Phosphate binders in CKD: bad news or good news?

Hyperphosphatemia has long been considered an important contributor to the mineral and bone disorder associated with CKD (CKD-MBD). Already 50 years ago, Slatopolsky et al. showed that as GFR decreased, fractional phosphate excretion rose because of inhibition of tubular phosphate reabsorption by parathyroid hormone (PTH), thus preventing serum phosphorus to rise. In the last 10 years, fibroblast growth factor 23 (FGF23) has been progressively recognized as another major regulator, if not the most important factor, in controlling renal phosphate excretion and avoiding phosphate retention in CKD. FGF23 exerts this action by activating its main receptor, FGFR-1, in the proximal tubular epithelium, with Klotho as an obligatory coreceptor. These adaptive mechanisms allow serum phosphorus to stay normal or near normal until CKD stages 4–5. The contribution of hyperphosphatemia to the pathogenesis of secondary hyperparathyroidism and hence osteitis fibrosa in CKD is well established. Moreover, hyperphosphatemia favors the development of soft tissue calcifications, including vascular calcification. Most importantly, a significant association has been identified between high serum phosphorus levels and mortality, both in CKD patients and in the general population. The associations may be both direct and indirect, through concomitant changes in circulating hormone levels such as PTH and FGF23. It must be pointed out, however, there is no high-quality evidence to date based on randomized controlled trials (RCTs) that normalizing serum phosphorus improves hard patient outcomes. Available evidence is only circumstantial, mostly based on association studies. The recent Kidney Disease— Improving Global Outcomes guideline on CKD-MBD suggests to reduce elevated phosphorus levels toward the normal range in CKD stage 5D and to keep serum phosphorus normal in CKD stages 3–5. Note, these are suggestions based on weak evidence. Although both extremely high and extremely low serum phosphorus levels clearly are associated with major complications, there is an intermediate gray zone with optimal serum phosphorus targets difficult to define for the different stages of CKD. Several means are available to reduce high serum phosphorus levels. They include dietary phosphate restriction, use of phosphate binders, and phosphate removal by effective dialysis in patients with ESRD. In general, limiting oral phosphate intake cannot be achieved effectively without reducing protein intake. This may be useful in CKD stages 3–5, but not necessarily in CKD stage 5D, where it could even do more harm than benefit. In any case, efficient long-term dietary phosphate restriction proves to be difficult in the majority of patients. In those on renal replacement therapy, the use of high-efficiency dialysis procedures alone allows optimal hyperphosphatemia control. However, most patients prefer standard dialysis regimens with which phosphate control is generally insufficient. The prescription of oral phosphate binders therefore remains the principal therapeutic approach. The degree of reduction of hyperphosphatemia that can be achieved with the various types of available binders is comparable. It appears to depend more on patient acceptance and tolerance than on differences in phosphate-binding capacity. It is fair to say that in patients with CKD stage 5D, it is often difficult, if not impossible, to achieve normal or near normal serum phosphorus values. All this brings us back to the issue of optimal hyperphosphatemia control. Is it useful at all to take phosphate binders? Are some types of binders superior to others in terms of patient outcomes? The authors of two recent observational studies examined the question of whether the prescription of phosphate binders compared with no phosphate binder prescription led to better patient outcomes. Unfortunately, we are left with opposite conclusions. Isakova et al. compared patients who began treatment with phosphate binders during the first 90 days after initiating hemodialysis with those who did not receive binders during that period. They found, using different types of analyses, that treatment with phosphate binders was independently associated with decreasedmortality risk compared with no treatment. However, Winkelmayer et al., who compared use versus nonuse of calcium-containing phosphate binders in incident dialysis patients, did not observe any mortality differences, using multivariate and propensity score-matched Cox regression models. Several RCTs were undertaken to identify possible superiority of one type of phosphate binder to another in terms of Published online ahead of print. Publication date available at www.jasn.org.