Rationale, design, and baseline characteristics of the CArdiovascular safety and Renal Microvascular outcomE study with LINAgliptin (CARMELINA®): a randomized, double-blind, placebo-controlled clinical trial in patients with type 2 diabetes and high cardio-renal risk

Abstract

Background: Cardiovascular (CV) outcome trials in type 2 diabetes (T2D) have underrepresented patients with chronic kidney disease (CKD), leading to uncertainty regarding their kidney efficacy and safety. The CARMELINA^® trial aims to evaluate the effects of linagliptin, a DPP-4 inhibitor, on both CV and kidney outcomes in a study population enriched for cardio-renal risk.

Methods: CARMELINA^® is a randomized, double-blind, placebo-controlled clinical trial conducted in 27 countries in T2D patients at high risk of CV and/or kidney events. Participants with evidence of CKD with or without CV disease and HbA1c 6.5-10.0% (48-86 mmol/mol) were randomized 1:1 to receive linagliptin once daily or matching placebo, added to standard of care adjusted according to local guidelines. The primary outcome is time to first occurrence of CV death, non-fatal myocardial infarction, or non-fatal stroke. The key secondary outcome is a composite of time to first sustained occurrence of end-stage kidney disease, ≥ 40% decrease in estimated glomerular filtration rate (eGFR) from baseline, or renal death. CV and kidney events are prospectively adjudicated by independent, blinded clinical event committees. CARMELINA^® was designed to continue until at least 611 participants had confirmed primary outcome events. Assuming a hazard ratio of 1.0, this provides 90% power to demonstrate non-inferiority of linagliptin versus placebo within the pre-specified non-inferiority margin of 1.3 at a one-sided α-level of 2.5%. If non-inferiority of linagliptin for the primary outcome is demonstrated, then its superiority for both the primary outcome and the key secondary outcome will be investigated with a sequentially rejective multiple test procedure.

Results: Between July 2013 and August 2016, 6980 patients were randomized and took ≥ 1 dose of study drug (40.6, 33.1, 16.9, and 9.4% from Europe, South America, North America, and Asia, respectively). At baseline, mean ± SD age was 65.8 ± 9.1 years, HbA1c 7.9 ± 1.0%, BMI 31.3 ± 5.3 kg/m², and eGFR 55 ± 25 mL/min/1.73 m². A total of 5148 patients (73.8%) had prevalent kidney disease (defined as eGFR < 60 mL/min/1.73 m² or macroalbuminuria [albumin-to-creatinine ratio > 300 mg/g]) and 3990 patients (57.2%) had established CV disease with increased albuminuria; these characteristics were not mutually exclusive. Microalbuminuria (n = 2896 [41.5%]) and macroalbuminuria (n = 2691 [38.6%]) were common.

Conclusions: CARMELINA^® will add important information regarding the CV and kidney disease clinical profile of linagliptin by including an understudied, vulnerable cohort of patients with T2D at highest cardio-renal risk. Trial registration ClinicalTrials.gov identifier-NCT01897532; registered July 9, 2013.

Keywords: Cardiovascular diseases; Clinical trial, phase IV; Diabetes mellitus, type 2; Diabetic nephropathies; Dipeptidyl-peptidase IV inhibitors; Linagliptin; Research design; Treatment outcome.

Fig. 1

10-year mortality in T2D by kidney disease manifestation in the United States. The dashed line indicates mortality in persons without diabetes or kidney disease (the reference group, 7.7%). The numbers above bars indicate excess mortality above the reference group. Error bars indicate upper limits of the 95% confidence intervals. Republished with permission of the American Society of Nephrology from Afkarian et al. [21]; permission conveyed through Copyright Clearance Center, Inc. *Impaired GFR was defined as GFR ≤ 60 mL/min/1.73 m². GFR glomerular filtration rate, T2D type 2 diabetes

Fig. 2

Design of the CARMELINA^® trial. *Additional glucose-lowering therapy may be given on top of study medication if HbA1c > 7.5%; investigators are encouraged to treat all other CV risk factors in accordance with local or regional standards of care. ^†Participants who stop study drug early are observed until study end (not just until 30 days after the end of treatment with study drug). CV cardiovascular, HbA1c glycated hemoglobin A1c

Fig. 3

Statistical testing for the primary and secondary endpoints. For the final analysis, the first hypothesis (non-inferiority of the primary endpoint [3P-MACE]) will be tested at the one-sided alpha-level of 2.5%. In case of significance, the null hypothesis (H₀) is rejected in a confirmatory sense and the next set of hypotheses (two separate hypothesis tests) will be tested: (a) test of the primary endpoint for superiority and (b) test of the composite renal endpoint for superiority. To adjust for multiplicity, a sequentially rejective multiple test procedure will be applied. Both one-sided hypotheses H_0(Sup1) and H_0(Sup2) will be tested separately, at the initial alpha-levels of ×0.2 alpha for the primary endpoint and ×0.8 alpha for the composite renal endpoint, respectively. If both null hypotheses cannot be rejected at these initial alpha-levels, the procedure stops and for neither endpoint can superiority be declared. After having shown superiority for one of these endpoints, the used alpha can be shuffled to the other test: If H_0(Sup2) is rejected at the alpha-level of ×0.8 alpha, then H_0(Sup1) can be tested at the full alpha-level of 2.5% (one-sided). If H_0(Sup1) is rejected at the alpha-level of ×0.2 alpha, then H_0(Sup2) can be tested at the full alpha-level of 2.5% (one-sided). *Cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke. H₀ null hypothesis, 3P-MACE 3-point major adverse cardiovascular events

Fig. 4

Proportion of patients included in the CARMELINA^® trial with established CV disease, prevalent kidney disease, or both. *110 patients without established CV disease had eGFR ≥ 60 mL/min/1.73 m² and UACR ≤ 300 mg/g. ^†Defined as albuminuria (UACR ≥ 30 mg/g or ≥ 30 μg albumin/min or ≥ 30 mg albumin/24 h) and prevalent macrovascular disease (≥ 1 of the following: confirmed history of myocardial infarction; advanced coronary artery disease; high-risk single-vessel coronary artery disease; history of ischemic or hemorrhagic stroke; presence of carotid artery disease; presence of peripheral artery disease). ^‡Defined as eGFR < 60 mL/min/1.73 m² or macroalbuminuria (UACR > 300 mg/g). CKD chronic kidney disease, CV cardiovascular, eGFR estimated glomerular filtration rate, UACR urinary albumin-to-creatinine ratio

Fig. 5

Prognosis of CKD in the CARMELINA^® trial population by eGFR and albuminuria categories.*CKD chronic kidney disease, eGFR estimated glomerular filtration rate, KDIGO Kidney Disease: Improving Global Outcomes, UACR urinary albumin-to-creatinine ratio

Fig. 6

Proportion of patients with reduced kidney function at baseline (eGFR < 60 mL/min/1.73 m²) in CARMELINA^® compared to previously reported CV outcome trials of non-insulin glucose-lowering drugs for T2D. *eGFR < 50 mL/min/1.73 m². CV cardiovascular, eGFR estimated glomerular filtration rate, QW once weekly, T2D type 2 diabetes