Perspectives on Chronic Kidney Disease With Type 2 Diabetes and Risk Management: Practical Viewpoints and a Paradigm Shift Using a Pillar Approach

CKD is defined as an abnormality in kidney structure and/or function based on the following criteria: eGFR <60 mL/min/1.73 m² and/or albuminuria (urinary albumin excretion rate ≥30 mg per 24 hours or UACR ≥30 mg/g) for >3 months (8).

Both UACR and eGFR are used for staging of CKD. As noted previously, eGFR is a marker of kidney function, and UACR is a marker of kidney damage. If the eGFR decreases from a patient’s baseline value (and this is replicated on retesting), this suggests worsening of kidney function; if the UACR increases from the baseline value (and this is replicated on retesting), this suggests worsening kidney damage. Of note, moderately increased urine albumin levels—previously called microalbuminuria and defined as UACR 30–300 mg/g—typically appears before a significant decline in eGFR in patients developing CKD (14–16). The KDIGO heat map (Figure 1) can be used to establish whether decreases in eGFR and/or increases in UACR are clinically significant and can be used to support a CKD diagnosis and to guide treatment (4,8–10,12). CKD is commonly attributed to diabetes and/or hypertension, but other possible causes include glomerulonephritis, polycystic kidney disease, systemic infection, autoimmune disease, drug toxicity, vascular diseases, and environmental exposures (8).

Diabetic kidney disease (DKD) is defined with similar criteria as CKD but occurs in the setting of diabetes and in the absence of other causes of CKD (17). In most patients with diabetes, CKD should be attributable to diabetes if one of the following conditions applies: 1) UACR >300 mg/g, 2) UACR 30–300 mg/g in the presence of diabetic retinopathy (DR), or 3) UACR 30–300 mg/g in type 1 diabetes of at least 10 years’ duration (18). eGFR alone may be a less sensitive screening method for DKD because many patients with DKD may demonstrate normal eGFR in the early years after diagnosis (18). The term “CKD associated with type 2 diabetes” is now more commonly used than “DKD” and will be used hereafter in this article.

The term “diabetic nephropathy” may be used when referring to CKD associated with type 2 diabetes, but the terms are not fully interchangeable: “CKD associated with type 2 diabetes” refers to the structural and functional alterations associated with diabetes, whereas “diabetic nephropathy” refers to histological findings on biopsy (19). However, the 2022 KDIGO guidelines suggest that the term “diabetic nephropathy” is an “outdated term” with “no consensus definition” (12). Thus, to avoid possible confusion, we will not use this term in this review except when it is used in the reference cited.

Approximately one in three adults with type 2 diabetes may have CKD, and as many as 90% of adults with CKD and 40% of adults with severe CKD are unaware of their kidney disease (20). These statistics underscore the importance of informing patients with diabetes, as well as clinicians, about the significance of screening for CKD.

Abnormalities in kidney structure may precede reductions in kidney function, and both presentations may be associated with complications such as cardiovascular disease (CVD) and metabolic disease (8,21). The most widely used and generally accepted index of kidney function is the eGFR (8). An important test for kidney damage in CKD is the UACR test, where (as noted previously) an above-normal level of albuminuria is a marker of kidney damage (8). The UACR can be determined from a spot urine collection or via a 24-hour urine collection; generally, the spot UACR test is preferred because the 24-hour collections are more burdensome with minimal differences between the two techniques in prediction or accuracy (4). Therefore, albuminuria and eGFR independently influence prognosis of CKD associated with type 2 diabetes, as represented in the KDIGO heat map (Figure 1), where albuminuria and/or a low eGFR indicate an increasing risk of CKD progression, frequency of visits, and need for referral to a nephrologist (8–10).

The ADA 2023 Standards of Care recommends that patients with type 2 diabetes should be screened at least annually for albuminuria using the spot urine test for UACR and should have their eGFR assessed, and this should be done irrespective of treatment (evidence grade B) (4). Naturally, consideration should be made of factors other than kidney damage, such as infection or strenuous exercise, that may also elevate UACR (8,22). Patients with established CKD associated with type 2 diabetes should have their UACR and eGFR monitored one to four times per year depending on their CKD stage (evidence grade B) (4). referral criteria and impact of early versus late referral on clinical outcomes below and Table 2 provide the ADA’s criteria for specialist referral (4,8,11).

