SODIUM GLUCOSE COTRANSPORTER TYPE 2 INHIBITORS IN THE CONTROL OF ALBUMINURIA IN PATIENTS WITH CHRONIC KIDNEY DISEASE, STUDY IN A MEXICAN CENTER

Introduction:

Chronic kidney disease (CKD) has been described as a disease that stands for a serious public health problem in Mexico and worldwide. Its cause is multifactorial and is strongly associated with the most prevalent chronic diseases in our population, such as diabetes and hypertension. Its impact on public health is reflected in the high demand for human, economic, and infrastructure resources that its treatment requires. It is also the second leading cause of years of life lost in Latin America.

Current guidelines have defined chronic kidney disease as a heterogeneous disorder characterized by abnormalities in kidney function or structure, with implications for health. Functional abnormalities are related to decreased glomerular filtration rate (GFR), while structural abnormalities are inferred by markers of kidney damage that include increased albuminuria and abnormalities in urinary sediment and imaging.

Kidney disease is classified according to the cause and severity of GFR and albuminuria, as this classification is crucial for prognosis and treatment. GFR is an index of kidney function, and a decrease reflects a reduction in the number of nephrons or in the GFR of each individual nephron. Albuminuria, on the other hand, serves as a marker of kidney damage, showing dysfunction in the permeability of the glomerular barrier to macromolecules. According to Obrador et al., 2016, chronic kidney disease can be defined as a decrease in GFR <60 ml/min/1.73 m² for more than 3 months or an elevated albuminuria-creatinuria ratio during the same period.

Risk factors such as age, diabetes, systemic arterial hypertension, and obesity promote albuminuria through various mechanisms. Albuminuria, in turn, can lead to complications such as hypoalbuminemia, hyperlipidemia, deep vein thrombosis, pulmonary embolism, and deterioration of GFR. The main complications related to decreased GFR and increased albuminuria include cardiovascular diseases, fluid overload, anemia, malnutrition, infections, cognitive impairment, and frailty. 

Sodium-glucose cotransporter type 2 (SGLT-2) inhibitors decrease glucose concentrations in an insulin-independent manner by inhibiting its reabsorption in the renal proximal tubule.

The Empagliflozin Cardiovascular Outcomes and Mortality in Type 2 Diabetes (EMPAREG OUTCOME) and Canagliflozin Cardiovascular Assessment Study (CANVAS) studies showed a significant reduction in cardiovascular events with the use of empagliflozin and canagliflozin, respectively, in patients with type 2 diabetes. In addition, these studies proved that SGLT2 inhibitors have significant renoprotective effects, although the exact mechanisms have not yet been fully elucidated.

The effect of dapagliflozin on albuminuria in the DECLARE-TIMI 58 (Dapagliflozin Effect on Cardiovascular Events) study included 17,160 patients with type 2 diabetes, with measurements of urinary albumin-creatinine ratio at baseline, 6 and 12 months, and later every year. Dividing albuminuria into <15, 15-30, 30-300 and >300mg/g, a favorable effect was proved, suggesting that SGLT-2i can be used as part of the primary prevention of diabetic kidney disease.

Nagasu et al., 2021, based on the Japan Chronic Kidney Disease Database, conducted a study in which they divided patients who were receiving or not receiving SGLT-2i treatment. In a total of 1,033 patients, a smaller decrease in the glomerular filtration rate was seen in those treated with SGLT-2i.

In Mexico, a consensus was reached for the use of SGLT-2i derived from numerous studies worldwide, where their beneficial effect on the vasculature and the progression of diabetic nephropathy was demonstrated, improving the quality of life of these patients.

Methods:

A quasi-experimental, homodemic, longitudinal, and prospective study was conducted, including patients between 18-70 years old with GFR between 20-65 ml/min/1.73m² who presented any degree of proteinuria in a general urine test, determined by any method, and who attended an outpatient clinic. Two groups were followed: the first was administered standard antiproteinuric treatment, and the second was added dapagliflozin 10mg every 24 hours.

All patients signed informed consent, following international research study standards and the Official Mexican Standard 004-SSA3-2012 on electronic health record information systems.

Follow-up was conducted for 13 months, during which laboratory results were recorded every 3-4 months. At the end, the obtained data were described for numerical variables as mean and standard deviation, and for qualitative variables as frequency and percentage. To assess the relationship between variables, the OR of prevalences was calculated, and Chi-square was applied for independent groups with a p < 0.05. Survival for the groups was determined using the Kaplan-Meier and Nelson-Aalen survival curve estimators. Additionally, a Cox analysis was performed to find factors related to the events of interest.

Results:

The information obtained in this study was analyzed with the statistical program R (R Core Team, 2023). Distinctive characteristics and variables of the sample were compared, Kaplan-Meier survival curves were constructed, and different prediction models for the events of interest were analyzed with Cox analysis.

