Pro/Con debate: the calculation on calcium balance in dialysis lowers the dialysate calcium concentrations (pro part)

The equality in Equation (1) can only be achieved through reliable estimation of CaAbs and prescription of equivalent total Ca removal by selection of an appropriate CdiCa++. A kinetic model of Ca absorption as a function of Ca intake and calcitriol concentration [6] is depicted in Figure 1. The model was derived from a reported series of single meal studies of CaAbs over a Ca intake range 200–1250 mg and calcitriol 8–80 pg/mL [7–10], and in Figure 1 it is extrapolated to a Ca intake of 3000 mg/day that might be expected with very high doses of Ca-based phosphorus (P) binders. It can be seen in Figure 1 that modelled CaAbs conforms to a family of logarithmic curves suggesting saturation of calcitriol-dependent absorption as Ca intake increases. The absorption curves for normal levels of calcitriol (20– 40 pg/mL) would be seen in normal subjects and in HD patients treated with relatively low doses of Vit D analogues—calcitriol, 0.5–0.75 μg/dialysis, and Hecterol and Zemplar, 3–5 μg/dialysis. Ca absorption was recently reported with Ca intake distributed over 3 meals per day as diet, Ca alone (1425 mg Ca), diet plus Sevelamer (1340 mg Ca) and diet plus CaCO3 (3040 mg Ca total) [11]. The amount of CaAbs was determined from measurement of daily 24-h urinary calcium excretion over 1 week in 12 healthy normal subjects. The results of this study are also plotted in Figure 1 where they can be seen to fit the curves modelled for normal calcitriol concentration perfectly out to 3000 mg/day intake and provide strong evidence for the validity of the CaAbs model. A CaAbs curve calculated from data reported by Kopple and Coburn in 1971 [1] is also plotted in Figure 1 and illustrates the striking increase in CaAbs over the past 30 years from 0 mg/day to ∼200 mg/day with an intake of 1500 mg/day and low doses of Vit D3 analogues. Ca intake and absorption values calculated from Heinrich data are listed in Figure 2 where the removal required per dialysis is also calculated and shown to be quite substantial for both sevelamer (500 mg/dialysis) and CaCO3 (650 mg/ dialysis). A kinetic model of Ca absorption as a function of serum calcitriol and Ca intake. Summary Ca balances reported by Heinrich et al. [11]. Note that with both sevelamer and CaCO3, there is substantial Ca absorption which would require a removal of 500–600 mg of Ca during each three times weekly haemodialysis in order to achieve neutral Ca balance. where [ΔCpCa++·VECW] is the product of total change in plasma Ca++ during dialysis multiplied by extracellular fluid volume (VECW) estimated as 1/3 of VU. Note that KMP can vary between 1.00 (ΔCpCa++·VECW = 0) and 0 (MCa = 0). We have observed an average KMP = 0.70 in calcium modelling studies to date. Panel A in Figure 3 depicts an in vivo plot of Equation (1) derived from 26 sets of serial measurements (unpublished results of ongoing studies) of CdiCa++ and CdoCa++ (every 10 min) and three serial CpiCa++ values during 3.5–4-h high flux dialyses at blood flow 400, dialysate flow 500 and average total ultrafiltrate 3.0 L with polysulfone (Optiflux 180, FMC) dialysers. From the regression line, it can be seen that removal of 600 mg of Ca predicts a required mean inlet Ca++ concentration gradient of −0.9 mEq/L. Panel B in Figure 3 depicts the predicted change in plasma Ca++ during dialysis, ΔCpiCa++, as a function of the total amount removed and KMP. Note that very little change in CpiCa++ is predicted for JdCaT ≤ 200 mg that is the region of primarily convective removal. As JdCa increases, CpiCa++ is predicted to fall linearly with slope determined by KMP. With an average KMP of 0.70, the predicted fall in CpiCa++ would be 0.60 mEq/L with a removal of 600 mg Ca. Intradialytic Ca flux modelled for a 3.5-h dialysis with total Qf 2.5 L and DCa++ = 0.250 L/min. See the text for further discussion. We recently performed a Ca kinetic modelling (CKM) survey to calculate CaAbs on 320 Fresenius Medical Care (FMC) patients in three dialysis facilities using the most recent monthly treatment data including pre- and postdialysis CpiCa++, serum P and UKM measurements. The kinetic analysis showed median CaAbs = 150 mg/day and required removal 350 mg/dialysis (substantially less than shown in Figure 2 because of lower dietary Ca intake). We then modelled the CdiCa++ which would result in neutral Ca mass balance in those patients, and the cumulative frequency distribution of the results is shown in Panel C of Figure 2. The median modelled CdiCa++ was 2.2 mEq/L and ranged from 1.5 to 3.0 in individual patients as a function of CpiCa++, CaAbs and KMP in individual patients. The actual CdiCa++ prescribed in these patients was 2.50 in 80% and 2.25 in 20%, so in many, the kinetic analysis suggested substantial positive Ca balance. Targeting diffusive removal of Ca during dialysis will result in some fall in CpiCa++ as quantified above and may carry a theoretical risk of increased arrhythmias and or stimulation of increased PTH secretion during dialysis. The corrected QT interval (QTc) in the electrocardiogram has been reported to increase with CdiCa++ <3.0 resulting in a suggestion that CdiCa++ <3.0 should be avoided [14,15] even though abnormal QTc's and arrhythmias were not observed in the studies. These two studies in fact resulted in mild hypercalcaemia with these levels of CdiCa++ and the reduction of QTc observed might more appropriately be considered a side effect of hypercalcaemia. In view of the very high incidence of vascular calcification and substantial interdialytic CaAbs with modern dialysis therapy, it is not rational to prescribe positive Ca balance during dialysis by using high CdiCa++. It is possible that an inappropriate increase in PTH secretion might occur with regularly prescribed negative Ca balance during dialysis and require concomitant therapy with a calcimimetic agent. These theoretical issues must be evaluated in the context of the theoretical risk associated with chronic Ca and P overload with appropriate prospective studies. Finally, all studies of management of mineral metabolism in HD patients should include consideration of the effects of the total therapeutic regimen on Ca mass balance. None of the many studies of Ca-based versus non-Ca-based binders have included the standardization of prescribed CdiCa++ and a dose of Vit D3 analogues which have such powerful effects on Ca mass balance [16]. Conflict of interest statement. Author is a consultant at Fresenius Medical Care.