Phosphate control in end-stage renal disease: barriers and opportunities

Abstract

Hyperphosphatemia is a nearly universal complication of end-stage renal disease that is widely recognized as one of the most important and most challenging clinical targets to meet in the care of dialysis patients. Left untreated, it can lead to bone pain, pruritus and worsening secondary hyperparathyroidism. Data from observational studies demonstrate that an elevated serum phosphorus level is an independent risk factor for mortality, and that treatment with phosphate binders is independently associated with improved survival. Experimental studies provide support for the epidemiologic findings: phosphate excess promotes vascular calcification, induces endothelial dysfunction and may contribute to other emerging chronic kidney disease-specific mechanisms of cardiovascular toxicity. On the basis of this evidence, clinical practice guidelines recommend specific targets for serum phosphorus levels in the dialysis population. The purpose of this review is to summarize common challenges in meeting these targets and to identify potential opportunities for improvement.

INTRODUCTION

Management of patients undergoing dialysis requires careful attention to multiple clinical parameters, which encompass dialysis access and adequacy, hypertension and volume control, anemia, nutrition and chronic kidney disease mineral bone disorder (CKD-MBD). Among these, hyperphosphatemia is increasingly recognized as an important problem. Indeed, existing guidelines recommend specific targets for serum phosphorus levels in the dialysis population [1, 2] despite lack of randomized clinical trials that have established efficacy of lowering serum phosphorus levels in improving clinical outcomes.

The basis for the growing recognition of the importance of hyperphosphatemia stems from epidemiologic and experimental data that link phosphate excess with adverse outcomes. Observational studies have consistently reported that an elevated serum phosphorus level is an independent risk factor for mortality in end-stage renal disease (ESRD) [3, 4]. Additionally, treatment with phosphate binders is independently associated with decreased mortality in dialysis compared with no treatment [5, 6]. Moreover, elevated extracellular phosphorus may directly stimulate vascular smooth-muscle cells to undergo phenotypic changes that predispose to vascular calcification [7, 8], which is one mechanism for the increased mortality in ESRD [9].

Left untreated, hyperphosphatemia can lead to pruritus, bone pain and worsening CKD-MBD. Uninterrupted stimulation of the parathyroid glands by elevated extracellular phosphorus concentrations, especially when accompanied by decreased extracellular ionized calcium concentrations, and markedly reduced serum calcitriol levels leads to increased parathyroid hormone (PTH) production [10]. This promotes diffuse polyclonal hyperplasia followed by monoclonal nodular hyperplasia [11]. The ensuing severe hyperparathyroidism may worsen hyperphosphatemia via phosphorus efflux from the skeleton. In some cases, PTH-induced skeletal changes may progress to form brown tumors, which are the product of destruction of local bone by rapid bone resorption, with hemorrhage and reparative granulation tissue replacing normal marrow. Phosphate excess has also been linked to endothelial dysfunction [12] and to elevated fibroblast growth factor 23 (FGF23), which contributes to left ventricular hypertrophy [13] and is an independent risk factor for mortality in ESRD [14].

In this review, we summarize common challenges in meeting the targets specified in the guidelines, while identifying potential opportunities for improvement.

BARRIERS IN ACHIEVING CONTROL

Recent publications have reported on difficulties in meeting the recommended targets for serum phosphorus levels in ESRD patients [15, 16]. Below, we identify potential barriers to implementation of the existing guidelines.

Phosphate handling in ESRD

Absence of renal phosphate excretion is a leading factor that impedes phosphate control in ESRD patients, especially among those without residual renal function (RRF). A typical Western diet is comprised of ∼1500 mg of dietary phosphorus per day, of which the gastrointestinal tract absorbs ∼800 mg per day. This absorption takes place in the small intestine via passive diffusion and active transport [17, 18]. The latter process occurs via a sodium-phosphate co-transporter (NPT)2b located on the luminal surface of enterocytes, and is upregulated by low dietary phosphorus intake and increased calcitriol [17, 18]. Importantly, recent animal data suggest that intestinal phosphate transport via NPT2b plays a greater role in total phosphate absorption than previously thought [19, 20], and clinical investigators have speculated that in ESRD patients taking phosphate binders a transient upregulation of NPT2b could blunt binder effectiveness [21].