NICE provides a comprehensive list of recommendations on the diagnosis, assessment, and initial investigations for CKD (11,23), which contrasts to the ADA’s summary recommendations (with evidence grading) in relation to CKD associated with type 2 diabetes. Screening (or initial investigations) should be considered for patients with risk factors for CKD (including type 2 diabetes), with incidental findings suggestive of kidney disease (e.g., elevated serum creatinine and/or eGFR <60 mL/min/1.73 m²) and/or with possible clinical features of CKD (11,23). Albuminuria and/or eGFR should be followed up within 3 months to determine the CKD stage (11,23). Additionally, patients with CKD may have symptoms in late stages, such as uremic fetor, pallor, cachexia, cognitive impairment, tachypnea, dehydration, hypertension, peripheral edema, peripheral neuropathy, and/or foamy urine (11,23); such symptoms are rarely seen in early-stage CKD (stages 1 and 2). Clinicians should also check their patients’ nutritional status, BMI, blood pressure, A1C, and lipid profile to optimize CVD risk factors. NICE also provides detailed specialist referral criteria, which are described in referral criteria and impact of early versus late referral on clinical outcomes below and in Table 2 (4,8,11).

The 2012 and 2022 KDIGO guidelines (8,12) do not include formal recommendations regarding screening for CKD or how often testing for kidney disease markers should be done, but they do note that public health policies should include screening in high-risk populations such as those with diabetes (8). Furthermore, the KDIGO guidelines include a table that provides expanded criteria for the definition of CKD and can be used by nonnephrologist physicians and other health care professionals to assist in the detection of CKD. In addition to the definition of CKD described previously in this article, these criteria include urinary sediment abnormalities, renal tubular disorders, pathological and structural abnormalities (as markers of kidney damages), and history of kidney transplantation (8). Although not included as a formal recommendation, the KDIGO 2012 guidelines recognize the importance of early detection of CKD (8). Avoiding a delay in diagnosis or early intervention to prevent progression of CKD can confer morbidity and mortality benefit at a lower cost than kidney transplantation (8). KDIGO criteria for referral to a specialist are provided in referral criteria and impact of early versus late referral on clinical outcomes below and in Table 2 (4,8,11).

As noted previously, the ADA 2023, KDIGO 2022, and NICE clinical practice guidelines differ in some respects in their recommendations for screening and early diagnosis of CKD. Although treatment guidelines are not intended to define a standard of care (12), the implications to clinical practice of differences in recommendations may lead to confusion (in terms of which guideline to follow and when) and nonadherence (24,25). In the case of CKD with type 2 diabetes, the ADA and KDIGO are working together to harmonize their treatment guidelines (13). Regarding screening, the ADA/KDIGO consensus report notes that screening should occur annually from the point of diabetes diagnosis in people with type 2 diabetes, with persistent abnormalities defining CKD (13). This alignment elevates the importance of screening for CKD both at diagnosis of type 2 diabetes and every year thereafter.

The ADA 2023 Standards of Care recommends screening for DR via an initial dilated and comprehensive eye examination by an ophthalmologist or optometrist at the time of a type 2 diabetes diagnosis (Chapter 12; evidence grade B) (26). If there is no evidence of DR for ≥1 year of annual examinations and glycemia is well controlled, then examinations may be repeated every 1–2 years (evidence grade B) (26). However, retinopathy status should be reassessed when intensifying glycemic control. In the SUSTAIN-6 (Trial to Evaluate Cardiovascular and Other Long-term Outcomes with Semaglutide in Subjects with Type 2 Diabetes) study, the number of retinopathy complications was significantly higher with semaglutide, a glucagon-like peptide 1 (GLP-1) receptor agonist, than with placebo, and semaglutide profoundly reduced A1C versus placebo (27). However, it is important to note that the patients in the SUSTAIN-6 trial were at high cardiovascular risk, >82% had preexisting retinopathy, and >58% were taking insulin at baseline, so it is not clear whether the retinopathy complications seen with semaglutide are a direct drug effect or the result of reduced A1C. Indeed, the SUSTAIN-6 authors noted that application of such an association is unclear, but a direct effect of semaglutide still cannot be ruled out.

