Background
Sudden cardiac death (SCD) is the leading cause of death in hemodialysis. Coronary events and atherosclerotic risk factors are only weakly associated with SCD in hemodialysis, which suggests that lethal ventricular arrhythmias in this population may largely contribute to high rates of SCD [
1,
2]. On hemodialysis, arrhythmias can occur during and shortly after receiving treatment, independent of traditional cardiovascular risk factors [
3,
4]. Despite the increased risk of SCD, there is no consensus on dialysis-related risk factors for the development of arrhythmias.
Calcium, potassium, and magnesium may contribute to the development of arrhythmias during hemodialysis as these cations play a major role in development of the ventricular action potential and propagation of the electrical impulse [
4‐
6]. During hemodialysis, a rapid decrease in serum calcium or potassium can result in prolonged QT interval and increased QT dispersion [
6‐
9]; however, findings from observational studies and randomized control trials in hemodialysis are inconsistent. In hemodialysis, serum calcium levels fluctuate due to routine intermittent nature of the treatments and certain medications. The optimal serum calcium range, and type of calcium measurement remains unclear as guidelines are targeted to lower risk of vascular calcification [
10].
Observational studies of prevalent dialysis patients demonstrated associations of calcium and potassium concentrations with the risk of arrhythmias and arrhythmic events. For example, low dialysate calcium concentration was associated with longer corrected QT (QTc) interval (> 440 ms); larger serum-to-dialysate calcium gradient was associated with a higher risk of sudden cardiac arrest [
4,
11]; and low dialysate potassium concentration and low and high predialysis potassium levels were associated with an increased risk of sudden cardiac arrest [
12]. Conversely, findings from a trial of 3883 prevalent hemodialysis participants report that baseline dialysate calcium and serum-to-dialysate calcium gradient are not associated with cardiovascular outcomes including arrhythmic events [
13]. While many of these studies examined the relationship of calcium and potassium with arrhythmic events – often cross-sectionally, fewer studies have examined repeated measures of arrhythmic risk and compared various measures of longitudinal serum and dialysate electrolytes to the development of arrhythmias such as QTc in incident hemodialysis patients. The objective of this study was to examine the cross-sectional and longitudinal associations of serum and dialysate electrolytes with QTc interval and prolongation in incident hemodialysis patients.
Discussion
Among incident hemodialysis participants, prolonged QTc is very common between dialysis sessions and persists over time. Electrolytes are critical for normal pacing of the heart and low ionized calcium and potassium levels were consistently associated with abnormally prolonged QTc. In contrast, magnesium levels, dialysate, and serum-dialysate gradient for calcium concentrations were not independently associated with prolonged QTc. The results suggest that closer monitoring of lower levels of serum ionized calcium and serum potassium may be important in assessing arrhythmic risk in in-center hemodialysis patients.
Abnormally prolonged QT interval has been associated with adverse cardiovascular outcomes in past studies among the general and CKD population. In the general population, longer QT intervals were predictive of all-cause mortality and cardiovascular mortality in large cohorts over 3 to 30 years follow-up periods, but it is not consistent in all cohorts [
20‐
27]. In the chronic kidney disease population, several studies including the Cardiovascular Health Study (CHS) found that among individuals with CKD or patients receiving hemodialysis or peritoneal dialysis, longer QT interval was associated with an increased risk of death, heart failure, coronary heart disease, and sudden death [
28‐
30]. While QT intervals were also strongly associated with cardiovascular risk factors in these studies, prolonged QT interval remained an independent risk factor for adverse outcomes even after adjusting for traditional risk factors. Prolonged QT intervals may play an important role in risk stratifying individuals with a high risk of cardiovascular mortality.
Electrocardiographic changes, such as QTc prolongation, are common in those on hemodialysis and represent a potential common pathway leading to arrhythmias and potentially sudden cardiac death [
4,
6,
31,
32]. We compared multiple measures of serum, ionized, and dialysate electrolytes (calcium, potassium, magnesium) with QTc prolongation, a well-recognized arrhythmogenic risk factor, cross-sectionally and longitudinally, and found that potentially modifiable factors, lower ionized calcium and potassium, consistently and independently predicts QTc prolongation, even after adjusting for use of QT prolonging medications.
Only 5% of the participants were on cinacalcet and thus it is unlikely that this contributed to lower calcium levels and high burden of prolonged QTc interval in our population. Although one third of the participants were receiving a QT prolonging medication, most of which was furosemide, this did not significantly modify the association between calcium and QTc measures. Studies such as the HEMO and DOPPS, which include predominantly prevalent dialysis patients have reported an increased risk of mortality in calcium levels less than 8.5–9.0 mg/dl [
5,
33,
34]. We demonstrate that although the associations of serum calcium with QTc prolongation were not statistically significant, lower levels of ionized calcium were associated with increased risk of QTc prolongation.
Our findings also suggest that low predialysis serum potassium is associated with prolonged QTc at baseline and longitudinally, supporting recent studies demonstrating a significant association between serum potassium level and sudden cardiac event or mortality in dialysis patients [
12,
35,
36]. Serum potassium has been shown to increase the risk of sudden cardiac arrest and all-cause mortality in dialysis patients with serum potassium levels less than 5.1–5.3 mEq/l or ≤ 4.0 mmol/l. [
12,
35,
36] Similarly, we demonstrate that in incident in-center hemodialysis patients, lower levels of serum potassium are associated with longer QTc interval and may consequently contribute to developing other arrhythmias. A recent study of chronic kidney disease participants demonstrated that hyperkalemia (> 5 mEq/l) rather than hypokalemia (< 3.5 mEq/l) was associated with a higher risk of sudden cardiac event [
37]. Due to our limited sample size, we could not asses this U-shaped relationship with potassium.
