Dataset Information

Assessing reproducibility and utility of clustering of patients with type 2 diabetes and established CV disease (SAVOR -TIMI 53 trial).

ABSTRACT:

Objective

To assess the reproducibility and clinical utility of clustering-based subtyping of patients with type 2 diabetes (T2D) and established cardiovascular (CV) disease.

Methods

The cardiovascular outcome trial SAVOR-TIMI 53 (n = 16,492) was used. Analyses focused on T2D patients with established CV disease. Unsupervised machine learning technique called "k-means clustering" was used to divide patients into subtypes. K-means clustering including HbA1c, age of diagnosis, BMI, HOMA2-IR and HOMA2-B was used to assign clusters to the following diabetes subtypes: severe insulin deficient diabetes (SIDD); severe insulin-resistant diabetes (SIRD); mild obesity-related diabetes (MOD); mild age-related diabetes (MARD). We refer these subtypes as "clustering-based diabetes subtypes". A simulation study using randomly generated data was conducted to understand how correlations between the above variables influence the formation of the cluster-based diabetes subtypes. The predictive utility of clustering-based diabetes subtypes for CV events (3-point MACE), renal function reduction (eGFR decrease >30%) and diabetic disease progression (introduction of additional anti-diabetic medication) were compared with conventional risk scores. Hazard ratios (HR) were estimated by Cox-proportional hazard models.

Results

In the SAVOR-TIMI 53 trial based dataset, the percentage of the clustering-based T2D subtypes were; SIDD (18%), SIRD (17%), MOD (29%), MARD (37%). Using the simulated dataset, the diabetes subtypes could be largely reproduced from a log-normal distribution when including known correlations between variables. The predictive utility of clustering-based diabetic subtypes on CV events, renal function reduction, and diabetic disease progression did not show an advantage compared to conventional risk scores.

Conclusions

The consistent reproduction of four clustering-based T2D subtypes can be explained by the correlations between the variables used for clustering. Subtypes of T2D based on clustering had limited advantage compared to conventional risk scores to predict clinical outcome in patients with T2D and established CV disease.

SUBMITTER: Aoki Y

PROVIDER: S-EPMC8604302 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Importance:An elevated level of urinary albumin to creatinine ratio (UACR) is a marker of renal dysfunction and predictor of kidney failure/death in patients with type 2 diabetes. The prognostic use of UACR in established cardiac biomarkers is not well described. Objective:To evaluate whether UACR offers incremental prognostic benefit beyond risk factors and established plasma cardiovascular biomarkers. Design, Setting, and Participants:The Saxagliptin Assessment of Vascular Outcomes Recorded in Patients With Diabetes Mellitus-Thrombolysis in Myocardial Infarction (SAVOR-TIMI) 53 study was performed from May 2010 to May 2013 and evaluated the safety of saxagliptin vs placebo in patients with type 2 diabetes with overt cardiovascular disease or multiple risk factors. Median follow-up was 2.1 years (interquartile range, 1.8-2.3 years). Interventions:Patients were randomized to saxagliptin vs placebo plus standard care. Main Outcomes and Measures:Baseline UACR was measured in 15?760 patients (95.6% of the trial population) and categorized into thresholds. Results:Of 15?760 patients, 5205 were female (33.0%). The distribution of UARC categories were: 5805 patients (36.8%) less than 10 mg/g, 3891 patients (24.7%) at 10 to 30 mg/g, 4426 patients (28.1%) at 30 to 300 mg/g, and 1638 patients (10.4%) at more than 300 mg/g. When evaluated without cardiac biomarkers, there was a stepwise increase with each higher UACR category in the incidence of the primary composite end point (cardiovascular death, myocardial infarction, or ischemic stroke) (3.9%, 6.9%, 9.2%, and 14.3%); cardiovascular death (1.4%, 2.6%, 4.1%, and 6.9%); and hospitalization for heart failure (1.5%, 2.5%, 4.0%, and 8.3%) (adjusted P?<?.001 for trend). The net reclassification improvement at the event rate for each end point was 0.081 (95% CI, 0.025 to 0.161), 0.129 (95% CI, 0.029 to 0.202), and 0.056 (95% CI, -0.005 to 0.141), respectively. The stepwise increased cardiovascular risk associated with a UACR of more than 10 mg/g was also present within each chronic kidney disease category. The UACR was associated with outcomes after including cardiac biomarkers. However, the improvement in discrimination and reclassification was attenuated; net reclassification improvement at the event rate was 0.022 (95% CI, -0.022 to 0.067), -0.008 (-0.034 to 0.053), and 0.043 (-0.030 to 0.052) for the primary end point, cardiovascular death, and hospitalization for heart failure, respectively. Conclusions and Relevance:In patients with type 2 diabetes, UACR was independently associated with increased risk for a spectrum of adverse cardiovascular outcomes. However, the incremental cardiovascular prognostic value of UACR was minimal when evaluated together with contemporary cardiac biomarkers. Trial Registration:clinicaltrials.gov Identifier: NCT01107886.