Although no formal recommendations on DR screening in relation to CKD are included in the ADA 2023 or KDIGO guidelines, DR has been shown to be a highly specific indicator for the diagnosis of diabetic nephropathy (28). Furthermore, several studies have suggested that DR severity in patients with type 2 diabetes can be used to predict CKD severity and/or progression risk at diagnosis (28–30). Thus, among other screenings, clinicians should evaluate DR severity in patients with type 2 diabetes at the time of diagnosis and monitor kidney function in patients with severe DR, as suggested (30).

The ADA (4), NICE (11), and KDIGO (8) have each provided detailed guidance regarding the timing or threshold for referral of patients to a nephrologist (Table 2). For example, the ADA recommends consultation with a nephrologist when there is a continuously rising UACR and/or a continuously declining eGFR, if there is uncertainty about the etiology of kidney disease, for difficult management problems, for CKD progression to stage 4 (eGFR <30 mL/min/1.73 m²), and/or when there is urinary sediment, nephrotic syndrome, or the absence of retinopathy in type 1 diabetes (Table 2) (4,8,11).

Early, appropriate referral to a nephrologist is associated with reduced mortality and hospitalizations and better dialysis preparation (31). A meta-analysis of 40 studies involving 63,887 patients with CKD found that almost one-third (n = 20,678) were referred late (defined as <1 to 6 months before starting dialysis) (31). Those referred early compared with those referred late had reduced mortality and hospitalizations, better uptake of peritoneal dialysis, and earlier placement of arteriovenous fistulae. Differences in mortality and hospitalizations were not explained by the prevalence of comorbid disease or serum phosphate levels.

Another study found that many patients with CKD are referred late (32). An analysis of electronic health records from 2017 to 2019 suggested that 54.6% of patients with an eGFR <30 mL/min/1.73 m² had not been referred to a nephrologist (32). The analysis also found that patients of younger age, with complex medical histories, and treated by primary care providers (PCPs) at an academic medical center were more likely to be referred (32). Furthermore, both a regional shortage of nephrologists (33) and a lack of PCP awareness of guideline recommendations for referral (34) may contribute to low rates of referral. Other barriers to (early) referral and overall effective comanagement of patients with CKD include a lack of effective collaboration tools for facilitating timely adequate information exchange between PCPs and specialist (nephrology) services, a lack of clear understanding of roles and responsibilities, and a need for greater access to specialist advice (35).

Most people with diabetes and CKD do not get a kidney biopsy. According to the ADA, referral for a kidney biopsy is recommended in patients with type 1 diabetes when additional or nondiabetic causes of kidney disease are suspected (4). The NICE guidelines provide a list of kidney disease markers and symptoms and a requirement of family history of kidney disease as indications for referral for a renal ultrasound; here, a nephrologist may consider that the patient needs a kidney biopsy (11). The KDIGO 2012 guidelines provide some general guidance concerning kidney biopsy referral criteria. Patients with a decline in eGFR without markers for kidney damage may be referred for a biopsy to look for evidence of parenchymal lesions (8). However, although a kidney biopsy is required to diagnose diabetic glomerulopathy definitively, in most cases careful screening of patients with diabetes can identify those with a high likelihood having CKD associated with type 2 diabetes without the need for a kidney biopsy (18). A biopsy should only be performed if it is essential to confirm a diagnosis and the benefits outweigh the risks (such as bleeding) and costs (Table 2) (4,8,11).

Although prevention of or reduction in the risk of end-stage kidney disease (ESKD) is an accepted clinically meaningful end point of CKD treatment, several surrogate end points have been developed to facilitate clinical trials (36). Levey et al. (37) proposed that a UACR reduction of 30% in 6 months or an eGFR slope reduction by 0.5–1.0 mL/min/1.73 m² over 2–3 years is a threshold reliably associated with significant treatment effects on kidney disease progression under certain conditions. (These numbers vary with sample size.) Using eGFR slope is also more appropriate and valid than changes of albuminuria, but it requires more attention to acute effects and a longer follow-up period (37). The ADA 2023 Standards of Care recommend that, in patients with CKD who have ≥300 mg/g urinary albumin, a reduction of ≥30% in mg/g urinary albumin is recommended to slow CKD progression (evidence grade B) (4).

