New insights to fibroblast growth factor 23 in kidney transplant.

Some abnormalities in mineral metabolism are evident even at very early stages of chronic kidney disease (CKD) and are important determinants of subsequent bone and cardiovascular disease.1,2 A decade ago, fibroblast growth factor 23 (FGF23) was recognized by a few as the protein responsible for several rare inherited and acquired syndromes of osteomalacia and rickets.3,4 Only recently have studies demonstrated the significance of FGF23 in mineral metabolism in the larger population of patients with CKD and ESRD.5–7 In 2007, Fliser et al.8 reported that higher FGF23 levels were strongly associated with progression of non diabetic CKD. The next year, Gutierrez et al.9 published that incident hemodialysis patients with higher FGF23 levels were at substantially greater risk for all-cause mortality. These and other studies generated considerable interest in the role of FGF23 in mineral metabolism homeostasis in CKD. Recent reports extend these findings by showing that FGF23 levels are elevated at very early stages of CKD,2 and the associations of FGF23 with all-cause mortality or cardiovascular disease extend to patients with stages 3 to 4 CKD,10 and even to individuals with ostensibly normal kidney function.11 In this issue of JASN, Wolf et al.12 report that higher FGF23 levels associate with the composite outcome of all-cause mortality or kidney allograft loss among 984 stable transplant recipients. Most of the participants were several years after transplantation. This finding is significant for several reasons. First, in conjunction with other literature, this study demonstrates that higher FGF23 levels identify patients at increased risk for adverse outcomes across the spectrum of CKD. Second, because most kidney transplant recipients have survived an extended period on dialysis before receiving an allograft and often have a high burden of vascular disease, the study demonstrates that FGF23 continues to provide risk information in this late stage of disease. Last, given the pattern of other mineral metabolism abnormalities observed in kidney transplant recipients, the study provides new insights into potential mechanisms, as described further next. With these discoveries come new challenges. Among the most pressing is a better understanding of mechanisms responsible for the link of FGF23 with adverse outcomes.7 Several possibilities require special consideration. A main biological function of FGF23 is to increase urine phosphorus excretion.4,13 A wealth of data spanning from the laboratory to population-based studies implicates hyperphosphatemia as a key factor inducing and promoting arterial calcification.14 –16 Thus, perhaps high FGF23 levels are linked with mortality through alterations in phosphorus homeostasis. Several studies have investigated this possibility. Consistently and observed again in the article by Wolf et al. in this issue, statistical adjustment for serum phosphorus levels measured concurrently with FGF23 does not attenuate its relationship with outcomes.8 –12 However, phosphorus may remain an important intermediary nonetheless. Contemporary clinical laboratories precisely measure serum phosphorus levels, typically with coefficients of variation 3%. However, there is considerable biological variability in serum phosphorus levels within individuals over time.17 This is analogous to serum glucose, for which one can precisely determine the blood level at a given moment, but it gives a mere snapshot of average glucose levels over time. Thus, one possibility is that higher FGF23 levels increase in response to higher phosphorus levels18 and may serve as a more accurate indicator of time-averaged serum phosphorus exposure than serum phosphorus levels themselves, analogous to a hemoglobin A1c as an indicator of time-averaged glucose levels. Future studies with repeated measurements of serum phosphorus over time are required to investigate this possibility. However, the study by Wolf et al. in this issue provides some early insights arguing against this hypothesis. Kidney transplant recipients frequently develop posttransplantation hypophosphatemia. Although this abnormality often wanes with time, a subgroup remains persistently hypophosphatemic.19 In the study by Wolf et al. did not provide repeated measures of serum phosphorus, subgroup analysis demonstrated that higher FGF23 levels remain strongly associated with death or allograft loss in patients with serum phosphorus in the lowest tertile ( 2.9 mg/dl) at study enrollment.12 Although not definitive, these data suggest that viewing FGF23 as a marker of time-averaged serum phosphorus may be too simplistic. Published online ahead of print. Publication date available at www.jasn.org.