Hyperkalemia: getting to the heart of the matter.

Hyperkalemia is a common and life-threatening complication frequently seen in patients with end-stage renal disease (ESRD), advanced chronic kidney disease (CKD) and acute kidney injury. Indeed, acute hyperkalemia is one of the most common reasons for patients requiring emergency dialysis [1]. However, hyperkalemia does not affect all patients in the same manner. Some of them persistently exhibit chronic hyperkalemia without ostensible signs or symptoms, whereas others are clearly symptomatic with the same plasma potassium concentration ([K]p). The former group appears to develop an undefined compensatory mechanism to mitigate the effects of long-term hyperkalemia. Thus, a significant variation appears to exist in the tolerance to hyperkalemia. A [K]p that causes no signs, symptoms or changes on the electrocardiogram (ECG) for one patient with chronic hyperkalemia may subject another patient to significant risk, presumably due to greater cell membrane depolarization. To date, there is no predictor of the degree of hyperkalemia a patient can tolerate without an adverse event, except for the ECG and even this direct assessment of electrical cardiac conduction has its limitations (see below). Thus, there is no diagnostic test that can determine what [K]p is acceptable for that particular patient. This knowledge would be particularly useful for the treatment of various diseases. For example, in CKD, certain drugs (angiotensin-converting enzyme inhibitors, angiotensin receptor blockers and mineralocorticoid receptors blocker) often cause hyperkalemia. Despite this, each of these drug classes has been shown to be beneficial for cardiac and renal protection. Thus, their potential to cause hyperkalemia can be significant and may outweigh the potential benefit for an individual patient. Indeed, hyperkalemia is a frequent end-point that requires discontinuation of these beneficial medicines. If there was a way to determine what degree of hyperkalemia was acceptable in an individual patient, then a more aggressive treatment plan could be enacted for that patient. As of now, most physicians feel obliged to treat hyperkalemia, particularly significant hyperkalemia ([K]p > 6.0 mEq/L) as potentially dangerous. This is based on epidemiological data that demonstrate an association of hyperkalemia with morbidity and mortality in different patient populations. Until a test is developed which would allow a physician to deem an elevated [K]p as tolerable for a particular patient, this practice should continue. All this is not to say that hyperkalemia never affects diagnostic tests. Hyperkalemia is manifested on an ECG, but the correlation between ECG changes and [K]p is imprecise: only in 50% of patients with [K]p >6.5 mEq/L changes will be seen on an ECG. A common manifestation of mild hyperkalemia (5.5–7.0 mmol/L) in ECG is the ‘tented T-waves,’ which are characterized as tall, peaked and narrow-based [2]. Unfortunately, not all patients with mild hyperkalemia exhibit tented T-waves. If there was another accurate ECG measurement that could predict whether the hyperkalemia was life-threatening, this would be a major advance to direct therapy. An ECG is a relatively noninvasive procedure that is widely available, which makes it an excellent candidate for this use. The article by Green et al. [3] from Manchester in this issue of Nephrology, Dialysis, and Transplantation aims to do just this. The authors propose a possible predictive tool based on ECG reading for hyperkalemia in ESRD and CKD stage 5. In this study, the authors examine the utility of the ratio of the T-wave to the R-wave (T:R) and whether it is more useful than the presence of tented T-waves as a predictor of hyperkalemia in patients with ESRD. The authors report that tenting was no more common in cases of hyperkalemia compared with normal serum potassium and was less common than left ventricular hypertrophy. T:R was less sensitive, but more specifically identified hyperkalemia with a serum potassium >6.0 mmol/L. They also noted that no clinical feature exhibited a correlation with the likelihood of developing abnormal T-waves in hyperkalemia. Finally, they report that abnormal T-waves in patients with hyperkalemia had greater all-cause mortality in their patient population with a mean follow-up of 3.5 years compared with those with normal T-waves. Unfortunately, this study did not provide us with a tool for guiding management of hyperkalemia within an individual patient. Neither T-wave shape or T:R ratio offered a good, predictive method for the management of hyperkalemia in ESRD. The authors remarked that younger individuals have a higher rate of ‘tented’ T-waves and that older patients naturally have lower T-waves. Because the majority of patients in the ESRD population are older