Background: Hemodialysis patients are at increased risk for progressive coronary artery calcification; however, the development and progression of this disease process in patients new to hemodialysis is unknown.

Method: One hundred and twenty-nine patients new to hemodialysis were randomized to receive calcium-containing phosphate binders or the non-calcium phosphate binder sevelamer hydrochloride. Subjects underwent electron beam computed tomography scanning (EBCT) at entry into the study and again at 6, 12, and 18 months.

Results: One hundred and nine patients underwent baseline and at least one additional assessment of coronary calcification. At baseline, 37 % of sevelamer-treated and 31 % of calcium-treated patients had no evidence of coronary calcification. No subject with a zero coronary artery calcium score (CACS) at baseline progressed to a CACS >30 over 18 months. Subjects with a CACS >30 at baseline showed progressive increases in CACS in both treatment arms (P < 0.05 for each time point in both groups). Subjects treated with calcium-containing phosphate binders showed more rapid and more severe increases in CACS when compared with those receiving sevelamer hydrochloride (P = 0.056 at 12 months, P = 0.01 at 18 months).

Conclusion: New hemodialysis patients with no evidence of coronary calcification showed little evidence of disease development over 18 months independent of phosphate binder therapy. However, subjects with evidence of at least mild coronary calcification had significant progression at 6, 12, and 18 months. Use of calcium-containing phosphate binders resulted in more rapid progression of coronary calcification than did use of sevelamer hydrochloride.

The RIND study [10] was one of the first studies to demonstrate the possible adverse effect of calcium-based binders on VC in a hemodialysis population. The results of the RIND study were similar to the earlier Treat To Goal study [11] in that both studies challenged the widespread use of calcium-based binders and suggested that they may adversely affect a well-validated surrogate endpoint, namely, VC. Furthermore, in a preplanned secondary endpoint analysis of the RIND study with a median follow-up of 44 months, patients randomized to sevelamer had a significantly lower mortality rate compared to those randomized to calcium-based binders [12]. However, it is important to recognize that RIND was simply not powered to detect mortality differences given its small sample size. Furthermore, a fundamental problem with both the RIND and Treat To Goal studies was a lack of a specified treatment protocol in the calcium- and sevelamer-based arms with use of different doses and types of calcium-based binders. This clearly undermines the utility of the data generated. However, both RIND and Treat to Goal were, at the time of publication, hypothesis-generating studies challenging the paradigm of phosphorus control based on the ubiquitous use of calcium-containing binders.

Background: Previous clinical trials showed that progression of coronary artery calcification (CAC) may be slower in hemodialysis patients treated with sevelamer than those treated with calcium-based phosphate binders. Because sevelamer decreases low-density lipoprotein cholesterol (LDL-C) levels, we hypothesized that intensive lowering of LDL-C levels with atorvastatin in hemodialysis patients treated with calcium acetate would result in CAC progression rates similar to those in sevelamer-treated patients.

Study Design: Randomized, controlled, open-label, non-inferiority trial with an upper bound for the non-inferiority margin of 1.8.

Setting and Participants: A total of 203 prevalent hemodialysis patients at 26 dialysis centers with serum phosphorus levels greater than 5.5 mg/dL, LDL-C levels greater than 80 mg/dL, and baseline CAC scores of 30–7,000 units assessed by means of electron beam computed tomography.

Interventions: 103 patients were randomly assigned to calcium acetate and 100 patients to sevelamer for 12 months to achieve phosphorus levels of 3.5–5.5 mg/dL. Atorvastatin was added to achieve serum LDL-C levels less than 70 mg/dL in both groups.

Outcomes and Measurements: The primary endpoint was change in CAC score assessed by means of electron beam computed tomography.

Limitations: Treatment assignment was not blinded. The 1.8 a priori margin is large, CAC is a surrogate outcome, duration of treatment was short, and dropout rate was high.

Conclusions: With intensive lowering of LDL-C levels for 1 year, hemodialysis patients treated with either calcium acetate or sevelamer experienced similar progression of CAC.

The original CARE study [13] was an 8-week, blinded RCT that randomized patients to sevelamer or calcium acetate. Unlike RIND or Treat to Goal, there was a specified protocol for the management of hyperphosphatemia in the study. CARE demonstrated that patients allocated to calcium acetate had significantly lower serum phosphorus levels than those allocated to sevelamer, though transient hypercalcemia occurred in 16.7 % of patients randomized to calcium acetate. The data from CARE suggested that calcium acetate provided better biochemical control of serum phosphorus as well as being more effective.