Under normal circumstances, renal excretion is the primary mechanism of maintaining normal phosphorus homeostasis. CKD patients manifest a compensatory reduction in tubular phosphorus reabsorption mediated by increased levels of serum PTH, FGF23 and phosphorus. However, once the glomerular filtration rate drops below 30 mL/min/1.73 m², phosphate excretion declines to a point that hyperphosphatemia starts to develop [22, 23]. Based on conservative estimates, if an anuric patient consumes 900–1500 mg of phosphorus per day and after we factor in dietary phosphate absorption and the amount of phosphorus removed by routine dialysis (see below), an excess of about 300–500 mg of phosphate per day will need to be removed by phosphate binding [23]. Without additional interventions to address this difference, hyperphosphatemia will develop.

Phosphate removal by dialysis

Patients on dialysis are typically prescribed a diet that contains an average intake of phosphorus of about 900 mg/day (or 6300 mg per week), but in reality could be as high as 1500 mg/day (or 10 500 mg per week) [24]. Absorption of phosphorus is normally about 60% of that ingested, but could be as high as 80% in the presence of calcitriol or as low as 40% in the presence of phosphate binders [25]. Thus, assuming that 50% of the ingested phosphorus is absorbed, the weekly amount of phosphorus excess to be removed by dialysis could be >5000 mg. The actual phosphorus removal by the standard sessions of dialysis lasting 4 h each three times weekly can range from 600 to 1200 mg per treatment, or 1800–3600 mg per week. Nocturnal hemodialysis on average removes 600–1200 mg per session or 3000–8400 mg per week, peritoneal dialysis averages 300–360 mg of phosphate removal per day or 2100–2520 mg per week and short daily hemodialysis provides variable results (Table 1) [25]. Thus, the limited amounts of phosphorus removed by standard dialysis modalities hamper our ability to attain the target phosphorus levels in the ESRD population.

Table 1.

Effect of various dialysis modalities on phosphate removal

Dialysis modality	Schedule	Phosphate removal
Conventional hemodialysis	4–5 h three times a week	600–1200 mg/session 1800–3600 mg/week
Peritoneal dialysis	Continuous	300–360 mg/day 2100–2520 mg/week
Nocturnal hemodialysis	6–10 h, 5–7 nights per week	600–1200 mg/day 3000–8400 mg/week
Short daily dialysis	1.5–3 h, 5–7 days per week	Variable

Effects of binders

Another barrier is the limited ability of phosphate binders to bind phosphorus. There is a maximum limit of phosphorus that can be bound to a unit of binder, which sets a ceiling effect on binder potency. A recent literature review by Daugirdas et al. [26] established the phosphorus-binding capacities (PBC) of the different existing binder treatments. PBC refers to the in vivo phosphorus-binding ability of each binder and can be used to compare the potencies of different binder prescriptions. A typical daily phosphate binder regimen in the USA provides a mean binding capacity of ∼250 mg/day [24]. But as indicated earlier, ∼300–500 mg of absorbed dietary phosphorus will need to be bound daily by phosphate binders in a typical anuric ESRD patient consuming a Western diet and receiving standard three times weekly dialysis. Therefore, the limited binding capacity of existing phosphate binder regimens may hinder phosphate control in a large number of dialysis patients.

Effects of vitamin D

The use of vitamin D in the treatment of secondary hyperparathyroidism of ESRD may lead to an increase in dietary phosphorus absorption and thus creates another hurdle to overcome in achieving target phosphorus levels. Ramirez et al. [27] reported that the dietary phosphate absorbed in untreated, vitamin D deficient hemodialysis patients, after ingestion of a meal containing an average of 308 ± 3 mg of phosphorus, was 186 ± 35 mg (60% of intake). Following calcitriol treatment, the dietary phosphorus absorbed after ingesting 315 ± 4 mg of phosphorus was 272 ± 16 mg (80% of intake). In five normal matched controls, the average phosphorus absorbed after an average intake of 303 ± 4 was 242 ± 30 (80% of intake) [27]. Based on these and similar findings for active vitamin D analogs, current guidelines suggest decreasing or discontinuing the dose of active vitamin D in ESRD patients if the serum phosphorus levels continue to be above the target range [1, 2]. These recommendations highlight the complexity of managing hyperphosphatemia in ESRD patients who also have secondary hyperparathyroidism that is usually responsive to treatment with active vitamin D.

