Higher levels of fibroblast growth factor 23 (FGF23) were associated with a greater risk of mortality and progression to end-stage renal disease (ESRD) in patients with chronic kidney disease (CKD), reported a large prospective study published in the Journal of the American Medical Association (2011;305:2432-2439).
The mortality association extends to earlier stages of CKD what has been demonstrated in patients with end-stage renal disease (ESRD), and the ESRD link reinforces what has been seen in smaller studies of the CKD population.
“We've long felt that the higher FGF23 level could be an earlier biomarker to define which patients have abnormalities of phosphorus metabolism before they manifest an elevation in their serum phosphate,” said senior author Myles Wolf, MD, MMSc, Assistant Dean for Translational and Clinical Research at the University of Miami Miller School of Medicine.
“If this were proven to be true, then we could target disordered phosphorus metabolism therapeutically when the blood level of phosphorus is normal but the FGF23 is high.”
Tamara Isakova, MD, MMSc, also of the Miller School of Medicine, was the lead author.
Chronic Renal Insufficiency Cohort
The largest study of its kind to date, this investigation drew upon 3,879 patients with Stages 2 through 4 CKD in the Chronic Renal Insufficiency Cohort (CRIC). The analysis was from the time of enrollment, which began in June 2003, to death, voluntary withdrawal, loss to follow-up, or the end of 2008. It was one of several analyses in the cohort study supported by multiple grants from the National Institutes of Health.
The median follow-up period for the cohort was 3.5 years, during which 266 participants died and 410 reached end-stage renal disease.
Median FGF23 levels were significantly higher in those who died (234 RU/mL) or reached end-stage renal disease (236 RU/mL) than in those who remained event-free (133 RU/mL).
Multivariate analysis adjusting for demographic characteristics and chronic kidney disease-specific risk factors did not significantly change the relationship between FGF23 and mortality. Participants in the highest quartile were 4.3 times more likely to die than those in the lowest quartile, and the intermediate quartiles had intermediate risks.
Contrary to the researchers' expectations, FGF23 was more strongly associated with mortality than were traditional cardiovascular disease- and chronic kidney disease-specific risk factors.
“When you look at FGF23 as a risk factor for mortality in kidney patients versus other well-known, powerful risk factors—most notably the amount of proteinuria and the severity of renal dysfunction—what was amazing to us was that a high FGF23 not only out-performed serum phosphate as a predictor of mortality, but it also outperformed proteinuria and GFR as a predictor of mortality,” Dr. Wolf said. “This finding gives a sense for just how powerful this risk factor is as a predictor of mortality.”
Not Ready for Clinic
Dr. Wolf cautioned that FGF23 probably is not ready for broad clinical application. “The only reason to do a diagnostic test is if it can influence the clinician to make a change in management plans, and that change either helps the patient feel better, live longer, or suffer fewer complications,” he said.
“Before FGF23 testing becomes more widely used in clinical practice, we need to show that making a treatment decision on the basis of an FGF23 test will improve outcomes for the patient. Then it would be ready to be used in clinical practice.”
There can be significant variability of FGF23 levels among healthy subjects, as well as among patients with CKD, Dr. Wolf said. However, normal subjects seldom exhibit the very high levels of the biomarker that are seen in CKD.
“Interestingly, there is quite a bit of stability of FGF23 within individual patients,” he said. “We have observed high correlation in the FGF23 levels in patients who have had two visits in our studies, separated by up to six months,” he added, drawing upon unpublished data from other ongoing studies.
Dr. Wolf said he is not aware of any longitudinal study that has measured decline in renal function and changes in FGF23 in individual patients. CRIC aims to fill that gap with a subset of patients who are being evaluated annually.
“It will tell us whether or not one can more accurately assess an individual's risk using serial measurements versus just one,” he said.
Levels of vitamin D were available in 333 participants at baseline and 1,159 participants at the Year 1 follow-up visit, according to the study's online supplemental material.
“Adjusting for those didn't alter the results for FGF23,” Dr. Wolf said. “If elevated FGF23 is lowering vitamin D levels, could that be one of the mechanisms of the adverse outcomes? We don't know that yet.”
This is an important research paper, said Bryan Kestenbaum, MD, MS, who participated in the peer-review process of the paper prior to its publication. Dr. Kestenbaum is Associate Professor in the Division of Nephrology and the Kidney Research Institute at the University of Washington.
“There have been other papers in CKD showing small-scale associations of FGF23 with clinical disease—coronary calcium, left ventricular mass, progression of chronic kidney disease,” he said. “Here it was done on a much broader scale and in a much more general population, and with the hard outcome of mortality.”
It is “way too early” to consider using FGF23 as a biomarker for clinical practice, Dr. Kestenbaum added. “For a marker to be used in clinical practice, it has to be modifiable with some kind of therapy; it has to reliably predict future events.
“We don't have any medications that can reliably alter FGF23 levels; there is conflicting evidence about phosphate binders,” he said. “And we certainly don't have any data that acting on FGF23 levels will change the risk profile.”
Yet as a research tool, FGF23 is an interesting biomarker that appears to be able to tell years ahead of time the risk of future events.
“We have to put more effort into understanding how FGF23 is associated with these adverse outcomes,” Dr. Kestenbaum said. “What are the mechanisms? How does it influence risk? What tissues is it acting on? Essentially, what is it a marker of?”
Some researchers have suggested that FGF23 is “a high-power marker of phosphate balance … it can give a flavor of hyperphosphatemia over a long period of time,” he noted.
“If FGF-23 is truly a marker of aberrant phosphorus metabolism, the paper underscores the clinical relevance of impaired phosphorous metabolism in chronic kidney disease prior to starting dialysis.”
Questions of Causality
FGF23 itself may be the harmful agent. The current understanding is that FGF23 acts through the membrane protein Klotho, which is only present in the kidney, brain, and parathyroid glands, Dr. Kestenbaum said. “So, if this is having an adverse effect on other organs, it may be doing it with a Klotho-independent mechanism that we don't yet understand.”
L. Darryl Quarles, MD, Director of the Division of Nephrology and Associate Dean for Research at the University of Tennessee Health Science Center, also focused on potential mechanisms of action.
“If the association is causative, then we need to understand the molecular pathway behind its effect on mortality and progression of renal disease and develop therapeutic strategies that prevent elevation of FGF23 in CKD and ESRD patients,” he said.
“But if it is simply a uremic marker, then interventions to prevent FGF23 from increasing would have no effect on mortality or progression of renal failure.”
Even if FGF23 plays no causal role in mortality and progression of renal disease, it does have a role in regulating mineral metabolism. One possible hypothesis, according to Dr. Quarles, is that the reduction in 1,25-dihydroxivitamin D and 25-hydroxivitamin D seen in CKD “may be purposeful, caused by FGF23 stimulation of Cyp24-mediated degradation of vitamin D analogues, which limits vitamin D-dependent gastrointestinal phosphate absorption as well as secondary hyperparathyroidism, thereby helping to maintain phosphate homeostasis in the face of declining ability of a diseased kidney to excrete phosphate.
“If true, these concepts will fundamentally change how we view the pathogenesis of secondary hyperparathyroidism and how we treat this disorder,” he said. “If the body is purposefully reducing active vitamin D production, then giving vitamin D replacement therapy will further stimulate FGF23 and lead to potential toxic effects.
“Future approaches may focus on correcting the primary events—such as phosphate retention or other factors stimulating FGF23 release—and less on vitamin D supplementation and treatment with active vitamin D analogues.”