Phosphate levels in patients treated with low-flux haemodialysis, pre-dilution haemofiltration and haemodiafiltration: post hoc analysis of a multicentre, randomized and controlled trial

Whether convective therapies allow better control of serum phosphate (P) is still undefined, and no data are available concerning on-line haemofiltration (HF). The objectives of the study are to evaluate the effect of convective treatments (CTs) on P levels in comparison with low-flux haemodialysis (HD) and to evaluate the correlates of serum phosphate in a post hoc analysis of a randomized clinical trial. This analysis was performed in the database of a multicentre, open label and randomized controlled study in which 146 chronic HD patients from 27 Italian centres were randomly assigned to HD (70 patients) or CTs: on-line pre-dilution HF (36 patients) or on-line pre-dilution haemodiafiltration (40 patients). CTs did not affect P (P = 0.526), calcium (Ca) (P = 0.849) and parathyroid hormone levels (P = 0.622). P levels were associated with the use of phosphate binders including aluminium-based phosphate binders (P < 0.001) and sevelamer (P < 0.001), pre-dialysis bicarbonate levels (P < 0.001) and pre-dialysis blood K levels (P < 0.001). On multivariate analysis (generalized linear model), serum P was again largely unassociated with CTs (P = 0.631). Notably, participating centres were by far the strongest independent correlate of serum P, explaining 45.3% of the variance of serum P over the trial and this association was confirmed at multivariate analysis. Bicarbonate (P < 0.001) and, to a weaker extent, serum K (P = 0.032) were independently related to serum P. In comparison with low-flux HD, CTs did not significantly affect serum P levels. Participating centres were the main source of P variability during the trial followed by treatment with phosphate binders, serum bicarbonate and, to a weak extent, serum potassium levels (ClinicalTrials.gov Identifier: NCT011583309).