CARE-2 [14] was designed to control for the lipid-lowering effect of sevelamer by determining the risk of VC in patients on sevelamer and calcium acetate who had comparable lipid control. This was achieved by using atorvastatin to lower LDL cholesterol to less than 70 mg/dl in both groups. The headline data demonstrated a comparable increase in coronary artery calcification in both groups. The short duration of the study coupled with the high dropout rate means that data should be interpreted with caution. Furthermore, virtually all the patients in the calcium acetate arm received atorvastatin, whereas 21 % of patients in the sevelamer arm did not receive any atorvastatin. Notwithstanding these significant limitations, both the CARE studies provide an important counterbalance to the heavily pharma-supported narrative that cheaper, calcium-containing binders lead to adverse outcomes by aggravating VC. It is based on the hypothesis that VC is highly predictive of mortality in the dialysis population and that VC can be further aggravated by the use of calcium-based binders. This is perhaps an oversimplistic model of the pathogenesis of VC, which is not a result of passive deposition of calcium and phosphorus in the vessel wall but rather a complex process tightly regulated by inhibitors of VC such as fetuin A and matrix gla protein [15].

Elevated serum phosphorus and calcium are associated with arterial calcification and mortality in dialysis patients. Unlike calcium-based binders, sevelamer attenuates arterial calcification, but it is unknown whether sevelamer affects mortality or morbidity. In a multicenter, randomized, open-label, parallel design trial, we compared sevelamer and calcium-based binders on all-cause and cause-specific mortality (cardiovascular, infection, and other) in prevalent hemodialysis patients. A total of 2,103 patients were initially randomized to treatment, and 1,068 patients completed the study. All-cause mortality rates and cause-specific mortality rates were not significantly different. There was a significant age interaction on the treatment effect. Only in patients over 65 years of age was there a significant effect of sevelamer in lowering the mortality rate. There was a suggestion that sevelamer was associated with lower overall, but not cardiovascular-linked, mortality in older patients. We suggest that further research is needed to confirm these findings.

DCOR [16] in essence is a negative outcome study with no beneficial effect of sevelamer hydrochloride on mortality. Although there appeared to be a survival benefit in patients over the age of 65 years, the “real world” clinical significance of this is not clear. Furthermore, this benefit was not associated with a reduction in cardiovascular mortality, which is somewhat surprising given that the putative benefits of sevelamer hydrochloride have been postulated to be due to a reduction in VC. DCOR is a large, adequately powered study, with a hard primary outcome (mortality) and adequately balanced baseline variables between both groups. The flaws of this study reside in the unusually high dropout rate (49 %), the absence of intention to treat analysis, and the inherent biased nature of an open-label design. A preplanned intention to treat analysis was later published and again showed no beneficial effect of sevelamer on mortality [17]. The high dropout rate may reflect in part issues with the tolerability of sevelamer.

DCOR remains a landmark study in the management of hyperphosphatemia in dialysis patients and again challenged expert, opinion-based guidelines that had promoted the use of sevelamer on the basis of beneficial effects on surrogate endpoints such as VC.

DCOR highlights the difficulty in conducting outcome studies on mortality targeting only one variable (such as phosphorus) in complex, comorbid dialysis patients where the risk of death is likely to be multifactorial. One plausible explanation for the difference between DCOR and the previously positive outcomes seen in the RIND study is that DCOR was conducted in a prevalent dialysis population, and therefore many of the subjects will already have established VC, thereby attenuating any potential impact of sevelamer on VC progression. However, it is clear that from this large, well-powered study that there is no hard evidence to support the use of sevelamer over cheaper calcium-based binders in a prevalent hemodialysis population in the absence of hypercalcemia.

Background and Objectives: Dietary phosphorus overload and excessive calcium intake from calcium-containing phosphate binders promote coronary artery calcification (CAC) that may contribute to high mortality of dialysis patients. CAC has been found in patients in early stages of non-dialysis-dependent CKD. In this population, no study has evaluated the potential role of phosphorus binders on mortality. This study aimed to evaluate all-cause mortality as the primary endpoint in non-dialysis-dependent CKD patients randomized to different phosphate binders; secondary endpoints were dialysis inception and the composite endpoint of all-cause mortality and dialysis inception.

Design, Setting, Participants, and Measurements: This is a randomized, multicenter, non-blinded pilot study. Consecutive outpatients (n = 212; stage 3–4 CKD) were randomized to either sevelamer (n = 107) or calcium carbonate (n = 105). Phosphorus concentration was maintained between 2.7 and 4.6 mg/dl for patients with stage 3–4 CKD and between 3.5 and 5.5 mg/dl for patients with stage 5 CKD. The CAC score was assessed by computed tomography at study entry and after 6, 12, 18, and 24 months. All-cause mortality, dialysis inception, and the composite endpoint were recorded for up to 36 months.

Results: In patients randomized to sevelamer, all-cause mortality and the composite endpoint were lower; a non-significant trend was noted for dialysis inception.

Conclusions: Sevelamer provided benefits in all-cause mortality and in the composite endpoint of death or dialysis inception but not advantages in dialysis inception. Larger studies are needed to confirm these results.