Calcification in CKD: No Closer to the Cure

Vascular calcification is highly prevalent in patients with end-stage renal disease (ESRD) and may be an important mechanism linking kidney failure with excess cardiovascular risk. Approximately 65% of incident and 85% of prevalent adult hemodialysis patients have calcification of the coronary arteries, as detected by electron beam computed tomography.1, 2 Pioneering basic science work has significantly advanced understanding of the mechanisms underlying dystrophic soft tissue calcification that occurs in chronic kidney disease (CKD).3, 4, 5 These studies have demonstrated arterial calcification to be a highly active process, analogous to bone formation, regulated by a growing list of promoters and inhibitors that are altered by kidney failure.

While disturbances of normal mineral metabolism, particularly phosphate, are considered to be the primary cause of vascular calcification in ESRD patients, evidence suggests the possibility of an additional iatrogenic component: the administration of excessive dosages of calcium salts to control phosphate and parathyroid hormone levels.6, 7 ESRD represents a state of altered calcium homeostasis, in which skeletal mineralization is impaired and urinary calcium excretion absent.8 Without a natural calcium reservoir or a means to remove excess calcium from the body, ESRD patients may incur significant positive calcium balance due to the intake of calcium-based binders and exposure to high calcium concentrations in the dialysate. Clinical studies of ESRD patients have reported an association of higher oral calcium dosage with greater coronary artery calcification scores.6, 7 Three clinical trials have found lesser progression of coronary artery calcification in kidney disease patients treated with sevelamer, a non–calcium-based phosphate binder, compared to those treated with a calcium-based binder.1, 9, 10 While not definitive evidence of harm, these findings motivated the recommendation of an upper limit of 1,500 mg of prescribed daily elemental calcium for patients with ESRD.11

In addition to binding phosphorus in the gastrointestinal tract, sevelamer also binds to dietary cholesterol and subsequently lowers low-density lipoprotein cholesterol (LDL-C) levels.12 This non–mineral metabolism property of sevelamer motivated the clinical trial by Qunibi et al13 that appears in this issue of AJKD. The study hypothesis was that the relative protection of sevelamer on coronary calcification might have been due, in part, to its cholesterol-lowering effects, rather than procalcification properties of calcium-based binders. In the current study, 203 prevalent hemodialysis patients with known coronary calcification were randomly assigned to therapy with either calcium acetate or sevelamer as a phosphate binder, with the addition of atorvastatin to achieve similar target LDL-C levels.

The rationale for the study design is questionable because there is limited evidence that LDL-C plays an important role in coronary calcification in populations with or without CKD. LDL-C concentrations are unrelated to coronary calcification scores in prevalent hemodialysis patients, and are unrelated to the progression of preexisting coronary calcification in more than 2,800 participants in the Multi-Ethnic Study of Atherosclerosis (MESA) study.2, 14 Intensive lipid lowering with atorvastatin had no effect on 1-year progression of coronary calcium scores in 471 participants without CKD with prevalent coronary calcification, and the 4D (Die Deutsche Diabetes Dialyse) trial demonstrated no effect of intensive LDL-C lowering on cardiovascular events in hemodialysis patients, yet coronary calcification is hypothesized to be a key surrogate of cardiovascular outcomes in this setting.15, 16, 17 In the current study, calcification progressed relentlessly in both treatment groups despite the substantial reduction of LDL-C levels to less than 70 mg/dL.

Even if LDL-C were related to coronary calcification, it is difficult to interpret findings from a clinical trial that compares 1 treatment to 2 different treatments. Sevelamer and atorvastatin reduce cholesterol concentrations via completely different mechanisms, and statins possess many properties beyond their effects on LDL-C.18 The preferential addition of atorvastatin to 1 treatment arm for the sole purpose of equalizing LDL-C concentrations obscures the interpretation of any differences or similarities between the treatment groups in this study.

In the current study, coronary calcification scores increased by 52% in the calcium acetate group and 57% in the sevelamer group after 1 year (P = 0.9). While it is tempting to conclude a lack of difference between the treatment groups, many other factors may have accounted for the observed similarity, including (1) loss of nearly half of the study participants during follow-up, (2) nontrivial measurement error in coronary calcification scores when scores are high,19 (3) potential for partial or nonadherence to assigned therapies, (4) relatively short duration of follow-up, and (5) small number of participants. In fact, many common errors in clinical studies produce bias toward the null and potential for type II error, in which no difference between treatment groups is observed when in fact a true difference exists. Even in the setting of a very large clinical trial, a proven treatment standard with large effect size, and precisely measured end points, it remains difficult to prove noninferiority.20 It is not appropriate to apply the term noninferiority to this study, which is simply a small negative study with many possible reasons for the negative findings, not proof that the 2 treatments are equal. In all fairness, coronary calcification progressed substantially in both treatment groups at rates similar to the natural history of the disease; it is likely that a placebo group would have been statistically noninferior to both treatment groups under these study conditions.

These findings contrast those of 3 previous clinical trials that demonstrated lesser progression of coronary calcification with sevelamer compared to calcium binders. In the Treat-to-Goal study, 200 prevalent hemodialysis patients were randomly assigned to sevelamer versus a calcium-based phosphate binder.9 After 1 year, mean changes in coronary calcification scores were −46 Agatston units in the sevelamer group versus +151 Agatston units in the calcium group. The RIND (Renagel in New Dialysis) trial randomly assigned 129 incident hemodialysis patients to sevelamer or calcium for 18 months; mean increases in coronary calcification scores were +138 Agatston units in the sevelamer group versus +338 Agatston units in the calcium group.1 Finally, Russo et al assigned 90 predialysis patients with stages 3 to 5 CKD and no clinical diabetes or coronary heart disease to a low-phosphorus diet plus either sevelamer, calcium carbonate, or placebo for 24 months.21 Mean changes in coronary artery calcification scores were +178 Agatston units in the placebo group, +133 Agatston units in the calcium carbonate group, and +38 Agatston units in the sevelamer group. The current study by Qunibi et al most closely relates to the Treat-to-Goal study of prevalent dialysis patients; however, progression of calcification was substantially greater in the current study, possibly due to a higher proportion of diabetic patients. Control of calcium and phosphate concentrations was similar in the 2 studies, though parathyroid hormone was somewhat better controlled in Treat-to-Goal study. Disparate findings from these relatively small clinical trials emphasize the role of sampling variation and the variability of coronary artery calcium measurements.

There is currently no effective treatment available for vascular calcification. From bench to bedside, novel therapies are needed to inhibit aberrant smooth muscle mineralization and to enhance deposition of calcium and phosphate in the skeleton, where these minerals belong. In the meantime, the current study does little to change clinical practice standards or current national guidelines. Nephrologists should continue to control serum phosphorus levels while recognizing the potential for dystrophic calcification that may result from excessive calcium loading, as well as from other agents that can increase serum calcium and phosphate levels. Particular caution is advised for patients with known calcification, or those with calcification risk factors, such as diabetes and higher serum phosphate and calcium concentrations.