Currently, the use of total versus corrected or ionized calcium is debated when considering routine screening in bundled lab testing and also treating targets according to the KDIGO clinical practice guidelines [
38]. Routine measurement of serum ionized calcium is not cost-effective and the use of albumin-corrected calcium remains less specific and non-superior compared to other measures of serum calcium [
38,
39]. Our study suggests that ionized calcium is a more consistent and stronger predictor of longer QT interval and QT prolongation, independent of serum potassium levels and other traditional risk factors. As 30% of our study population were found to have QT prolongation at follow-up, most of whom had persistent QT prolongation, closer attention to serum ionized calcium or potassium levels may be important to help mitigate risk of arrhythmias over time. This may be critical especially in dialysis, as the ionized calcium level is the set point for physiological actions including electrophysiology and muscle contraction, and may be more valuable in diagnosis hyperparathyroidism [
40].
In recent studies and clinical guidelines, most concerns regarding serum calcium point to potential calcium loading and the risk of vascular calcification [
11,
38,
41]. Vascular calcification is common in ESRD [
42], and both hypercalcemia and vascular calcification are associated with adverse cardiac events and mortality [
43‐
45]. The 2017 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines weakly suggests that hypercalcemia should be avoided [
10]. Although our findings do not show a significant association with serum calcium and prolonged QTc, we show that low levels of ionized calcium is a strong predictor of QTc prolongation, a known arrhythmogenic risk factor, suggesting that methods to monitor and avoid low ionized calcium levels may reduce the risk. Given the concerns for the potential contribution of higher serum calcium levels to vascular calcification, the optimal strategy for maintaining calcium balance in dialysis definitely requires further prospective studies.
Interestingly, our study demonstrates a lack of association between dialysate or serum-dialysate gradient measures and QTc prolongation, even though a few studies have demonstrated significant associations with cardiovascular disease [
11]. A case-control study of 502 cardiac arrest cases and 1632 matched controls reported that a dialysate potassium concentration < 2 mEq/l was associated with higher odds of cardiac arrest [
12]. We were unable to examine low dialysate potassium concentrations as only 2% of our participants received < 2 mEq/l. In contrast, results from the EVOLVE trial reported no association between baseline dialysate calcium or serum-dialysate calcium gradient with sudden cardiac death [
46]. Our results are similar to findings from the EVOLVE trial in that although the direction of the dialysate calcium estimates suggest that lower dialysate calcium concentrations may predict higher arrhythmic risk, the association was not statistically significant. This is most likely due to a much smaller sample size in each category of dialysate calcium. Our study does differ from previous dialysis studies as we examined QTc intervals on interdialytic days and examined electrolytes using multiple predialysis and interdialytic measurements over follow-up.
A limitation of our study is that a smaller proportion of participants completed follow-up cardiovascular evaluations at 1 year; however, the ECG measurements and serum and dialysate electrolytes did not differ by follow-up status. Therefore, while the follow-up cohort may be underpowered, it is less likely that it is biased. Another limitation is the different timing of electrolyte measurements that may potentially influence the observed associations as ionized calcium was done on the same day as the QTc interval measurement and total calcium was performed on routine monthly dialysis testing. Future studies are needed to carefully examine how the timing of measurement can affect the relationship between electrolytes and QT interval. Despite these limitations, important strengths need to be acknowledged. This is a large cohort of incident in-center hemodialysis patients with standardized cardiovascular and clinical characterization and also a large minority population often underrepresented in studies; therefore, findings from this study may be generalizable to these populations.
Acknowledgements
We dedicate this manuscript to the memory of our friend and colleague, Dr. W.H. Linda Kao, and her enduring work on improving patient outcomes in chronic disease. We thank our participants, nephrologists and staff of the DaVita and MedStar dialysis units in the Baltimore area who contributed to the PACE study. We would like to thank the participation of dialysis practices in Baltimore, in particular, the Mid-Atlantic Nephrology Associates and the Nephrology Center of Maryland. We thank the PACE study and the Johns Hopkins Clinical Research Unit staff for their effort, Kimberly Keck, and the members of the Data Safety Monitoring Board of the study, Drs. Paul Scheel, Luis Gimenez and Roger Blumenthal.
We thank the PACE Study Endpoint Committee: Bernard G. Jaar, MD, MPH (Chair); Michelle M. Estrella MD, MHS; Stephen M. Sozio MD, MHS, MEHP; Rulan S. Parekh MD, MS; N’Dama Bamba MD; Wei Tsai MD, MS, MPH; Geetha Duvuru, MD; Julia Scialla MD, MHS; Teresa K. Chen, MD, MHS; Jose Manuel MonroyTrujillo, MD; Frances-LLena Capili, MD; Ijaz Anwar, MD; Lili Zhang, MD; Manisha Ghimire, MD; Raghotham Narayanaswamy, MD; Ramya Ravindran, MD; Svetlana Chembrovich, MD; Stefan Hemmings, MD, Steven Menez, MD.