From cold dialysis to isothermic dialysis: a twenty-five year voyage.

Twenty-five years have passed since our group described for the first time the role played by Temperature (T) in cardiovascular stability. ‘Cold’ treatments prevented the hypotension induced by ‘warm’ treatments, whether in haemodialysis (HD) or in isolated ultrafiltration [1] or in haemofiltration [2,3]. Though some notes of caution were sounded [4], all researchers studying the problem in the years that followed confirmed the role played by T in short-term [5–9] as well as in long-term studies [10,11]. As compared to standard HD or ‘warm’ HD, that is with dialysate T of 37–37.58C, ‘cold’ HD, that is with dialysate T of 35–35.58C, ensures better cardiovascular stability. In 1997, reduction of dialysate T was recommended by the DOQI Guidelines as a means to prevent intradialytic hypotension [12]. A systematic review recently published found that ‘intradialytic hypotension occurred 7.1 (95% CI, 5.3–8.9) times less frequently with cool-temperature dialysis. A total of 22 studies comprising 408 patients were included, all studies were of crossover design and relatively short duration’ [13]. Lastly, the European Best Practice Guideliness on cardiovascular instability, announced at the 2006 ERA-EDTA Congress, scored only cold dialysis with evidence level I, among the different dialysis techniques usually adopted to prevent intradialytic hypotension. The terms ‘warm’ and ‘cold’ HD, though immediately understandable, are in reality too simplistic to describe the complex interrelations between the thermal profile of the dialytic treatment and its impact on the patient’s body T and cardiocirculatory function. Since 1984 a good correlation (r1⁄4 0.75) between variations in dialysate T in the 34–388C range and concomitant changes in patients’ body T has been found [14]. However, in standard HD with a dialysate T of 378C no thermal exchanges, on average, took place in the extracorporeal circuit, in that the T of the blood in the venous line was equal to or slightly lower than that of the arterial line [14]. Despite this thermoneutral behaviour however, an average rise in patient’s body T of 0.5–0.78C takes place [14,15], and this is the reason for defining standard HD as a ‘warm’ treatment. However, further analysing the behaviour of individual patients, differentiated responses (with some subjects who tended to warm up and others to cool down), for the same T of the dialysate were described [15]. To better understand the physio-pathological mechanisms and the clinical implications underlying dialytic hyperthermia, it is useful to review the physiology of the relations between T and arterial pressure, to evaluate whether the haemodynamic profile during HD is in accordance with physiology and finally, to study what determines thermal balance in the course of HD.