The treatment of hyperphosphataemia in CKD: calcium-based or calcium-free phosphate binders?

The recent publication of the novel KDIGO™ chronic kidney disease (CKD) guidelines in 2009 [1] clearly underlined the importance of bone and vascular health for patients with CKD. Our understanding is rapidly increasing of how closely disturbances in mineral metabolism are linked to bone and vascular disease (the so-called bone–vascular axis). A decade ago, the first published studies on this deleterious interplay focused upon dialysis patients [2,3]. However, it has recently become evident that patients in the early stages of CKD also exhibit profound pathological changes in the bone–vascular axis, termed CKD–mineral bone disorder (CKD–MBD) [4]. CKD–MBD is a descriptive term that encompasses serum biochemistry changes such as hyperphosphataemia or hyperparathyroidism, renal osteodystrophy, and the problem of vascular calcification [5]. Our current understanding of these disorders has switched the focus of interest towards patients with earlier forms of CKD, as prevention of occurrence (primary prevention) might be easier than prevention of progression. In the light of overall rather disappointing results of interventional trials in patients with end-stage renal disease (ESRD), the idea that primary prophylaxis may be promising is fascinating. This concept is supported by the data from lipid modification trials indicating that statin therapy in ESRD is not effective and probably starts too late [6]. In contrast, statin therapy in patients with moderate CKD lowers cardiovascular risk comparable to patients with normal renal function. The data from the Pravastatin Pooling Project (PPP) demonstrated that pravastatin significantly reduced the incidence of cardiovascular events in people with or at risk for coronary disease and concomitant moderate CKD [7]. In this respect, we are awaiting the SHARP trial results. In line with these data, in patients with moderately reduced kidney function, ‘subclinical’ or modestly elevated serum phosphorus levels may pose an appreciable threat to patients [8,9]. In a recent issue of the New England Journal of Medicine, we found of interest the review article by Tonelli et al. [10] on the use of oral phosphate binders in CKD patients undergoing dialysis. The revision of the literature in the field appears to be complete. However, we would like to discuss critically the authors' conclusions on considering calcium-based phosphate binders to be the first-line therapy for dialysis patients. In contrast to the strict focus on available evidence-based medicine data as chosen by Tonelli et al., we suggest a more integrative and translational approach to this crucial question. In our opinion, several issues have not been sufficiently appreciated by Tonelli et al. [10]: for example, the data regarding occurrence of hypercalcaemic episodes, the development of vascular calcifications as a surrogate for cardiovascular disease [11], and the deleterious interaction of inflammation, dyslipidaemia and vascular disease. Although prolonged hypercalcaemia is not a common event in dialysis patients [12], any serum calcium levels below overt hypercalcaemia do not necessarily mean a well-balanced calcium metabolism given the poor reflection of serum spot measurements with overall total body calcium and with direction of calcium flux. In our opinion, several issues have not been sufficiently appreciated by Tonelli et al. [10]: for example, the data regarding occurrence of hypercalcaemic episodes, the development of vascular calcifications as a surrogate for cardiovascular disease [11], and the interaction of inflammation, dyslipidaemia and vascular disease. Serum phosphate levels in uraemic patients can be controlled by the reduction of oral phosphate intake, an adequate dialysis schedule and the use of intestinal phosphate binders. During homeostasis, ~ 60% of the ingested phosphate is absorbed via the gut. This amount is increased by (active) vitamin D. Strict dietary counselling may reduce daily oral intake of phosphate. However, following daily dietary protein recommendations of 1000–1200 g/kg/body weight, CKD patients ingest a consistent amount of at least 900–1000 mg phosphate per day [13]. It is a common observation that rigid dietary phosphorus restriction is rarely followed by most patients in the long term. Moreover, the phosphate content of food may vary considerably according to industrial procedures [14]. Considering that standard intermittent haemodialysis, as currently performed thrice weekly for 4–5 h, is usually unable to maintain phosphorus balance at acceptable levels, the use of phosphate binders appears to be an essential treatment to control phosphate overload. Isakova et al. [15] investigated in a prospective cohort of incident haemodialysis patients if treatment with any type of phosphate binder versus no phosphate binder was associated with improved outcome. And indeed, such a survival benefit was actually detected and surprisingly present not only in patients with very high but also in patients with rather low levels of serum phosphate [15]. Approaching the question of optimal phosphate binder therapy (calcium-containing versus calcium-free), we would like to address the following issues that we feel are of outstanding importance in our everyday clinical practice: Sevelamer hydrochloride (SH) has been developed as the first