Nocturnal but not short hours quotidian hemodialysis requires an elevated dialysate calcium concentration

Interest in quotidian (daily) hemodialysis (HD) is growing. Some advocate short-hours high-efficiency daily HD (SDH) and others long-hours slow-flow nocturnal HD (NH) while the patient is asleep, both being used 5 to 7 d/week. The London Daily/Nocturnal Hemodialysis Study was the first attempt to obtain data of SDH and NH that may be compared with conventional thrice weekly HD (CH). This was a 4-yr observational study designed to enter and follow 40 patients: 10 receiving SDH, 10 receiving NH, and 20 receiving CH. The CH patients were cohort control subjects matched for each SDH and NH patient by age, gender, comorbidity, and original dialysis modality (in-center, home, self-care, or satellite HD). All SDH and NH treatments were at home. Data collection to December 2001 was analyzed. Then enrollment had been completed and all patients had been followed for 15 mo, eight SDH plus six NH for 18 mo, seven SDH plus six NH for 21 mo, and seven SDH and five NH for 24 mo. This report gives data on calcium and phosphorus metabolism in these patients. All patients were initially dialyzed against a 1.25-mmol/L calcium bath. Predialysis serum calcium levels became lower in NH versus SDH patients by the first month and at 9 mo, were 2.67 +/- 0.25 mmol/L (M +/- SD) in SDH, 2.40 +/- 0.16 mmol/L in NH, and 2.52 +/- 0.21 mmol/L in CH (SDH versus NH, P = 0.038; SDH versus CH versus NH, NS). Predialysis phosphorus levels were better controlled by NH than by SDH or CH, and with NH, all phosphate binders were discontinued. By 12 mo, a rise in bone alkaline phosphatase was seen in NH patients (but not in SDH or CH patients), which peaked at 15 to 18 mo (NH 191 IU/L +/- 70; SDH 82 +/- 34; CH 80 +/- 36; P < 0.002) and similarly with intact parathyroid hormone (iPTH) levels (NH 159 pmol/L +/- 75; SDH 13.1 +/- 10; CH 18 +/- 18; P < 0.00001). Because of these changes, the dialysate calcium concentration was increased to 1.75 mmol/L for the NH patients. Postdialysis calcium then rose to 2.57 +/- 0.21, and alkaline phosphatase and iPTH normalized completely by 21 mo. These observations prompted mass balance studies that showed that a 1.25-mmol/L calcium dialysate was associated with a mean net calcium loss of 2.1 mmol/h of dialysis time, whereas 1.75-mmol/L calcium dialysate provides a net gain of 3.7 mmol/h. In addition, the mass balance studies showed that phosphate removal by NH (43.5 +/- 20.7 mmol) was significantly (P < 0.05) higher than by SHD (24.2 +/- 13.9 mmol) but not by CH (34.0 +/- 8.7 mmol) on a per-treatment basis. With the increased frequency of treatments provided by quotidian dialysis, the weekly phosphorus removal (261.2 +/- 124.2 mmol) by NH was significantly higher than by SDH (P = 0.014) and CH (P = 0.03). This allowed the discontinuation of P binders in the NH group, which in turn eliminated approximately 8 g elemental Ca/wk oral intake. This, together with a 4 g elemental Ca/wk dialysate loss induced by a 1.25-mmol/L Ca bath, explains the changes in Ca, alkaline phosphatase, and iPTH seen in the NH patients. The SDH patients have weekly dialysis times similar to CH and still require P binders and do not become Ca deficient using 1.25-mmol/L Ca dialysate. With NH but not SDH, an elevated dialysate Ca concentration is required.