ASSOCIATION BETWEEN LONG-TERM OUTCOMES AND BIOMARKERS IN PATIENTS WITH CHRONIC KIDNEY DISEASE: A 15-YEAR FOLLOW-UP STUDY

Estimated glomerular filtration rate (eGFR), albuminuria, and causative kidney diseases are key prognostic factors for chronic kidney disease (CKD). Although several biomarkers have been reported as predictors of CKD outcomes, few studies have evaluated their association with long-term prognosis beyond 10 years. We previously reported that methylglyoxal (MG), a precursor of advanced glycation end products, predicted 3-year outcomes in CKD patients, particularly those with diabetic nephropathy (DN). This study aimed to investigate whether multiple serum and urinary biomarkers measured from stored samples are associated with long-term outcomes in the same CKD cohort. 

We recruited 150 patients with CKD who had received outpatient care at two public hospitals from 2009 to 2010 for a three-year prospective study. The inclusion criteria were age over 20 years, and eGFR below 60 mL/min/1.73 m2. Patients with an eGFR below 45 mL/min/1.73 m2 were preferentially enrolled, while those with serious infections, active inflammation, active malignancies, liver diseases, or a serum creatinine greater than 5.0 mg/dL were excluded. This study was subsequently extended with a 12-year retrospective follow-up, resulting in a total observation period of 15 years. Serum biomarkers included MG, 3-deoxyglucosone, pentosidine, intact parathyroid hormone, 1,25-(OH)2 vitamin D, Fibroblast growth factor 23 (FGF-23), and hepcidin. Urinary biomarkers included 8-hydroxy-2'-deoxyguanosine and liver-type fatty acid-binding protein. The primary endpoint was a composite of all-cause mortality, end-stage kidney disease (ESKD), and cardiovascular events (CVE). Survival analysis was performed using the Kaplan-Meier Method. Associations between biomarkers and outcomes were assessed using multivariate Cox proportional hazards models, adjusted for age, sex, eGFR, diabetes, systolic blood pressure (sBP), and urinary protein.

The mean age of the patients was 62 ± 12 years, and 97 patients (65%) were male. Among the 150 patients, 20 (13%) were diagnosed with DN, 45 (30%) with hypertensive nephrosclerosis (HTN), and 85 (57%) with chronic glomerulonephritis (CGN) or other kidney diseases. Baseline parameters were as follows: urinary protein, 0.99 g/gCr (interquartile range, 0.36–3.03); eGFR, 25.0 ± 12.1 mL/min/1.73 m²; MGO, 402 nmol/L (276–556); and FGF-23, 83 pg/mL (57–126). During follow-up, 119 patients (79%) experienced an event: 95 (63%) developed ESKD, 14 (9%) died, and 10 (7%) had CVEs. Event-free survival at 15 years were 41% for CKD stage 3; 6% for stage 4; and 3% for stage 5. Among the primary CKD diseases, DN and HTN showed poorer prognosis in that order than CGN or others, although the differences were not statistically significant. In multivariable analysis, eGFR (HR 0.52, 95% CI 0.42–0.64), urinary protein (HR 1.58, 95% CI 1.32–1.87), and log-transformed FGF-23 (HR 2.27, 95% CI 1.02–5.02) were independently associated with long-term outcomes. MG predicted short-term outcomes but not long-term prognosis.

In this 15-year follow-up study, long-term prognosis was extremely poor, with nearly 80% experiencing ESKD, CVEs, or death. In addition to eGFR and proteinuria, FGF-23 was independently associated with long-term outcomes. The content presented in this abstract was submitted for the 68th Annual Meeting of the Japanese Society of Nephrology.