phosphate binder that contains neither calcium nor aluminium [16]. The main study designed to evaluate the efficacy and safety of sevelamer was a prospective randomized controlled trial (TTG study), designed to compare the efficacy of sevelamer versus calcium carbonate on serum phosphate, calcium and calcium–phosphate product levels, and also to compare the progression of coronary and aortic calcification assessed by EBCT [17]. Serum calcium profile remained among normal values in both the sevelamer- and the calcium-treated group. However, hypercalcaemic episodes occurred less frequently in sevelamer-treated patients (incidence of hypercalcaemia: 16% in calcium carbonate and 5% in the sevelamer group, P = 0.04). Although experimentally yet to be proven, intermittent peaks of circulating calcium, i.e. hypercalcaemic episodes, might potentially cause even more harm than steady-state calcium levels leading to the same cumulative mineral burden [18]. Additionally, calcium carbonate administration was found to be associated with a higher median percent increase in coronary (25% vs 6%, P = 0.02) and aortic (28% vs 5%, P = 0.02) calcium scores compared with sevelamer. In a secondary post hoc analysis of the same study, Raggi et al. [19] showed that patients who received calcium salts showed a significant reduction in thoracic vertebral bone density along with an increased coronary calcification score. Therefore, calcium-based phosphate binders may enhance progression of vascular calcification both directly and indirectly: by increasing systemic calcium load and by reducing the calcium storage in bone [18]. In contrast, two other prospective studies deny any advantages of calcium-free phosphate binding at first sight (the BRIC [20] and the CARE-2 trial [21]). However, these two trials are rather incomparable to the TTG trial due to significant differences in study design and patient characteristics: in the CARE-2 trial, a high pre-existing vascular calcification and vascular risk factor burden were present, and in the BRIC study, the bone disease induced by liberal aluminium usage may have blunted any beneficial effects of calcium-free phosphate binding [20,21]. Moreover, in the BRIC study, the calcium concentration in the dialysate was reduced to 2.5 mEq/L in all those patients with low bone turnover at baseline (54% of patients in the sevelamer group and 63% in the calcium group, respectively), having a potential positive effect on the vascular calcification progression in both treatment arms [20]. We would like to come back to the aforementioned issue of primary prevention. One prospective study in CKD patients not yet on dialysis with mild levels of coronary artery calcification (CAC) at baseline also showed beneficial results of calcium-free phosphate binders [22]. In these patients, only sevelamer prevented CAC progression, while both low-phosphate diet alone and administration of calcium carbonate were associated with CAC progression. In a cohort of 127 incident haemodialysis patients, Block et al. [23] confirmed the protective effect of sevelamer on coronary calcifications. Moreover, the authors demonstrated that treatment with sevelamer was associated with a significant survival benefit when compared with calcium-based phosphate binders. A large prospective randomized survival study, the DCOR trial, was undertaken in order to test if finally sevelamer administration was associated with a survival benefit in ESRD patients [24]. The DCOR randomized 2103 patients on haemodialysis up to 45 months to either sevelamer hydrochloride or calcium-based phosphate binder treatment. Both all-cause and cause-specific mortality rates were not significantly different. A predetermined analysis revealed that in patients over 65 years of age, there was a significantly higher mortality rate in patients treated with calcium-based phosphate binders [24]. Furthermore, in a pre-scheduled secondary analysis of the DCOR trial, St Peter et al. [25] showed that treatment with sevelamer reduced all-cause hospitalizations compared with calcium-based binders. Recently, sevelamer carbonate, a modified and buffered form of sevelamer hydrochloride, has been studied in patients with chronic kidney disease [26] and in dialysis patients [27]. Serum bicarbonate levels only decreased with sevelamer hydrochloride, but both sevelamer carbonate and sevelamer hydrochloride demonstrated equal efficacy in controlling serum phosphorus. Therefore, sevelamer carbonate may have advantages over sevelamer hydrochloride in the treatment of hyperphosphataemia in both CKD and haemodialysis patients, where bicarbonate levels may decrease after switching from calcium-based binders to sevelamer hydrochloride [28]. Aiming at reduction of calcium supply, sevelamer is not the only alternative. In recent years, new clinical data have shown the efficacy and safety of lanthanum carbonate in controlling hyperphosphataemia in dialysis and CKD patients. Lanthanum carbonate has been widely investigated in pre-clinical and clinical studies, demonstrating a good efficacy in reducing serum phosphate levels without major side effects. In a recent long-term report up to 6 years, Hutchison et al. [29] demonstrated that lanthanum carbonate was not associated with any adverse changes in liver enzymes and bilirubin levels. Reduction of pill burden might be an important advantage of lanthanum carbonate treatment: a