Adherence

Poor adherence to treatment regimen in the ESRD population is yet another barrier in reaching the target phosphorus levels. The high prevalence of side effects associated with binders, like diarrhea, nausea or chalky taste in the mouth, along with the high pill burden, and the variable individual meal patterns contribute to poor adherence. In a cross-sectional assessment of 233 patients from three dialysis units, Chiu et al. [28] reported a high pill burden in this patient population, with almost half of them taking >20 pills per day. This was implicated as a possible reason for a significantly lower adherence and higher serum phosphorus levels in these patients [29]. In a systematic review of the literature, Karamanidou et al. [30] found that nonadherence to phosphate-binding medication is a serious problem, with reported rates varying between 22 and 72%. Younger age, regimen complexity, patients' beliefs about the treatment and perceived social support may predict poor adherence [30]. Another important contributor could be the high prevalence of low health literacy in the dialysis population, which may make it more difficult for patients to comprehend the importance of complex regimens [31]. Hence, poor adherence further impedes the ability to achieve target phosphorus levels.

OPPORTUNITIES IN ACHIEVING CONTROL

Emerging data have pointed to potential pathways to improve phosphate control. We highlight some of these below.

Sources of dietary phosphate

Educating patients about sources of dietary phosphorus and how to make intelligent choices when selecting how and what to eat could become a very useful and effective tool in achieving phosphate control. It is known that the main source of phosphate is diet, but its bioavailability depends on factors like source (animal or plant derived), and type (organic or inorganic) (Table 2). Organic phosphorus is mainly found in protein-rich food groups, including animal and vegetarian sources of protein and dairy products, while inorganic phosphates are widely used as preservatives, color improvement components and flavor enhancers (Table 3) [32]. Organic phosphorus from dairy products, meat, poultry and fish is readily available as inorganic phosphorus after a series of hydrolytic processes [33], which is why diets rich in animal proteins have been linked to the development of hyperphosphatemia in CKD patients [34]. On the other hand, phosphorus derived from plants is found in the form of phytic acid or phytate [35], and because humans do not express the enzyme phytase, the bioavailability of this form of phosphorus is low [36]. This is supported by an interesting finding from an animal study [37] showing that urinary excretion of phosphorus, which is a commonly used surrogate marker of dietary phosphate absorption, was higher in a meat-based diet compared with a plant-based diet. Moe et al. [38] also addressed this particular point in CKD patients by comparing the effects of meat and vegetarian diets on phosphorus homeostasis, while keeping the caloric intake balanced. After 1 week of the dietary intervention in this crossover study, a vegetarian diet significantly decreased serum phosphorus levels [38].

Table 2.

Selection of high phosphate foods

Low bioavailability (10–30%)

Organic

Plant

Seeds

Beans

Legumes

Peas

Nuts

Bread

Almonds

Peanuts

Lentils

Chocolate

Moderate bioavailability (40–60%)

Organic

Animal

Dairy products

Meat

Poultry

Fish

Egg yolk

Egg white

High bioavailability (100%)

Inorganic

Additives

Chicken nuggets

Pot pies

Hot dogs

Bacon

Deli meats

Canned meat

Processed cheese

Instant pudding and sauces

Refrigerated bakery products

Muffins

Ready to eat cereal

Breakfast bars

Colas

Fruit juices

Nestea cool iced tea

Dry powdered beverages

Liquid nondairy creamer

Flavored milk

This table is reproduced from Houston et al. [57], with permission from Academic Press, Elsevier. © 2013 Elsevier Inc. All rights reserved.

Table 3.