Clinicians should be aware that, although these surrogate end points are intermediate outcomes (i.e., not final clinical outcomes of interest) that can be tested in a laboratory, a strong mathematical association between the surrogate and clinical end points does not always guarantee surrogacy in a clinical context (36). Surrogate markers can be influenced by the presence of acute drug effects (38), day-to-day fluctuation caused by analytical bias (39,40), beneficial effects of treatment in patients with fast progression of disease (proportional treatment effects) (41), and patients’ age (42).

In addition, criteria for CKD associated with type 2 diabetes that are based on albuminuria may not be applicable to all patients because some patients develop advanced disease without showing albuminuria, and others with albuminuria do not demonstrate pathological evidence of kidney damage (18). Among U.S. adults with diabetes from 1988 to 2014, although the overall prevalence of CKD associated with type 2 diabetes (defined as albuminuria ≥30 mg/g, reduced eGFR <60 mL/min/1.73 m², or both) did not change significantly, the prevalence of albuminuria declined, and the prevalence of reduced eGFR increased (43). The ADA 2023 guidelines also note the existence of frequent cases of reduced eGFR without albuminuria in patients with diabetes (4).

Furthermore, the drugs used to treat CKD can cause multiple effects beyond the target effect, and off-target effects can contribute to the ultimate effect on clinically meaningful kidney outcomes (36). For example, renin– angiotensin system (RAS) inhibitors reduce blood pressure and albuminuria, both of which contribute to renoprotection, but may also elevate blood potassium levels. Therefore, multiple risk parameters and risk scores (36) that integrate all known drug-induced effects are potentially more reliable than single markers to help clinicians make more appropriate treatment decisions. A scoring system that integrates multiple short-term drug effects (i.e., changes in systolic blood pressure, albuminuria, potassium, hemoglobin, cholesterol, and uric acid) was generated to predict the long-term effect on kidney and cardiovascular outcomes. These scores provided better prediction of the drug effect on hard kidney outcomes than single markers (36,44).

Lifestyle modifications and self-management are important elements of risk reduction for type 2 diabetes– associated complications (12,45). Lifestyle modifications include losing excess weight, consuming fewer simple sugars and saturated fats, adopting healthy eating habits, increasing physical activity (to at least 150 minutes/week), and smoking cessation (6,12,45). Additionally, drug treatments that help to optimize blood glucose levels (such as SGLT2 inhibitors or metformin) and blood pressure (RAS inhibitors) and/or for lipid management (statins) may be needed if lifestyle modifications alone are not sufficient to reduce risk (6,12,45,46). Aspirin may be appropriate for those at high risk of atherosclerotic CVD. Risk factor reassessment should be completed every 3–6 months (12).

The use of drug treatments to support risk mitigation in type 2 diabetes (and in CKD associated with type 2 diabetes) is important, and treatment guidelines provide information on the treatment thresholds that should be reached before drug initiation. However, what may not be as clear in the guidelines is what the recommended approach should be when a patient is experiencing adverse effects. Should the drug be discontinued permanently? Should it be discontinued and reintroduced at a lower dose? Should an alternative drug be used? None of the guidelines provide specific detailed guidance on this point, perhaps because of the wide heterogeneity of the type 2 diabetes population. Thus, clinicians should continue to use a holistic and patient-centered approach when supporting high-risk patients.

The ADA/KDIGO consensus report does include summary guidance according to adverse event risk versus benefit for various drug classes (13). For example, using an SGLT2 inhibitor as a glucose-lowering drug also has the benefit of potentially reducing progression of CKD with no notable increased risk of adverse events, but if a thiazolidinedione is used as a glucose-lowering drug in patients with type 2 diabetes who are at high risk of heart failure, there is no evidence of overall benefit, but there is an increased risk of adverse effects.

