Background

Individuals with type 2 diabetes mellitus are at increased risk of cardio-renal complications. Within the altered hormonal milieu among those with type 2 diabetes are changes in the insulin growth factor axis; recent studies have examined the role of insulin growth factor-1 (IGF-1) in the risk for complications from type 2 diabetes [1]. IGF-1 is a 70-amino acid peptide, structurally homologous to pro-insulin, synthesized mainly in the liver upon growth hormone (GH) stimulation [2]. Molecular studies have revealed that IGF-1 promotes cellular growth, inhibits cell apoptosis, stimulates glucose uptake by muscle and heart cells, and enhances glycogen, lipid, and protein synthesis [3, 4]. These metabolic effects are regulated by a complex interaction between GH, insulin, IGF-1, and 6 soluble high-affinity IGF-binding proteins [5].

Abnormal concentrations of IGF-1 are linked with obesity [6], metabolic syndrome [7], type 2 diabetes [8], atherosclerosis [9], heart failure (HF) [10, 11], and diabetic kidney disease (DKD) [12]. Furthermore, recent data have implicated insulin-like growth factor binding proteins (IGFBP) in cardio-renal risk in those with and without type 2 diabetes [13]. The most abundant peptide in the IGFBP family is IGFBP-3 [14]. It has a high affinity for IGF-1 and alters the interaction between IGF-1 and IGF-1 receptor. The ratio of IGF-1 to IGFBP-3 is considered as a parameter of IGF-1 bioactivity [15].

Further research is needed to determine the significance of IGF-1, IGFBP-3, and the ratio between the two in assessing cardio-renal risk, as previous studies have yielded conflicting results regarding their association with adverse clinical outcomes [16]. Furthermore, an understanding of how therapies with benefits on cardio-renal risk in diabetes intersect with concentrations of these peptides is unknown. Accordingly, in the present analysis, using data from CREDENCE trial (Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation; ClinicalTrials.gov Identifier: NCT02065791), we sought to investigate the association of IGF-1, IGFBP-3 and IGF-1/IGFBP-3 ratio with incident cardio-renal outcomes and evaluated effect of canagliflozin on their concentrations.

Methods

Study design and patient population

The trial design, baseline patient characteristics, and the main study results from the CREDENCE trial have been published previously [17, 18]. Briefly, CREDENCE was a placebo-controlled trial of canagliflozin 100 mg versus placebo in 4401 persons with type 2 diabetes and DKD. Study participants had a minimum glycated hemoglobin between 6.5% and 12.0% and were required to have an estimated glomerular filtration rate (eGFR) between 30 and 90 mL/min/1.73 m2 and urine albumin creatinine ratio (UACR) > 300 to 5000 mg/g. All subjects had to be treated with angiotensin converting enzyme inhibitor(ACEi) or angiotensin receptor blocker (ARB) at randomization.

In this analysis, only those study participants with available plasma for analysis of IGF-1 and IGFBP-3 at baseline were included (N = 2627). Plasma samples were collected at baseline, 1 year, and 3 years, and stored at – 80 ℃ degrees centigrade. IGF-1 was measured using an automated electrochemiluminescence immunoassay (Roche Diagnostics, Mannheim, Germany). This method is standardized against the WHO International Standard 02/254. Detection limit was 7 ng/mL and coefficient of variation for repeatability was ≤ 3.5%

There were 4 main goals of this analysis. First, we determined the distribution of biomarkers at baseline. Second, we evaluated canagliflozin’s effect on biomarker concentrations from baseline to 1 year and baseline to 3 years. Third, we evaluated the association between biomarker concentrations at baseline (or their change from baseline to Year 1) and cardiovascular (CV) and kidney outcomes. Clinical endpoints examined included the primary composite endpoint of CREDENCE (a composite of end-stage kidney disease, doubling of the serum creatinine level, or renal/CV death), the renal composite endpoint (a composite of end-stage kidney disease, doubling of the serum creatinine level, or renal death), as well as the composite of heart failure (HF) hospitalization or CV death, HF hospitalization, all-cause death, and CV death. Fourth, we evaluated the effect of canagliflozin on risk as a function of concentrations of IGF-1, IGFBP-3, or their ratio.