Hidden phosphate additives on food labels

Phosphate additives
Phosphate salt	Alternate names	Function	Product
Disodium phosphate	Sodium phosphate; dibasic; DSP/A; disodium monohydrogen; orthophosphate; disodium monophosphate	Texturizer	Imitation cheese, buttermilk
Monosodium phosphate	Monosodium dihydrogen orthophosphate; sodium phosphate disbasic	Emulsifier	Sports drink, whole egg
Potassium tripolyphosphate	Pentapotassium triphosphate; KTPP	Moisture retention	Poultry products
Sodium acid pyrophosphate	SAPP; disodium dihydrogen pyrophosphate; acid sodium pyrophosphate	Color	French fries, instant mashed potatoes
Sodium hexametaphosphate	HMP; sodium polyphosphate; Graham's salt	Reduce purge, emulsifier	Frozen fish fillet
Sodium tripolyphosphate	Sodium triphosphate; pentasodium triphosphate; STPP; STP	Flavor enhancer	Instant noodles, reduced sodium meats
Tetrasodium pyrophosphate	Sodium pyrophosphate; tetrasodium diphosphate; sodium diphosphate; TSPP	Moisture retention	Sausage, deli meats
Trisodium triphosphate	Sodium phosphate; tribasic; TSP; TSPA; trisodium monophosphate	Antimicrobial	Cheese

This table is reproduced from Houston et al. [57], with permission from Academic Press, Elsevier. © 2013 Elsevier Inc. All rights reserved.

Complexity is added to the problem by food additives, which represent mostly inorganic phosphorus in the form of phosphate salts that are rapidly dissociated and readily absorbed in the intestinal tract. It is thought that up to 90% of the inorganic phosphorus contained in these salts is absorbed as opposed to 40–60% of the organic phosphorus present in unprocessed foods [39]. It has been estimated that additives contribute as much as 1000 mg/day of additional phosphorus in a typical American diet [40]. A recent randomized trial of 279 ESRD patients with hyperphosphatemia evaluated the impact of limiting phosphorus-containing food additives on serum phosphorus levels [36]. After 3 months of instructing patients how to read food labels and to select alternative items when they detected a phosphorus-based additive on the label, the mean decline in serum phosphorus concentration was 0.6 mg/dL greater in the intervention group than in the control group [95% confidence interval (95% CI) −1.0 to −0.1 mg/dL]. In summary, dietary choices have an important role in phosphorus homeostasis in ESRD patients. Dietary recommendations that encourage some substitution of animal protein with vegetarian sources and provide education on the abundance of phosphorus additives in processed foods could potentially facilitate phosphate control.

Effects of residual renal function

Recently, the role of RRF in phosphate control was evaluated in a cross-sectional study from Europe [41]. In their analysis, the authors found a direct relationship between RRF and the achievement of phosphorus treatment targets. Additionally, patients with RRF in this study had less overall use of phosphate binders, suggesting that the presence of RRF may also result in a better quality of life and reduced treatment cost. While preservation of RRF may not be a feasible opportunity for improving phosphate control in the hemodialysis setting, patients undergoing peritoneal dialysis may potentially benefit from such efforts.

Effects of intensive dialysis

As the removal of phosphate in the conventional three times a week schedule is limiting, the use of longer and more frequent hemodialysis sessions was proposed to address phosphate control. The Frequent Hemodialysis Network (FHN) Daily trial [42] looked at the effect of short daily sessions of hemodialysis (1.5–2.75 h) six times a week compared with the traditional three times a week schedule in a cohort of 245 participants. At 12 months, the group randomized to frequent hemodialysis (n = 125) showed a reduction of 0.46 mg/dL (95% CI 0.13–0.78 mg/dL) in the mean serum phosphorus as compared with the three times a week group (n = 120). There was also a significant reduction in phosphate binder dose [43]. At the same time, the FHN Nocturnal trial [44] randomized 87 patients to three times per week conventional hemodialysis or to nocturnal sessions 6–8 h of hemodialysis six times per week. At 12 months, the group of frequent nocturnal hemodialysis showed a reduction in mean serum phosphorus of 1.24 mg/dL (95% CI 0.68–1.79 mg/dL) compared with the conventional hemodialysis group. Furthermore, 73% of the participants on the frequent nocturnal hemodialysis group did not need to use phosphate binders at 1 year of follow-up [43]. Secondary analysis of both studies revealed that the reduction in phosphorus in the more frequent dialysis groups was achieved even taking into consideration the interdialytic interval before the sample was taken, suggesting that this represents a true reduction in the predialysis serum phosphorus. These data show that more frequent and more prolonged sessions of hemodialysis result in better control of hyperphosphatemia, the allowance of more liberal diets, and reduced requirements of phosphate binders; however, the optimal frequency and length of individual dialysis sessions remain to be determined.