Pharmacological management of chronic conditions such as type 2 diabetes usually follows a linear, or stepwise, approach through which drugs are added (and optimized) or stopped based on their toxicity, efficacy, and/or patient-reported quality of life. Each treatment step requires a period of waiting to ascertain these effects before moving on to the next step. However, is a linear treatment approach always the most appropriate strategy for patients with a chronic progressive condition? In this section, we discuss a hypothetical treatment approach that uses multiple drugs simultaneously, each targeting a different biological pathway associated with disease progression. We first discuss an established example of a pillars of therapy hypothesis (with each pillar representing a different drug class) in the treatment of heart failure with reduced ejection fraction (HFrEF). Next, we focus on a pillar approach hypothesis for pharmacological management of CKD with type 2 diabetes.

Recently, the “four pillars of heart failure” was proposed as a strategy to treat patients with HFrEF (47). This four-pillars approach could reduce the risk of treatment delays and offer patients a health benefit compared with the current linear (stepwise) treatment approach (47). The linear approach to treating HFrEF involves initiation of first-line drug therapy with an ACE inhibitor plus a β-blocker, which is followed by a waiting period during which assessments and checks are made to determine whether the patient is responding to and tolerating treatment. Angiotensin receptor- neprilysin (ARN) inhibitors may replace ACE inhibitors (48). Additional therapies (e.g., a mineralocorticoid receptor antagonist) are then recommended for patients who do not respond to first-line therapy (48). This linear approach is standard across treatment guidelines, but it does introduce a time delay before advancing treatment to the next step in a population of patients who already have an acute, life-limiting condition (47).

In the proposed four-pillars approach to HFrEF, all four agents (ARN inhibitor, β-blocker, MRA, and SGLT2 inhibitor) (49) are initiated simultaneously, followed by optimization of dosing, when required (47). Clinical trial data from various studies support this approach. A meta-analysis using data from 58 randomized clinical trials found that combinations of some of these drugs (e.g., ARN inhibitor + β-blocker + MRA) provided incremental benefit in mortality and all-cause hospitalizations in patients with HFrEF compared with placebo; furthermore, the benefit appeared greater than single-drug class therapies versus placebo (50). Comparing data from three pivotal heart failure trials indirectly, Vaduganathan et al. (51) speculated that a combination of drugs from all four classes potentially reduces cardiovascular death and heart failure hospitalizations compared with conventional therapy with an ACE inhibitor or angiotensin receptor blocker (ARB) plus a β-blocker.

One of the main treatment goals for patients with type 2 diabetes is to maintain effective glycemic control, which reduces the risk of developing or slows the progression of diabetes-related complications such as CVD and/or CKD. The focus on glycemic control in type 2 diabetes is noted in the ADA 2023, KDIGO, and NICE guidelines (12,45,52). As noted previously, glycemic control in type 2 diabetes requires significant lifestyle changes on the part of patients, as well as pharmacological management (45,46).

Pharmacological management of type 2 diabetes follows a linear approach, and the initial drug therapy or therapies used depends on factors such as whether the patient has established CKD or CVD and/or their risk of developing either or both of these conditions based on factors such as their age, pregnancy status, and number and type of comorbidities (6,53,54). However, the progressive nature of type 2 diabetes means that first-line drug therapy may only be appropriate for a short period of time because of eventual loss of glycemic control (12,45).

Glucose toxicity resulting from chronic hyperglycemia (55) and failure to successfully treat patients with type 2 diabetes to their prescribed metabolic targets increase their risk of long-term microvascular and macrovascular complications such as CKD and cardiovascular events. Thus, delay before advancing to the next step in the linear treatment approach is a concern.

Once a CKD diagnosis is made, preventing kidney disease progression and reducing its cardiovascular impact are the focus of therapy, which includes maintaining glycemic and blood pressure control. Using the linear approach, drugs with different mechanisms of action may be introduced and optimized or removed in a step-up or step-down fashion guided by factors such as kidney function testing, tolerability, efficacy, and blood glucose levels. Regular testing can detect early signs of kidney damage and can also inform available treatment options at early stages.

A guide to the frequency of monitoring based on eGFR and albuminuria was provided in the KDIGO 2012 guidelines (Figure 1) (8–10). This guide may need to be adjusted according to an individual’s history and the underlying cause of kidney disease (8,9). However, the linear treatment approach in CKD introduces a delay between treatment steps because the time between kidney function tests, for example, may be several months, potentially allowing kidney disease to worsen given its progressive nature, particularly in patients with no or mild symptoms who may not seek medical help. For example, the nsMRA finerenone was approved in 2021 for the treatment of CKD associated with type 2 diabetes and has shown cardiorenal protective effects; however, this drug may be reserved by some health care professionals as a later-stage or second-line treatment option, so patients may not receive this drug until their kidney disease has progressed further (although this suggestion requires further exploring through real-world studies).

