MO056
CTERMINAL FGF23 VERSUS INTACT FGF23 AS CARDIOVASCULAR PREDICTOR IN CHRONIC KIDNEY DISEASE

INTRODUCTION AND AIMS: Patients with chronic kidney disease have an enormously higher risk of cardiovascular events than people with intact renal function. Beyond classical / traditional cardiovascular risk factor, chronic kidney disease - mineral and bone disorder (CKD-MBD) may contribute to this high cardiovascular risk. CKD-MBD comprises an increased expression of the phosphaturic fibroblast growth factor 23 (FGF-23). In several cohort studies, plasma levels of FGF-23 were associated with markedly increased risk of cardiovascular events. Of note, two different ELISAs for FGF-23 measurement are available, which either measure both biologically intact FGF-23 and its inactive fragments (“C-terminal FGF-23”), or only biologically intact FGF-23. Until now, most large cohort studies measured C-terminal FGF-23 as outcome predictor, while data on intact FGF-23 are very sparse. It has been claimed that most FGF-23 in advanced CKD is biologically active, and the use of two different ELISAs therefore dispensable. The purpose of this study was firstly to analyze the degree to which intact FGF-23 accumulates across the spectrum of CKD, and secondly whether intact FGF-23 outperforms cterminal FGF-23 as a predictor of cardiovascular events.

METHODS: Among 544 patients of our ongoing CARE for HOMe - study, we measured c-terminal FGF-23 and intact FGF-23 from plasma samples stored at study inclusion (varying from 2008 to 2014). Glomerular filtration rate (eGFR) was estimated with the MDRD equation; all patients had CKD G 2 - G 4 at study inclusion. We followed these patients annually to assess the occurrences of cardiovascular events, defined as coronary, carotid and peripheral arterial revascularization, stroke, acute myocardial infarction or death of any cause.

RESULTS: We found a rather weak correlation between c-terminal FGF-23 and intact FGF-23 (r = 0.398; p < 0.001). eGFR was correlated to a similar extent with c-terminal FGF-23 (r = 0.419) and intact FGF-23 (r = 0.339), but not with the intact / c-terminal FGF-23 ratio (r = 0,008; p = 0.789). During a mean follow-up of 4.6 ± 2.0 years, 146 cardiovascular events occurred. In univariate Kaplan-Meier analyses, tertiles of c-terminal FGF-23 and intact FGF-23 predicted cardiovascular events in CKD patients (log-rank test <0.001). In Cox regression analysis, c-terminal FGF-23, but not intact FGF-23, remained an independent outcome predictor after adjustment for eGFR (and subsequently for age, gender, diabetes mellitus, intact FGF-23 and plasma phosphorus).

CONCLUSIONS: Although it has been hypothesized that the most FGF-23 that accumulates in CKD is biologically active, i. e. intact FGF-23, our data show that the iFGF/cFGF - ratio remains stable across the spectrum of CKD. Surprisingly we found that C-terminal FGF-23 outperformed intact FGF-23 as a predictor of future cardiovascular events.