Effects of dual binder therapy

Approximately 30–50% of ESRD patients remain hyperphosphatemic on a single binder therapy [45]. In a recent study of American hemodialysis patients, the authors proposed that combined binder regimens might address this problem [24]. They estimated that the amount of excess phosphorus could be as high as 750 mg/day, and that a regimen consisting of a single phosphate binder on average had a PBC of ∼256 mg/day. Thus, the majority of phosphate binder prescriptions were found to be inadequate and the gap between the required and the actual binder therapy could be narrowed by the use of two binders simultaneously. Indeed, patients using two binders were found to have a higher average PBC than patients using one binder (451 versus 236 mg/day, respectively) [24]. This important study provides valuable insight into achieving better phosphorus control with the judicious use of dual binder therapy.

Recent studies suggest that niacin, which reduces intestinal phosphate absorption by blocking NPT2b [46], may be another promising therapy in phosphate control. Based on data from two randomized placebo-controlled trials, Ix et al. [47] examined the effect of niacin versus placebo on phosphorus levels in 261 patients with Stage 3 CKD. During 24 weeks of follow-up, the change in serum phosphorus was −0.40 mg/dL in patients receiving niacin (95% CI −0.46 to −0.34), with no changes in the placebo group (mean 0.02 mg/dL; 95% CI −0.05 to 0.10). There was no clinically relevant effect on calcium levels. The maximum effects were noted at Weeks 8–12 and were sustained during the entire follow-up. This promising approach needs to be tested in prospective studies in dialysis patients.

Effects of cinacalcet

The advent of cinacalcet as an alternative or adjunctive therapeutic agent for secondary hyperparathyroidism introduced a potential opportunity to minimize the impact of active vitamin D on serum phosphorus levels in ESRD. A recent study of 1716 Japanese hemodialysis patients with secondary hyperparathyroidism confirmed this possibility [48]. The authors found that the combination of starting cinacalcet and increasing active vitamin D was associated with a significant decrease in PTH, whereas the combination of starting cinacalcet and decreasing active vitamin D was associated with simultaneously achieving target serum phosphorus, calcium and PTH levels. Thus, use of cinacalcet with active vitamin D may lead to better control of both serum phosphorus and PTH levels in some patients.

FIRST, DO NO HARM

A large body of observational and experimental data linking phosphate excess with adverse outcomes provides sufficient rationale to advocate adequate phosphate control. However, given lack of randomized clinical trials establishing efficacy of lowering serum phosphorus levels in improving clinical outcomes in dialysis, providers need to be mindful of the safety of the existing approaches to lower serum phosphorus and be attune to potential unintended consequences of guideline implementation. We highlight some examples below.

Dietary restriction and malnutrition

Given the strong correlation between dietary phosphorus and protein intake, aggressive dietary phosphate restriction as an attempt to control phosphorus may result in protein calorie malnutrition [49]. An epidemiologic study of more than 30 000 maintenance hemodialysis patients followed for 3 years showed that a low predialysis serum phosphorus level in conjunction with a reduced estimated protein intake was associated with an increased risk of death [50]. In this same study, patients whose estimated protein intake increased over time showed the greatest survival, especially when this was associated with a decline in serum phosphorus levels, probably achieved by means of phosphate binders and not dietary restriction. The authors concluded that controlling serum phosphorus levels by restricting dietary protein intake might put patients at risk of protein malnutrition and increased risk of mortality. Given these and similar findings, renal professionals should be mindful of the potential risks of protein over-restriction and could consider tailoring dietary recommendations to leverage the existence of foods with naturally high protein but low in phosphorus, such as egg whites [51].