During the period between January 2023 and January 2024, a total of 63 patients were studied, of which 58 were included: 38 in the exposed group (treated with dapagliflozin) and 20 in the control group (standard antiproteinuric treatment). Five patients were excluded: one did not sign the informed consent, and the other four did not attend later consultations with albumin or protein determination in urine.

The mean age in the first group was 65 years ± 11.383, while in the second group it was 63.39 years ± 12.2. The rest of the demographic characteristics are described in Table 1.

TABLE 1. Baseline characteristics of the studied population

* p <0.05

Likewise, at the end of the study, the dapagliflozin group had an albuminuria of 666.54 mg/g ± 745; while the control group had 1350.19 mg/g ± 967.84 (p=0.14). The creatinine of the first group was 1.89 mg/dl ± 0.85, and for the second 2.83 mg/dl ± 1.339 (p=0.05).

During follow-up, the main goal was to prevent the deterioration of chronic kidney disease and, therefore, the need for renal replacement therapy. However, during the study, two patients in the dapagliflozin group needed renal replacement therapy (p=0.425).

Table 2. Behavior of the variables at the end of the study.

* p <0.05

To decide the therapeutic utility of sodium-glucose cotransporter type 2 inhibitors (iSGLT-2) controlling proteinuria in patients with chronic kidney disease, different variables such as hemoglobin, sodium, glycated hemoglobin, and serum albumin, among others, were compared between the groups. The comparison allowed deciding the existence or not of any relationship between the variables and the patient's exposure to SGLT-2 inhibitors. Our findings suggest that the patient's exposure to iSGLT-2 is not statistically significant in the presence of the different study variables. Table 3 shows the differences between the variables in the first measurement. Likewise, a comparison was made for the following measurements, where no significance was obtained either.

Table 3. Comparison of variables in the first measurement.

* p <0.05

The follow-up time was 13 months for both groups, to verify the effect of iSGLT-2 on albuminuria. Two events were defined as primary outcomes: an increase of 30% or more from baseline albuminuria, called “Progression,” and a decrease of 30% or more from baseline albuminuria, called “Improvement.” As seen in Table 4, the “Improvement group” was not statistically significant, however, in the “Progression group”, a decrease in risk was seen in 89%, that is, 11 out of 100 patients without dapagliflozin would increase albuminuria within 13 months.

Table 4. Primary results of the study, in terms of “Improvement” (>30% reduction in initial albuminuria) or “Progression” (>30% increase in initial albuminuria).

* p <0.05

According to the definition of the main events of the study, survival was determined by the Kaplan-Meier survival curve estimator for progression and the Nelson-Aalen survival estimator for improvement (See figures 1 to 10).

The survival curves of both events (improvement and progression) were compared for each of the variables: gender, diabetes, systemic arterial hypertension and the presence of some type of heart disease. In the case of the Improvement event, we did not find statistically significant evidence suggesting the existence of a difference between the groups. In the case of the “Progression” event, less progression was observed in the dapagliflozin group compared to the control group, when considering the variables: gender, diabetes, systemic arterial hypertension, some type of heart disease (p <0.001), see figures 2, 4, 6, 8 and 10.

Figure 1. Monitoring of “Improvement” over months (p=0.7)

Figure 2. Monitoring of “Progression” over months (p<0.001)

Figure 3. Improvement and gender (p=0.97)

Figure 4. Progression and gender (p<0.01)

Figure 5. Improvement and presence or absence of diabetes (p=0.68)

Figure 6. Progression and presence or absence of diabetes (p<0.01)

Figure 7. Improvement and presence or absence of systemic arterial hypertension (p=0.96)

Figure 8. Progression and presence or absence of systemic arterial hypertension (p<0.01)

Figure 9. Improvement and presence or absence of heart disease (p=0.64)

Figure 10. Progression and presence or absence of heart disease (p<0.01)

In addition to the nonparametric analysis of Kaplan-Meier and Nelson-Aalen, respectively, a Cox analysis was performed to identify those factors related to the events of interest.

In the case of time elapsed until the patient's improvement, the variables that had the greatest impact on the occurrence of the event of interest were:

• Having heart disease (HR 0.36, risk reduction of 64%, p=0.12)

• Hypothyroidism (HR 3.302, risk increase of 230.2%, p=0.02)

• Initial hemoglobin (HR 1.491, the higher the hemoglobin, the risk increases by 49.1%, p=0.006)

Obtaining a concordance of 71.7% for the prediction of the development of the event at 13 months (p=0.004).

