**5. Development of AFB-K**

One of the first studies on potassium profiled haemodialysis was published by Readelli (Readelli et al., 1996). This multicenter, randomised, cross-over, prospective study evaluated 42 patients with a known increase in Premature Ventricular Complexes (PVC) during dialysis (Holter-ECG recording). The author compared a haemodialysis model with constant K+ in the dialysate versus a model with decreasing K+ that maintained a constant K+ gradient between the blood and dialysate. A significant reduction of over 30% in PVC was recorded (both in terms of the number of PVC/h and as regards the PVC couplets/h) when performing potassium profiled dialysis. This PVC decrease is observed particularly in the first hour of dialysis, which corresponds to the greatest reduction in blood K+ values in the case of dialysis with constant K+. The result is obtained without affecting the K+ levels compared with conventional constant K+ haemodialysis.

Subsequently, Santoro (Santoro et al., 2002) published a study including 125 patients from a single centre in which dialysis with two techniques – one comprising AFB with constant K+ and another with potassium profiled dialysate (AFB-K) – resulted in slower K+ reduction with the latter technique (serum K+ being greater in the first hour of dialysis with the potassium profiling procedure). Both techniques reached the same final serum K+ levels. This study recorded a reduction in the number of intra-dialysis arrhythmias, particularly in those patients with a significant presence of ventricular ectopic beats. In conclusion, the safety and easy application of the AFT-K technique was confirmed, as well as the possibility of adapting the profile to each individual patient. In addition, this treatment was shown to reduce the impact upon repolarisation homogeneity and thus prevent cardiac arrhythmia.

Buemi et al. (Buemi et al., 2005) studied a series of 28 patients undergoing haemodiafiltration with constant K+ (HDF), compared with HDF involving variable K (HDFK). The authors examined the ECG tracings at different times during dialysis to measure the QT-interval, and analysed serum K, intra-erythrocyte K and the electrical membrane potential at rest. Intra-erythrocyte K remained constant. The electrical membrane potential at rest decreased over time, though the reduction proved significantly greater in HDF versus HDFK. In turn, the ECG tracing showed a reduction in QT dispersion and in the QT-interval with HDFK. The authors concluded that the risk of arrhythmia may be lower when using haemodialysis with variable K.

Subsequently, Santoro (Santoro et al., 2008) published a multicenter, randomised, crossover, prospective study in which 30 patients susceptible to arrhythmia were dialysed in a cross-over design with AFB involving constant K+ (2.5 mEq/L) or AFB with decreasing K in the dialysate bath (AFB-K). A tendency towards fewer arrhythmias is observed with AFB-K that becomes significant in the case of patients sensitive to dialysis. A reduction in the number of PVC per hour is observed both during dialysis and hours after the end of dialysis (with a maximum reduction in the number of PVC 14 hours after dialysis). In the first hour of treatment the serum K+ levels of the patients were lower in AFB than with AFB-K. The study concluded that the slow clearance of K+ may exert a protective effect.