Statistical analysis

Biomarkers were log transformed because of their skewed distribution. Median (interquartile) and count (frequency) were used to present continuous and categorical variables. Kruskal–Wallis, ANOVA, and chi-square tests were used to compare the baseline characteristics of study population across IGF-1 quartiles, IGFBP-3, and IGF-1/IGFBP-3 ratio as appropriate. To evaluate the effect of canagliflozin on biomarker concentrations, comparisons of geometric mean (95% CI) concentrations were performed in Years 1 and 3. For change from baseline to Year 1, a base linear model was constructed for each log-transformed biomarker at Year 1 by selecting important baseline covariates in patients randomized to placebo in the main study based on Bayesian Information Criterion. The candidate covariates at baseline included continuous variables: age, eGFR, body mass index, systolic blood pressure, hemoglobin A1c, duration of diabetes mellitus, UACR, log transformed NTproBNP and categorical variables: history of HF, and history of diuretic treatment. Cox proportional hazard regression was implemented to assess the association between biomarker concentrations with clinical outcomes, including treatment and treatment-by-biomarker interaction in the models with selected covariates. To do so, log-transformed concentrations of IGF-1 and IGFBP-3 or their ratio were evaluated with hazard ratio (HR) and 95% CI expressed per 1-unit change in each measure. Additionally, dichotomous cutoffs for IGF-1 based on age were also applied 19.

All hypotheses were 2-sided, with a p-value < 0.05 considered statistically significant. All statistical analyses were performed using the R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria. URL: https://www.R-project.org/).

Results

Additional file 1: Figure S1 details the study flow for the present analysis. The baseline study sample consisted of 2627 individuals with diabetic kidney disease.

Table 1 details the baseline characteristics of the study population across IGF-1 quartile groups. Patients with the highest quartile were younger, more likely to be male and Black, had a lower prevalence of coronary artery disease, lower eGFR level, lower systolic blood pressure, higher diastolic blood pressure, and lower diabetes mellitus duration compared to other quartiles. Baseline characteristics according to IGFBP-3 and IGF-1/IGFBP-3 ratio quartiles are detailed in Additional file 1: Tables S1, S2.

Table 1 Baseline characteristics of study population stratified by IGF-1 quartile
Full size table

Table 2 shows unadjusted geometric mean (95% CI) concentrations of biomarkers at baseline, Year 1, and Year 3. IGF-1, IGFBP-3 levels, and the IGF-1/IGFBP-3 ratio remained relatively constant during 3 years of follow-up among both canagliflozin and placebo groups. To explore treatment-related effect on biomarker concentrations, geometric mean ratio of Year 1/baseline was examined in adjusted analyses. In these adjusted models, treatment with canagliflozin did not significantly change concentrations of IGF-1 and IGFBP-3 over time (Additional file 1: Table S3).

Table 2 Unadjusted geometric mean (95% CI) concentrations of biomarkers at baseline, Year 1, and Year 3
Full size table

IGF-1 and IGFBP-3 were examined as continuous and dichotomous variables. To do so, elevated IGF-1 level is defined based on age-specific cutoffs, as the level tends to decrease significantly with age [19] (Additional file 1: Table S4). Patients with elevated IGF-1 were younger, were more likely to be male and Black, had lower eGFR at baseline, and had a longer duration of type 2 diabetes compared to patients with lower concentrations of the biomarker (Additional file 1: Table S5). Concentrations of IGF-1 and IGFBP-3 across chronic kidney disease (CKD) stages are detailed in Additional file 1: Table S6. Patients with stage 4 CKD had higher concentration of IGF-1 compared to other stages. IGFBP-3 concentrations were similar across CKD stages.

Additional file 1: Figure S2 depicts the association of continuous log IGF-1, IGFBP-3, and IGF-1/IGFBP-3 ratio with primary composite outcome using restricted cubic spline modeling. Higher IGF-1 levels and IGF-1/IGFBP-3 ratio were associated with a higher risk of primary composite outcome.

