Randomised Controlled Trial of Dialysate Cooling for the Preservation of Cognitive Function in People Receiving Haemodialysis

Cognitive impairment is increasingly recognised as a major problem in the 25,000 adults with kidney failure treated by haemodialysis in the United Kingdom. The onset of cognitive impairment happens early after starting haemodialysis, and has an accelerated course so that some degree of cognitive impairment is almost universal, and over two-thirds of patients have moderate to severe reductions on cognitive function testing. A systematic review and meta-analysis of cognition in people on haemodialysis found that their global cognitive function scores were 25% lower than the general population, particularly in the domains of working memory, attention, orientation and executive function. Furthermore, the one-year risk for developing multi-infarct dementia in people on haemodialysis is more than 7 times higher than the general population. Investigators previously studied brain white matter microstructure in haemodialysis patients using diffusion tensor magnetic resonance imaging (MRI) and demonstrated the presence of significant brain white matter damage in haemodialysis patients compared to aged-matched controls. The amount of brain white matter damage correlated with lower score in cognitive function tests. Cognitive impairment in people on haemodialysis is a major component of the frailty syndrome that is also common in this population and cognitive impairment leads to increased dependency, poor quality of life, and reduces the ability of patients to correctly take medications and follow other important aspects of treatment such as dietary and fluid restrictions. Furthermore, cognitive impairment is associated with reduced survival and higher rates of hospitalisation. This has a further detrimental impact on quality of life and results in additional costs to the health service.

Working memory is the ability to hold and manipulate information in mind in the short term, which enables completion of cognitive tasks. It is one of the executive functions and it is well established that these are predictors for ability to carry out activities of daily living such as managing medications and using transport, executive function test performance is a stronger predictor than general cognition for activities and is vital for decision making. Working memory has been suggested as a potential treatment target due to its importance as a predictor of function. Letter Number Sequencing (LNS) is a well established test for working memory, it is derived from the Wechsler Adult Intelligence Scale-III and Wechsler Adult Intelligence Scale-IV test batteries. The LNS has clear face validity as a test of working memory. The participant hears a sequence of letters and numbers and must sort them into separate sequences of letters and numbers in the correct order. Testing working memory using LNS in haemodialysis patients compared to healthy controls demonstrated significant impairment in haemodialysis patients' working memory. Currently there are no interventions proven to protect cognitive function in people on haemodialysis.

Although haemodialysis is life sustaining for those with kidney failure, it is increasingly recognised that the haemodynamic effects of the treatment itself have negative effects on the blood flow to various vascular beds. These effects manifest as intradialytic hypotension, myocardial hypoperfusion leading to myocardial stunning and brain hypoperfusion. Dialysate cooling was first tested as an intervention in the 1980s to reduce intradialytic hypotension. Since then, there has been a large body of evidence that demonstrated beneficial effect of dialysate cooling on blood pressure. Dialysate cooling has been found to improve the stability of blood pressure and reduce intradialytic hypotension, confirmed in two metanalyses.

In smaller studies, dialysate cooling has also been shown to protect against intradialytic cardiac dysfunction and ameliorate myocardial stunning. In a previous NIHR-funded randomised clinical trial, investigators tested the hypothesis that improved haemodynamic stability with dialysate cooling would reduce the development of ischaemic lesions in the brain as determined by diffusion tensor imaging (DTI) MRI. The investigators randomised 73 participants to dialysate cooling to 0.5○C below body temperature or standard dialysis at 37○C. After one-year the standard haemodialysis patients developed MRI changes consistent with ischemic brain injury (increased fractional anisotropy and reduced radial diffusivity), cooled dialysate improved hemodynamic tolerability, and this group were protected against the development of brain white matter changes. The trial was designed with DTI-MRI as the primary outcome measure, and this meant that the study was not powered or designed to assess the effect on cognitive function.

More recently, a large cluster randomised, registry-based pragmatic trial (MyTEMP) conducted in Canada failed to demonstrate that dialysate cooling led to survival benefit or reduction in major cardiovascular events. However methodological limitations, lack of granularity of patient-level data and absence of important patient-centred outcomes mean that these results do not address questions as to whether cooling the dialysate can improve patient-centred outcomes, including protection against the development of cognitive impairment.