BNP trajectory and the risk of heart failure hospitalisation in stage 2 CKM syndrome
K Teramoto, M Sato, S Ogata, K NishimuraAbstract
Background
Stage 2 cardiovascular-kidney-metabolic (CKM) syndrome includes a heterogeneous population with metabolic/renal diseases predisposing to cardiovascular diseases, including heart failure (HF). Within this group, brain natriuretic peptide (BNP) trajectories may be vital for precise risk stratification.
Purpose
Latent groups with differential BNP trajectories and incident HF were explored in stage 2 CKM syndrome.
Methods
Anonymized data from 7,979 Japanese individuals with stage 2 CKM syndrome identified using ICD-10 codes who had at least one BNP measurement in January 2019 (i.e., baseline) were analyzed. A joint latent class mixed model (JLCMM) was constructed using repeatedly measured BNP and incident HF hospitalisation. Age at baseline was included in the membership and longitudinal trajectory submodels to account for class membership likelihood and physiological associations with increasing BNP trend. The model was trained on 70% of the overall cohort and internally validated on the remaining 30%. Multivariable logistic regression was performed to identify risk factors for the classes.
Results
In a training dataset of 5,579 individuals (mean age 69.6 years, 48% female, median follow-up of 5.2 years), JLCMM identified 2 groups with distinct disease trajectories. Based on the model, 5,145 were assigned to Class 1 (69.3 years with an HF event rate of 0.41/100 person-years) and 434 to Class 2 (73.9 years and 5.78/100 person-years). Class 1 was characterized by low and stable BNP (predicted median baseline BNP of a 70-year-old: 27.3 pg/ml; β-coefficient for slope: -0.001, p-value=0.83). Class 2 was characterized by increased baseline BNP and a sharp annual increase (predicted median baseline BNP of a 70-year-old: 170.1 pg/ml; β for slope: 0.333, p<0.001) with a higher risk of HF hospitalization over time. The membership submodel found that age was associated with the likelihood of being in Class 2 (OR 1.03, 95%CI [1.01-1.06]). The trajectory submodel identified that, regardless of class, age accelerates the rate of BNP elevation (interaction between age and time, β=0.001, p=0.029). Nevertheless, both young and older individuals demonstrate a rapid increase in BNP and elevated HF risk in class 2 (Figure 1). The model produced a concordance index of 0.74 when applied in the internal testing cohort. Having CKD and hypertension at baseline were associated with increased odds of being classified as Class 2 (OR 3.20, 95% CI [2.64-3.86], and 1.46 [1.22-1.77], respectively).
Conclusions
Distinct BNP trajectory patterns and HF outcomes were identified in stage 2 CKM syndrome. While age accelerates the BNP trajectory regardless of the pattern, pathological progression, characterized by a synchronized, sharp increase in BNP trajectory and HF risk, appears to predominate over aging. BNP trajectory-based assessment may be helpful in identifying high-risk individuals with rapid pathological progression.Figure 1