Figure 1 demonstrates the association of continuous and dichotomous IGF-1, IGFBP-3, and IGF-1/IGFBP-3 with clinical outcomes. In the multivariable-adjusted model, 1-unit increase in log IGF-1 and IGFBP-3 was not associated with clinical outcomes (p-value > 0.1). However, elevated IGF-1 according to the age-specific cutoff was associated with the primary composite outcome (HR: 1.52, 95% CI 1.09–2.13, P: 0.01), renal composite outcome (HR: 1.65, 95% CI 1.14–2.41, P: 0.01) and all-cause mortality (HR: 1.52, 95%CI 1.00–2.32, P; 0.05). Also, an increase in the IGF-1/IGFBP-3 ratio was associated with primary, renal, CV death, and all-cause mortality outcomes (p values < 0.05). No treatment-by-biomarker interaction was present; thus, the effect of canagliflozin across quartiles of IGF-1, IGFBP-3, or their ratio was largely consistent relative to study outcomes (Fig. 2).

Fig. 1
figure 1

Association of IGF-1 and IGFBP-3 with clinical outcomes. In a multivariable-adjusted model, a 1-unit increase in log IGF1/IGFBP-3 ratio and elevated IGF-1 levels according to age-adjusted cutoff increased the risk of primary composite outcome, renal composite outcome, all-cause death, and CV death. *Per 1-unit increment. HR hazard ratio, CI confidence interval, IGF-1 insulin-like growth factor-1, IGFBP insulin growth factor binding protein. HF heart failure, CV cardiovascular

Full size image
Fig. 2
figure 2

Efficacy of canagliflozin on lowering risk of primary composite outcome across IGF-1, IGFBP-3, and IGF-1/IGFBP-3 ratio quartiles. Effect of canagliflozin across quartiles of IGF-1, IGFBP-3, or their ratio was largely consistent relative to study outcomes. IGF insulin-like growth factor, IGFBP insulin growth factor binding protein, HR hazard ratio, CI confidence interval

Full size image

Discussion

In this trial of patients with type 2 diabetes and CKD who were randomized to receive canagliflozin or placebo, we showed that baseline IGF-1 levels and IGF-1/IGFBP-3 ratio (but not IGFBP-3 concentrations) were associated with cardio-renal outcomes. Higher IGF-1 levels (for a given age) were associated with a greater risk of developing renal and all-cause mortality events. 3 years of therapy with canagliflozin did not significantly change IGF-1 and IGFBP-3 concentrations. Lastly, the benefits of canagliflozin to reduce cardio-renal events in this high-risk population were consistent across IGF-1 and IGFBP-3 strata. These findings provide evidence regarding the role of the IGF axis in risk for cardio-renal disease.

IGF-1 is an anabolic hormone that regulates cellular proliferation, apoptosis, and several metabolic pathways in the human body. Nearly all 98% of IGF-1 is bound to 1 of 6 IGFBPs in circulation. Owing to its longer half-life, IGFBP-3 is the most abundant member of the IGFBP family and accounts for 80% of all IGFBPs [20]; it binds to IGF-1 with high affinity and blocks IGF-1 access to the IGF-1 receptor. IGF-1 plays an intermediate role in glucose metabolism. Unlike GH, IGF-1 has a hypoglycemic effect by suppressing hepatic gluconeogenesis and stimulating peripheral glucose uptake [3]. IGF-1 levels increase as insulin resistance develops; with worsening of insulin resistance, the IGF-1 concentration reaches a plateau level, and subsequently, when glucose levels reach concentrations typical of type 2 diabetes, IGF-1 levels tend to decline [8, 21, 22]. Although the IGF-1/IGFBP-3 ratio is proposed as an index of IGF-1 bioavailability [23], it is essential to recognize other IGFBPs as they also modestly affect IGF actions [24].

While enhancing insulin sensitivity, growth-promoting properties of IGF-1 are proposed to play a role in developing complications of diabetes [1]. Emerging studies have shown the GH/IGF-1 axis alteration among patients with DKD [25] with IGF-1 production are thought to stimulate proliferation of renal mesenchymal cells and vascular proliferative changes [26]. Animal model studies have shown enhanced expression of renal IGF-1 receptor as a factor contributing to renal hypertrophy—a hallmark sign of DKD [27]. In addition, studies suggest that IGF-1 may have anti-natriuretic properties [4]. This is believed to occur through two mechanisms: firstly, it may directly affect the absorption of sodium by regulating the epithelial sodium channel in the distal nephron [28]. Secondly, IGF-1 overexpression may indirectly enhance the renin–angiotensin–aldosterone system. (RAAS) [29].

