oxidative stress and inflammation reduction are particularly compelling aspects of this combination therapy. A decrease in MDA (a marker of lipid peroxidation) and an increase in SOD activity were observed, indicating reduced oxidative stress and enhanced antioxidant capacity. Oxidative stress plays a key role in the development of diabetic complications, making its mitigation an important therapeutic target[38]. The findings of the present study suggest that CSCT may offer protection against oxidative damage, potentially lowering the risk of long-term complications. Additionally, reductions in TNF-α and CRP levels reflect systemic inflammatory decrease. Inflammation is increasingly recognized as a critical factor in insulin resistance and cardiovascular disease development[39]. Therefore, therapies, such as CSCT, which modulate inflammatory responses, not only control blood glucose but also provide broader health benefits. These effects likely arise from different yet converging pathways: Semaglutide directly inhibits NADPH oxidase activity and NF-κB signaling via the cAMP/PKA pathway, reducing pro-inflammatory cytokine production. Canagliflozin decreases proximal tubular sodium-glucose load, alleviating mitochondrial respiratory chain overload and ROS generation, thereby mitigating oxidative stress at its source[40,41]. Previous studies indicated that each drug alone can lower renal cortical oxidative stress markers, but the data of the present study further show that their combination produces additional inhibition of the inflammatory cascade[42]. Mechanistically, this is significant because persistent inflammation and oxidative damage are fundamental drivers of glomerulosclerosis and tubulointerstitial fibrosis in DN.

The safety comparison between the two groups provided clinical reassurance. Although semaglutide is known to potentially cause GI side effects, this combination did not increase the incidence of adverse events, likely due to the gradual dose escalation protocol[43]. No episodes of severe hypoglycemia occurred, supporting that semaglutide’s effects depend on blood glucose levels. Additionally, the rate of urinary tract infections aligned with the known safety profiles of both drugs, further reinforcing this finding. The results indicate that the combination therapy has a favorable risk–benefit ratio, making it more reasonable and reliable for clinical application.

Despite encouraging results, this study has several limitations to consider. First, although the study ensured group comparability through strict inclusion and exclusion criteria and baseline comparisons, it cannot completely avoid selection bias and the influence of unmeasured confounding factors. For instance, potential factors, such as patients’ dietary habits, daily exercise levels, and other unrecorded auxiliary medications, could affect renal function and metabolic indicators, leading to residual confounding. Second, as a single-center study, its generalizability to broader populations is limited. Patient demographics, comorbidities, and treatment adherence vary across different settings, potentially affecting observed outcomes. Additionally, the relatively short follow-up period limited the ability to assess long-term clinical outcomes. More time is needed to confirm the sustained efficacy and safety of CSCT. Larger sample sizes and longer observation periods could provide more reliable safety data. This study has two limitations regarding renal function assessment: First, the use of ACR instead of 24 hours urine protein quantification, while convenient, does not reflect the absolute amount of daily protein excretion. Second, the lack of significant changes in GFR over 6 months may be related to the initial hemodynamic effects of SGLT2is and the insufficient observation period to capture the long-term benefits of slowing eGFR decline.

Future prospective, multicenter randomized trials should address these limitations. Such studies could incorporate renal histology or advanced imaging biomarkers to better understand the structural changes underlying observed functional improvements. Investigating the molecular mechanisms of CSCT can offer insights into its mode of action and pave the way for targeted therapies for specific aspects of type 2 diabetes pathology. Exploring potential synergies between CSCT and existing treatments could enhance therapeutic options for patients with complex metabolic needs. Understanding the long-term effect of CSCT on microvascular complications like nephropathy and neuropathy is crucial.

CONCLUSION

For patients with type 2 DN, especially those whose proteinuria or metabolic indicators have not been adequately controlled with SGLT2i monotherapy, CSCT represents an effective and safe intensified treatment regimen. This study suggests that CSCT may be associated with improvements in metabolic and renal parameters, particularly with lower levels of urinary protein. These findings provide supportive evidence for the potential benefits of CSCT in potentially delaying disease progression and improving long-term patient outcomes.