multicentre, open-label trial including 367 dialysis patients showed effective phosphate control in patients taking lanthanum carbonate most of the time with three tablets per day—much less than a mean of 7.6 tablets of phosphate binders per day taken before starting with lanthanum carbonate [30]. However, cardiovascular outcome data—e.g. prospective trials investigating the effect upon vascular calcification or cardiovascular morbidity and mortality—are missing for lanthanum carbonate. Dose-dependent efficacy of calcium-based phosphate binders is well affirmed, and these agents have been our standard treatment over years. However, even standard treatment calcium dosages administered under strict surveillance as in prospective trials and in the absence of any worrisome occurrence of hypercalcaemic episodes [17,20,21] have been associated with worsening of coronary artery calcium score. We have to underline at this point that any potential side effects of calcium overload result from a cumulative effect of calcium supply that includes oral intake (diet and medication) as well as calcium shifts during dialysis. A warning debate with respect to calcium toxicity is incomplete without mentioning potential threats from a positive calcium balance induced by dialysis. In this regard, we would like to underline data showing that a dialysate calcium concentration of 1.25 mmol/L is associated with a little net calcium flux [31]. As already discussed with the BRIC study data, lowering dialysate calcium is likely linked with restoration of bone metabolism in patients with adynamic bone disease [20]. Along with restoration of bone metabolism, low levels of bath calcium are presumably associated with lower levels or with reduced progression of vascular calcification [32,33]. However, since lowering calcium bath concentrations poses the risk of a more difficult hyperparathyroidism control, then again, oral phosphate binders come into play. Calcium-based phosphate binders represent a significant source of daily calcium supply, which we recommend to be below 1.5 g elemental calcium per day. However, even this amount may overwhelm the calcium-buffering capacity of bone in specific dialysis patients. In the review by Tonelli et al. [10], it is clearly affirmed that all of the available phosphate binders are effective in lowering serum phosphate. Similarly, for any of the available drugs, gastrointestinal adverse effects are a possible limitation and may promote non-adherence to medication. In consequence, if efficacy and tolerability are comparable, the choice of a particular drug should take additional effects into account. And undoubtedly, a main one could be the economical aspects, in which case selection of calcium-based products would be the obligatory choice. However, irrespective of tolerability, efficacy and economical aspects, we would like to promote a personalized and individualized approach regarding phosphate binder choice, in which especially co-morbidities are integrated into the decision process of which binder type to prefer. We feel that the review of Tonelli et al. has not paid enough consideration to the fact that dialysis patients are not a homogenous community. Nephrologists must not forget that hyperphosphataemia and secondary hyperparathyroidism are nowadays regarded as complex clinical syndromes, properly integrated into CKD–MBD to underline clearly the strict pathophysiological link existing between biochemical derangements, bone lesions and vascular calcifications [34]. As a consequence, when facing the challenge of trying to correct serum hyperphosphataemia, we have to consider the intricate relationships existing among these three clinical variables. Accordingly, we would like to present three examples where co-morbidities or co-medication clearly pushes the choice of binder towards avoidance of calcium-based binders: The first example is a patient with adynamic bone disease, who may also be diabetic and already affected with vascular calcifications (not a rare example in our daily clinical practice, in fact). Based on all available data, there is no doubt that any excessive amount of calcium supply will contribute to progression of vascular calcification (in general, calcified patients are considered to be prone to progression) [35]. Secondly, the same threat could be true for a patient presenting with signs of chronic inflammation [36]. Chronic inflammation predisposes the patient to have a pro-atherogenic milieu and at the same time to have significantly reduced plasma levels of serum fetuin-A. Fetuin-A is important for inhibition of tissue precipitation of calcium and phosphate [37–39]. Also, in this type of patient, any calcium and/or phosphate burden is expected to worsen vascular calcifications even faster than in an otherwise ‘healthy’ CKD patient. Thirdly, the application of Coumadin (vitamin K antagonists) for long-term oral anticoagulation appears to promote vascular calcification. The putative mechanism behind that association is the blockade of matrix-gla protein (MGP) activation via reduced post-translational gamma-carboxylation induced by Coumadin [40]. So, we consider a patient on Coumadin not to be a good candidate for any avoidable calcium application since an important anti-calcification protection barrier is malfunctioning. We could present several further examples of particular patients in which prescribing a calcium-based