Despite the proposed mechanistic role of IGF-1 in DKD, results of clinical studies investigating the association between IGF-1 levels and kidney disease are inconclusive. In the NHANES study (National Health and Nutrition Examination Survey), Teppala and colleagues [30] showed that elevated IGF-1 levels were positively associated with CKD independent of conventional CKD risk factors. In line with this finding, Dittman and colleagues [31] demonstrated that increased serum IGF-1 levels were associated with decreased eGFR level. Contrary to these results, several studies have shown an association between low IGF-1 levels and adverse renal outcomes [32, 33]. In this study, we found a negative relationship between IGF-1 level and kidney function. One may assume that reduced renal excretion may play role in elevation IGF-1 concentration. However, this is possibly overly simplistic as elevated IGF activity may be linked with more severe kidney disease. Indeed, previous studies have established a connection between urinary excretion of insulin like growth factors and renal disease activity [34]; as a result, increased levels of IGF-1 in severe CKD patients may not be directly linked to renal excretion. Given biological rationale but mixed clinical studies, a rationale existed to examine the role of IGF-1 and IGFBP-3 in CREDENCE study participants.

Findings of our study corroborate studies indicating a detrimental association between elevated IGF-1 (when above age-specific cutoffs) as well as higher IGF-1/IGFBP-3 ratio on kidney function. As IGF-1 levels decrease by age, the findings from this study highlight the importance of considering IGF-1 age-specific cutoffs as well as incorporating the balance of IGF-1 bioactivity (reflected in the IGF-1/IGFBP-3 ratio) when studying IGF biology and cardio-renal risk. On the other hand, IGFBP-3 was not correlated with baseline kidney function and failed to predict any clinical events. IGFBP-3 has several IGF-1–dependent and IGF-1–independent functions [35]. The role of IGFBP-3 in type 2 diabetes and its complications requires further consideration.

The exact mechanism of the cardio-renal benefit of canagliflozin is still undetermined [36]. Beside lowering blood glucose, several mechanisms, including improved energy metabolism, vascular function, hemodynamic alterations, decreased oxidative stress, and reduction in inflammation, are proposed to mediate CV benefit of SGLT2 inhibitors. In this study, although higher IGF-1 levels were associated with renal outcomes, canagliflozin failed to lower IGF-1 levels and beneficial effects of canagliflozin in lowering adverse outcomes were observed across IGF-1 and IGFBP-3 levels. These findings imply that the beneficial effects of canagliflozin is likely independent of any changes in the IGF-1 axis. Lastly, it is important to highlight that a monoclonal antibody against IGF-1 receptor called teprotumumab has been developed for treatment of proptosis [37]. While hyperglycemia is a significant adverse effect, the existence of these agents offers the potential for their use in addressing DKD that involve excessive IGF activity.

This study had several limitations. First, biomarker data were unavailable for all participants; however, those in this post hoc analysis were similar to the main study. Second, more than 70% of study participants were White. A diverse research population can increase generalizability of our findings. Future studies need to implement the National Institutes of Health recommendations to conduct research studies with diverse ethnic backgrounds. Lastly, patients were followed up for 3 years; a longer duration may be required to investigate the association between IGF-1 and incident HF.

Conclusion

In conclusion, this study provides evidence that elevated IGF-1 levels or the ratio of IGF-1/IGFBP-3 is associated with a higher risk of kidney disease progression and all-cause mortality. Three years of therapy with canagliflozin failed to impact IGF-1 or IGFBP-3 levels. Nonetheless, the benefit of canagliflozin to reduce cardio-renal endpoints was preserved across strata of both biomarkers. These results affirm a role of IGF-1 or its activity in the progression of cardio-renal disease among individuals with type 2 diabetes and DKD.