For the progression event, the variables with the greatest relevance were the following:

• Treatment (HR 0.014, risk reduction 98.6%, p<0.001)

• Female gender (HR 0.223, risk reduction 77.7%, p=0.06)

• Age (HR 1.068, the older the age the risk increases 6.8%, p=0.08)

• Systemic arterial hypertension (HR 0.013, risk reduction 98.7%, p<0.001)

• Initial hemoglobin (HR 0.682, the higher the hemoglobin the risk decreases 31.8%, p=0.08)

• Initial serum creatinine (HR 0.061, the higher the creatinine the risk decreases 93.9%, p=0.01)

• Initial serum albumin (HR 4.351, the higher the albumin the risk increases by 335.1%, p=0.0593)

With the model, a concordance of 91.4% is obtained for the prediction of the event at 13 months (p<0.001).

Conclusions:

As described in the literature, the assessment of chronic kidney disease must consider the different risk factors, such as albuminuria, especially in patients with chronic degenerative diseases such as diabetes, systemic arterial hypertension or some type of heart disease.

Studies have shown that metabolic control of patients plays a crucial role in preventing the progression of kidney disease. Therefore, different drugs have been developed that, during the first phases of study, have shown favorable "collateral" effects for patients.

In studies such as that of Chu et al. (2019) and in post hoc studies such as that of Mozenson et al. (2021), the effect of sodium-glucose cotransporter type 2 inhibitors (SGLT-2 inhibitors) on the excretion of proteins, specifically albumin, in urine has been investigated.

As proved in the study by Hiddo J. L. Heerspink et al. (2020), where patients were followed for 2.3 years, a slight decrease in the glomerular filtration rate was observed in the first months, with subsequent stabilization of the same. In our work, the glomerular filtration rate increased, but this increase was not statistically significant, with respect to the initial renal function. However, when comparing the 2 groups, the group with dapagliflozin had a better eGFR at the end of the study.

Regarding albuminuria, there are works such as those by Jongs et al. (2021) and Heerspink et al. (2021) where they performed an analysis on the effect of dapagliflozin on this, resulting in patients with this medication decreased excretion, likewise, patients with severe albuminuria, the medication prevented an increase in this area. In our study, as we did not have such a large sample, we decided to divide the patients into patients with improvement (30% decrease in initial albuminuria) or progression (30% increase in albuminuria compared to the initial); observing that patients with dapagliflozin, although it did not reduce albuminuria excretion, did prevent the increase during follow-up, this being statistically significant.

With the above information, a Cox analysis was performed to determine the variables with the greatest statistical weight, developing a predictive model that can determine the risk that patients have of developing any of our events of interest. Both models tell us about the patient's initial conditions, so it should be useful to quickly determine if our patient would benefit from being treated with sodium-glucose cotransporter type 2 inhibitors.

About the “Improvement” model (decrease in initial albuminuria >30%), being a patient with a heart disease reduces the probability of improvement by 64%, hypothyroidism increases the risk by 230%, in the biochemical part; for each g/dl that the hemoglobin increases, the probability increases by 49.1%. Although the variable being a patient with a heart disease does not reach statistical significance, at the time of the model concordance, it has a predictive capacity of up to 71.7%.

On the other hand, in the Cox analysis for “Progression” (increase in initial albuminuria >30%), being treated with iSGLT-2 reduces the event by up to 98.6%; The female gender has a 77% lower risk compared to men; with regard to age, for each additional year of age, the risk increases by 6.8%; having systemic arterial hypertension is associated with a 98.7% reduction in risk. Thus, in the biochemical part of the patient's assessment, having higher hemoglobin, creatinine and serum albumin have a protective effect of 31.8%, 93.9% and an increased risk of 335.1%, respectively. Overall, the previous model has a concordance of 91.4%, which could predict which patients will have a faster progression of albuminuria, and indirectly a faster deterioration of renal function.

These models could help in screening the population at risk. However, there are variables that probably cause us contradiction when talking about the events of interest, as in the case of serum creatinine. The higher the serum creatinine, the Cox analysis determined a lower risk of progression of albuminuria. This is probably due to various mechanisms, whether renal damage or related to the main pathology. Likewise, the protective effect of high blood pressure on the progression of albuminuria could be due to the concomitant use of other drugs that indirectly reduce albuminuria. However, these arguments are beyond the scope of this work, so further studies will be needed that follow a larger number of patients, as well as the integration of variables that can help elucidate what has been discussed above.

Although the use of SGLT-2 inhibitors does not significantly reduce albuminuria in patients with chronic kidney disease, it can stop the progression of this disease, that is, prevent the increase in albumin excretion through urine. This would translate into a better quality of life for patients and could avoid the need for renal replacement therapy, which would be reflected in a reduction in health care costs for this type of patient.

It is important to highlight the need to use this type of treatment from the primary care level, since, in many cases, patients arrive at the level of care where we work with a significant deterioration of kidney function. It would be ideal if primary care physicians could help patients by using the probabilistic model discussed above.

I have no potential conflict of interest to disclose.

I did not use generative AI and AI-assisted technologies in the writing process.