A pillar approach could bring together the main drug classes much earlier or simultaneously while still allowing for dose optimization, thereby removing the delay between steps that occurs with linear treatment. Thus, the pillar approach may be appropriate for chronic progressive diseases such as CKD, although clinical research is needed to test this hypothesis in CKD associated with type 2 diabetes. Although current treatment guidelines largely support a linear approach for CKD associated with type 2 diabetes, it is worth considering the feasibility of implementing a pillar approach instead. Hereafter, we explore this question further.

In the linear treatment approach to CKD associated with type 2 diabetes, a drug is initiated based on baseline A1C, blood pressure, and/or kidney function, followed by periodic monitoring of whether individualized targets expected from the drug’s action are achieved without intolerable side effects (6,12). If the targets are not met with the drug, then dose adjustments and/or the addition or substitution of different drugs may be needed (12,56). For example, for patients with type 2 diabetes, the ADA 2023 guidelines (Chapter 9) recommend using drugs that provide adequate efficacy to achieve and maintain glycemic goals, such as metformin or other drugs, including combination therapy. For patients taking the maximum tolerated dose of an ACE inhibitor or ARB, an SGLT2 inhibitor is recommended to reduce CKD progression, or a GLP-1 receptor agonist with proven CVD benefit in cases where SGLT2 inhibitors are contraindicated/not tolerated (45). Similarly, for blood pressure control, the ADA guidelines (Chapter 10) recommend an ACE inhibitor or ARB at the maximum tolerated dose indicated for patients with hypertension, diabetes, and albuminuria, with at least annual monitoring of albuminuria, eGFR, and serum potassium levels (6). If patients do not meet blood pressure targets, then addition of or change to a calcium channel blocker and/or diuretic may be considered (6). MRAs are recommended for patients who do not meet targets after receiving three classes of antihypertensive medications (including a diuretic) (6). In these linear treatment strategies, different therapies are given to a patient depending on the patient’s response to medications and/or the extent of CKD progression.

For patients with type 2 diabetes and CKD, both the ADA 2023 guidelines and the ADA/KDIGO consensus report recommend use of an SGLT2 inhibitor in patients with an eGFR ≥20 mL/min/1.73 m² (and ≥200 mg/g urinary albumin [4]) to reduce CKD progression and cardiovascular risk (4,13). In the ADA 2023 guidelines, an nsMRA (finerenone) is recommended for patients with CKD who are at increased risk for cardiovascular events or CKD progression (evidence grade A) (4). The ADA/KDIGO consensus report also recommends an nsMRA for patients with type 2 diabetes and an eGFR ≥25 mL/min/1.73 m², normal serum potassium levels, and albuminuria (13).

In contrast to a linear approach, taking a pillar approach in CKD treatment would mean that drugs that may slow CKD progression would be introduced at an early disease stage (ideally at CKD diagnosis), ultimately reducing the risk of CKD progression to ESKD. A drug class that targets a specific biological pathway associated with diabetes- related complications, including CKD progression, could be regarded as one pillar in a multipillar treatment. Patients with type 2 diabetes, CKD, and an eGFR ≥30 mL/min/1.73 m² may be receiving metformin (12) and an ACE inhibitor or ARB if there is hypertension and albuminuria, although in the ADA 2023 guidelines, metformin is not regarded as the first-line treatment for blood glucose control (45). For example, as soon as patients with diabetes and hypertension have confirmed abnormal kidney function, they could receive finerenone and/or an SGLT2 inhibitor in addition to metformin (and/or another antihyperglycemic agent) and an ACE inhibitor/ARB, with the intension of limiting progression of CKD (Figure 2). An integrated approach involving multiple risk parameters and scores could be used for monitoring the efficacy and safety of multiple drugs (Figure 2). It should be noted that, in the KDIGO guidelines, the combination of low doses of metformin and an SGLT2 inhibitor is suggested for patients with type 2 diabetes and an eGFR ≥30 mL/min/1.73 m² as “a practical approach” due to different mechanisms of action between the two drug classes (12).