Pill burden

Aggressive dual phosphate binder therapy may favor better phosphorus control in maintenance hemodialysis patients [24]. However, this approach may be harmful in some patients due to polypharmacy, early satiety, extra fluid intake and poor health-related quality of life (HR-QOL) [28]. A cross-sectional study in 2009 [28] evaluated the total daily pill burden in 233 prevalent maintenance hemodialysis patients. Phosphate binders were the single largest contributor to the pill burden, representing 49 ± 19% of all pills taken. Interestingly, the pill burden in these patients was very high; ∼50% of them were taking >20 pills daily, even more than patients with congestive heart failure (10–11 pills a day), and patients with HIV, whose pill burden is <20 in most regimens [28]. A higher pill burden was an independent predictor of lower HR-QOL scores [28]. In conclusion, while dual binder therapy may be beneficial for some patients, it could also have a negative impact on others.

Vascular calcification

While the efficacy of existing phosphate binders to reduce serum phosphorus levels is well established, their long-term safety profiles are less clear. Recently, Block et al. [21] looked at the effects of three phosphate binders versus placebo on parameters of mineral metabolism in 148 CKD patients. Surprisingly, active therapy significantly increased calcification of the coronary arteries and abdominal aorta. While the group randomized to receive calcium acetate seemed to have more pronounced effects on vascular calcification, similar signals were also observed in the noncalcium-based arms. These unexpected results certainly illustrate the need for additional studies in the predialysis setting, and also point to the need to be cautious with prescription of phosphate-binding agents in the ESRD population.

Hypocalcemia with cinacalcet

The reported incidence of hypocalcemia with cinacalcet use in the initial studies was <5% [52, 53]. The recent EVOLVE study [54], a clinical trial of 3883 ESRD patients with secondary hyperparathyroidism randomized to receive cinacalcet or placebo, found that the rate of hypocalcemia was 12.4% in the cinacalcet arm when compared with 1.7% in the placebo arm. Nephrologists prescribing cinacalcet to patients who are dialyzed on a low-calcium bath and receive noncalcium-based binders and little or no active vitamin D should be particularly mindful of this potentially serious complication.

CONCLUSION AND PERSPECTIVE

Nephrologists, nurses and dieticians caring for ESRD patients face a number of difficulties in achieving desired serum phosphorus levels. The barriers are not limited to patient-related factors, but are instead multidimensional, involving various aspects of care delivery, consequences of therapy and complex pathophysiology. Novel strategies that may begin to address some of these barriers include education of patients on the impact of food additives and vegetarian diets on phosphate control; utilization of phosphate binder combinations; addition of cinacalcet to CKD-MBD regimens; and, in selected patients, delivery of more frequent and /or prolonged dialysis. Sustained efforts in delivering multidisciplinary care that addresses health literacy gaps and maintains safety surveillance will likely also yield beneficial results.

The role of NPT2b in intestinal phosphate absorption has gained increasing recognition in recent years [18, 19]. Future research may yield novel therapeutic options to directly target this important mechanism of enterocyte phosphate uptake. In addition, the contribution of sodium phosphate channels, such as PIT1, to phosphate-induced vascular calcification is also becoming more apparent, and we are continuing to uncover the intricacies of this complex and multiorgan process [55]. For example, recent work has shown that aldosterone induces a spironolactone-sensitive, PIT1-depedent osteoinductive signaling cascade that contributes to vascular calcification [56]. Clearly, phosphate control has implications beyond the serum phosphorus level, and it is highly likely that we will continue to learn about additional systemic effects of this important clinical problem.

FUNDING

T.I. was supported by grant K23DK087858 from the National Institutes of Health.

CONFLICT OF INTEREST STATEMENT

T.I. has received honoraria from Genzyme and Shire.