phosphate binder could raise similar hypothetical concerns. In our opinion, it is important to underline that the new syndrome, CKD–MBD, should be regarded as a complex web, where touching one point means having unpredicted changes in others. Accordingly, we should consider that available clinical data, even though mostly observational, suggest that some therapeutic strategies we have been following (e.g. unlimited calcium supply) are in fact associated with worse outcomes. As a consequence, it seems wise that when selecting a phosphate binder, a nephrologist should take into account the entire picture of CKD–MBD and not only focus on hyperphosphataemia. When deciding about optimal phosphate binder therapy, we propose to consider at least the following parameters as indicators for calcium avoidance: The list of additional concomitant conditions that may influence the individual prescription of phosphate binders is much longer (e.g. presence or absence of acidosis). Moreover, adherence is an important issue since in the daily clinical practice, it has been reported that phosphate binders represent up to half of the pills burned in dialysis patients [49]. Older age The prevalence of vascular calcification rises with age presumably because the growing skeleton is a calcium buffer [41]. Accordingly, the DCOR results suggest a survival benefit for elderly patients when treated with sevelamer. Male gender and post-menopausality Female renal patients are less frequently affected with coronary calcification [42,43], thus suggesting a possible protective role for sex hormones similar to non-CKD patients [44,45]. Diabetes The presence of diabetes is universally considered to be a risky condition for calcification. Low bone turnover The few available data indicate that having a low bone turnover is associated with increased calcifications [46]. Prevalent vascular and/or valvular calcification As suggested by KDIGO Clinical Practice Guideline for the Diagnosis, Evaluation, Prevention, and Treatment of CKD–MBD [47], baseline calcification is a strong risk factor for ongoing calcification. Inflammation As already discussed, affected patients may be at increased risk for calcification [48]. In summary, we are treating individuals, not protocol patients, who deserve an individualized approach. Moreover, as suggested by KDIGO clinical practice guidelines, the dose of calcium-based phosphate binders should be restricted in the presence of hypercalcaemia, vascular calcification, and/or adynamic bone disease, and/or if serum PTH levels are persistently low [47]. Recent publications have shown that there is a significant positive association between serum phosphorus levels in the normal range and the amount of vascular calcification [25,50]. In our opinion, we nephrologists have to learn that treating solely overt hyperphosphataemia underestimates the true danger of phosphorus burden. Many patients with CKD stages 4–5D have high serum phosphorus levels, and for them, the current KDIGO guidelines suggest maintaining serum phosphorus levels in the reference range [47]. However, we doubt that single-point measurements of serum phosphate are reliable indicators of phosphate metabolism. More sensitive markers for disturbances in phosphorus metabolism, such as phosphorus excretion in the urine or the serum FGF23 levels, may help us in estimating the true danger of silent hyperphosphataemia. In this context, it is of special interest that novel non-calcium-containing phosphate binders, such as sevelamer carbonate or lanthanum carbonate, are already approved for CKD patients not yet on dialysis [27,51,52]. And a very recent prospective trial in CKD patients showed that only sevelamer carbonate, but not calcium carbonate, could lower serum FGF23 levels [53]. However, disturbances of phosphate metabolism are not the only cardiovascular risk factor in patients with advanced or end-stage renal disease. At this point, the crucial question arises whether or not a multi-target approach towards reduction of cardiovascular risk factors might actually be more promising than a ‘one target–one hit’ approach. Indeed, many ‘one target–one hit’ studies have actually been of disappointing outcome (e.g. regarding anaemia management or statin treatment) in ESRD patients. Is there a million gadget all-in-one therapeutic device available covering at least some of the overwhelming cardiovascular risk factors in CKD patients? Well, certainly, the answer is no, but sevelamer might produce some enthusiasm. Originally developed as a phosphate binder, some additional modes of action beyond serum phosphate reduction have emerged that can be as or Sevelamer has been shown to influence phosphate metabolism, reduce FGF23 serum levels increase serum levels, reduce increase at least in some studies, decrease levels of protein even this large of beneficial of sevelamer are all not by phosphate may not the results of the DCOR trial, i.e. equal mortality data with the again, of patients may benefit from this individualized ‘one In the issue of the optimal phosphate binder choice a of evidence-based medicine for a particular type of binder do not mean that any for the use of non-calcium-containing phosphate binders are Although may not be to calcium-based binders in all patients, are probably beneficial for several among of interest has received in the by and has received in the by and and has received in the by and