In the CKD treatment of patients with diabetes, potential advantages of using a pillar approach are to 1) simultaneously target multiple pathways that contribute to CKD progression, 2) reduce some safety concerns (e.g., hyperkalemia caused by MRAs), 3) reduce the overall treatment duration by reducing the need for assessment and evaluation of kidney function between steps of therapy (discussed previously), and 4) minimize risks of silent CKD progression, which may not be captured by single surrogate markers (discussed previously). Potential disadvantages of a pillar approach to CKD treatment are 1) lack of safety and efficacy data from dedicated multicombination drug trials in real-world settings, 2) potential use of unnecessary drugs, 3) pushback from clinicians who are familiar with and prefer the current conventional approach, 4) resistance from patients who may reject the idea of polytherapy, and 5) costs and limitations in access to simultaneous drug treatments.

The current lack of clinical trials to evaluate the effect of combination therapies versus conventional first-line drug therapies for CKD associated with type 2 diabetes (and hence the opportunity to demonstrate additive or synergistic benefits) makes assessing the feasibility of the pillar approach challenging. However, guidelines have provided recommendations regarding the use of certain drug combinations (although not a three- or four-pillar combination). The KDIGO 2022 guidelines suggest a combination of an SGLT2 inhibitor and metformin for patients with type 2 diabetes (eGFR ≥30 mL/min/1.73 m²) but warn against combining an ACE inhibitor with an ARB (due to higher risk and marginal benefit) or combining an ACE inhibitor or ARB with a direct renin inhibitor due to safety concerns (12).

The potential inclusion of finerenone as a pillar in the treatment of CKD associated with type 2 diabetes was discussed previously. Here we give an overview of finerenone clinical trials.

Finerenone was first approved by the U.S. Food and Drug Administration (FDA) in 2021, which led to the eventual inclusion of this drug in the ADA and KDIGO guidelines as a therapeutic option for CKD associated with type 2 diabetes. The Finerenone in Reducing Kidney Failure and Disease Progression in Diabetic Kidney Disease (FIDELIO-DKD) (57–59) and Finerenone in Reducing Cardiovascular Mortality and Morbidity in Diabetic Kidney Disease (FIGARO-DKD) (60–62) studies demonstrated reduced CKD progression and cardiovascular events and hospitalization for heart failure in patients with CKD associated with type 2 diabetes who were treated with finerenone. The first approval of finerenone was based on the results of FIDELIO-DKD, which showed a reduced risk of sustained eGFR decline, kidney failure, cardiovascular death, nonfatal myocardial infarction, and hospitalization for heart failure compared with placebo in adults with CKD associated with type 2 diabetes.

Because patients treated with finerenone in FIDELIO-DKD and FIGARO-DKD received other types of drugs at baseline and/or post-baseline, the results from these trials may provide insight for researchers assessing the feasibility of pillar therapy for CKD associated with type 2 diabetes. All randomized patients in the trials received an ACE inhibitor or ARB at the optimized dose (per manufacturer’s protocol) at baseline (57,61). In FIDELIO-DKD, 4.4 and 10.9% of patients treated with finerenone also received an SGLT2 inhibitor at baseline or at any time during treatment (baseline + post-baseline), respectively (57,63). Importantly, although SGLT2 inhibitor use did not affect the reduction in UACR and key secondary composite outcomes, hyperkalemia events were fewer with finerenone in the SGLT2 inhibitor group (63). A subgroup analysis of patients receiving finerenone with an SGLT2 inhibitor at baseline showed a 55% lower risk of hyperkalemia compared with the overall group (hazard ratio 0.45, 95% CI 0.27–0.75) (64). Also in FIDELIO-DKD, 6.7 and 13.3% of patients treated with finerenone received a GLP-1 receptor agonist at baseline and at any time during the treatment (baseline + post-baseline), respectively (57,61). However, a subgroup analysis of patients receiving finerenone with a GLP-1 receptor agonist demonstrated no additional benefit of the GLP-1 receptor agonist for the primary kidney or secondary cardiovascular outcomes in patients treated with finerenone (65).