Project description:Background: In patients with type 2 diabetes mellitus, sodium-glucose cotransporter 2 inhibitors reduce the risk of hospitalization for heart failure (HHF). We assessed the effect of ertugliflozin on HHF and related outcomes.Methods: VERTIS CV (Evaluation of Ertugliflozin Efficacy and Safety Cardiovascular Outcomes Trial), a double-blind, placebo-controlled trial, randomly assigned patients with type 2 diabetes mellitus and atherosclerotic cardiovascular (CV) disease to once-daily ertugliflozin 5 mg, 15 mg, or placebo. Prespecified secondary analyses compared ertugliflozin (pooled doses) versus placebo on time to first event of HHF and composite of HHF/CV death, overall and stratified by prespecified characteristics. Cox proportional hazards modeling was used with the Fine and Gray method to account for competing mortality risk, and Andersen-Gill modeling to analyze total (first+recurrent) HHF and total HHF/CV death events.Results: A total of 8246 patients were randomly assigned to ertugliflozin (n=5499) or placebo (n=2747); n=1958 (23.7%) had a history of heart failure (HF) and n=5006 (60.7%) had pretrial ejection fraction (EF) available, including n=959 with EF ?45%. Ertugliflozin did not significantly reduce first HHF/CV death (hazard ratio [HR], 0.88 [95% CI, 0.75-1.03]). Overall, ertugliflozin reduced risk for first HHF (HR, 0.70 [95% CI, 0.54-0.90]; P=0.006). Previous HF did not modify this effect (HF: HR, 0.63 [95% CI, 0.44-0.90]; no HF: HR, 0.79 [95% CI, 0.54-1.15]; P interaction=0.40). In patients with HF, the risk reduction for first HHF was similar for those with reduced EF ?45% versus preserved EF >45% or unknown. However, in the overall population, the risk reduction tended to be greater for those with EF ?45% (HR, 0.48 [95% CI, 0.30-0.76]) versus EF >45% (HR, 0.86 [95% CI, 0.58-1.29]). Effect on risk for first HHF was consistent across most subgroups, but greater benefit of ertugliflozin was observed in 3 populations: baseline estimated glomerular filtration rate <60 mL·min-1·1.73 m-2, albuminuria, and diuretic use (each P interaction <0.05). Ertugliflozin reduced total events of HHF (rate ratio, 0.70 [95% CI, 0.56-0.87]) and total HHF/CV death (rate ratio, 0.83 [95% CI, 0.72-0.96]).Conclusions: In patients with type 2 diabetes mellitus, ertugliflozin reduced the risk for first and total HHF and total HHF/CV death, adding further support for the use of sodium-glucose cotransporter 2 inhibitors in primary and secondary prevention of HHF. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT01986881.