Because the supporting data for the pillar approach in CKD associated with type 2 diabetes are based largely on subgroup analyses derived from large studies in which multiple agents were not initiated simultaneously, dedicated carefully designed combination studies are necessary to evaluate the feasibility of a pillar approach. CONFIDENCE is an ongoing, parallel-group, double-blind, three-arm phase 2 trial to assess the efficacy and safety of finerenone plus the SGLT2 inhibitor empagliflozin compared with finerenone or empagliflozin alone in patients with CKD associated with type 2 diabetes. The primary end point is change from baseline in UACR, and secondary end points include changes in UACR and eGFR, acute kidney injury, hyperkalemia, and hypoglycemia. It will be interesting to see what the combined effect of an SGLT2 inhibitor plus finerenone has on outcomes compared with either treatment alone. Further subgroup analyses from FIDELIO-DKD and FIGARO-DKD and real-world data, possibly using multiple-risk parameters and scores, may provide further insight into the feasibility of a pillar approach in the treatment of CKD associated with diabetes.

In this review, we discussed the existing linear and hypothetical pillar treatment approaches to CKD associated with type 2 diabetes, with close reference to the ADA 2023 guidelines, supported where appropriate by the KDIGO guidelines (representing a global approach to kidney disease) and NICE guidelines (representing a country-specific approach to CKD outside of the United States). By simultaneously targeting the multiple pathways involved in CKD progression, as well as cardiovascular events, a pillar approach could potentially bring an additive/synergistic benefit to patients in the treatment of CKD associated with type 2 diabetes and its comorbidities. FDA approval of the nsMRA finerenone and a revised label indication for the SGLT2 inhibitor dapagliflozin, both in 2021, have expanded treatment options for patients with CKD, enabling clinicians to consider the possibility of a pillar therapy approach. Although it is still hypothetical and more clinical and real-world studies are needed, a pillar approach, combined with proactive early detection and early referral when needed, could enable the risks of disease progression and cardiovascular events in CKD associated with type 2 diabetes to be reduced and the condition to be more efficiently managed moving forward. A summary of the main themes of this article is presented in Figure 3.

Medical writing support was provided by Tomo Sawado, PhD, of Alligent – Envision Pharma Group and funded by Bayer Corporation. Envision Pharma Group’s services complied with international guidelines for Good Publication Practice (GPP4).

J.M. is a consultant and promotional speaker for Bayer, Boehringer Ingelheim, Eli Lilly, and Novo Nordisk; has conducted clinical research for Novo Nordisk; and is an advisory board member for Abbott, Bayer, Boehringer Ingelheim, Eli Lilly, Intarcia, Novo Nordisk, and Sanofi. S.D.-J. has led clinical trials for AstraZeneca, Bayer, Boehringer Ingelheim, and Novo Nordisk; has received fees from AstraZeneca, Bayer, Boehringer Ingelheim, Janssen, Merck, and Sanofi; holds equity interests in Aerami Therapeutics and Jana Care; and serves on the editorial boards of the American Journal of the Medical Sciences, BMJ Diabetes Research & Care, Experimental Biology & Medicine, Frontiers in Endocrinology, and Scientific Reports. V.F. has received research support (to his institution) or grants from Fractyl Health and Jaguar Gene Therapy; has received honoraria for consulting and lectures from Abbott, Asahi Kasei Pharma, AstraZeneca, Bayer, Novo Nordisk, and Sanofi; holds stock options with BRAVO4Health and Mellitus Health; has stock in Abbott and Amgen; and has a patent with BRAVO Risk Engine for Predicting Diabetes Complications (pending). J.J.N. has received consulting fees from Bayer, Novo Nordisk, and Sanofi and served on a speaker’s bureau for Dexcom. S.E.R. has received research funds (to her institution) from AstraZeneca and Bayer; is a member of a scientific advisory board for AstraZeneca, Bayer, and Teladoc; is president-elect of the National Kidney Foundation; and is an employee of Beth Israel Lahey Health.

All authors contributed to writing and reviewing the manuscript. J.M. is the guarantor of this work and, as such, had full access to all the data reported and takes responsibility for the integrity of the article content.