Skip Navigation


NDT Advance Access originally published online on November 28, 2007
Nephrology Dialysis Transplantation 2008 23(4):1378-1386; doi:10.1093/ndt/gfm765
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/4/1378    most recent
gfm765v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Severi, S.
Right arrow Articles by Santoro, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Severi, S.
Right arrow Articles by Santoro, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org


Calcium and potassium changes during haemodialysis alter ventricular repolarization duration: in vivo and in silico analysis

Stefano Severi1,*, Eleonora Grandi1,*, Chiara Pes1, Fabio Badiali2, Fabio Grandi3 and Antonio Santoro4

1 Biomedical Engineering Laboratory – Department of Electronics, Computer Science and Systems, University of Bologna, Cesena, Italy 2 Nephrology and Dialysis Division – Infermi Hospital, Rimini, Italy 3 Hospal S.p.A., Bologna, Italy 4 Malpighi Nephrology Division, Policlinico S. Orsola-Malpighi, Bologna, Italy

Stefano Severi, Biomedical Engineering Laboratory, D.E.I.S., University of Bologna, Via Venezia 52, I-47023 Cesena, Italy. Tel: +39-0547-339202; Fax: +39-0547-339208; Email: stefano.severi{at}unibo.it



   Abstract
Top
 Abstract
Introduction
 Subjects and methods
 Results
 Discussion
 Conclusions
 References
 
Background. Alterations of ventricular repolarization duration, as measured by the QT interval, are frequently observed in haemodialysis (HD) patients. The nature and the sign of these changes are not yet fully understood.

Methods. Different dialysate K+ and Ca2+ levels, leading to different end-HD plasma concentrations in the patient, have been tested in the present study in terms of their impact on QTc. A model of the human cardiomyocyte action potential (AP) has been used to assess in silico whether the changes in Ca2+ and K+ were able to justify at the cellular level the observed alterations of QTc.

Results. QTc was prolonged in HDs with low (1.25 mM) versus high (2 mM) Ca2+ (424 ± 33 versus 400 ± 28 ms, P < 0.05) and in HDs with low (2 mM) versus high (3 mM) K+ (420 ± 35 versus 399 ± 36 ms, P < 0.05). These alterations were confirmed at the cellular level by computational analysis showing prolongation of ventricular AP at low K+ and low Ca2+ at the same extent of the measured QTc variations. Numerical simulation predicted a critically long AP (and QT) when considering low K+ and Ca2+ simultaneously, suggesting the concurrent lowering of Ca2+ and K+ as a potential arrhythmogenic factor.

Conclusions. Numerical simulations of the ventricular AP may be useful to quantitatively predict the complex dependence of AP duration on simultaneous changes in Ca2+ and K+. Moreover, Ca2+ content in the dialysate should be designed not to critically lower serum Ca2+, especially in sessions at risk of end-dialysis hypokalaemia.

Keywords: calcium; electrolytes; electrophysiology; haemodialysis; ventricular repolarization



   Introduction
Top
 Abstract
 Introduction
Subjects and methods
 Results
 Discussion
 Conclusions
 References
 
Mortality from cardiac arrest and sudden death is unusually high in patients receiving maintenance haemodialysis (HD). Electrocardiographic alterations are frequently observed and the incidence of ventricular arrhythmias increases during and immediately after the HD session [1–3].

The QT interval is a recognized ECG marker of the ventricular repolarization and its prolongation has been associated with increased risk of sudden death in both pathological [4–6] and healthy [7] populations. For this reason, several studies addressed the analysis of QT, which is often significantly changed by the end of the dialysis treatment. Nevertheless, the results on the effect of HD on QT are not univocal. In some studies, the QTc interval increased during the dialysis session [8–14]. In other studies the QTc interval did not change, or decreased or showed a variable response [15–25]. Discrepancies between the studies may be attributed to population selection, methods of QT measurement [26,27] and variables related to the dialysis technique.

Apart from the above-mentioned confounding factors, HD therapy per se introduces several factors that can influence the duration of the QT interval. HD impacts body fluid composition, tissue hydration, electrolyte equilibrium and adrenergic activation. All these factors are known to have considerable effects on the excitability of the cardiac cells and potentially contribute to the observed QT changes. Since these factors act simultaneously and exert a complex influence on cardiac activity, it is difficult to assess their relative importance. In fact, mechanisms underlying QT alterations during haemodialysis are not completely understood, nor it is clear the clinical relevance.

Electrolyte disorders are one of the main HD-related factors that can cause QT interval alterations and cardiac arrhythmias, because of their involvement in the genesis, duration, morphology and propagation of the cellular action potential. The electrolytes that mostly influence the ventricular repolarization are K+ and Ca2+ [28]. In fact, in patients undergoing HD, prolongation of QTc was inversely correlated with variations in plasma Ca2+, suggesting that patients with the greatest reduction in Ca2+ had the greatest increases in QTc at the end of the HD session [14,16,18,20]. On the other hand, several authors pointed out the role of K+ removal as a pivotal arrhythmia-inducing factor associated with the dialysis treatment [1,22,29,30]. Since a HD session induces changes in both K+ and Ca2+ plasma concentrations, which mainly depend on the dialysate composition, an increasing interest has been devoted to a more rational design of dialysate K+ and Ca2+ contents, possibly time-profiled or targeted to the needs of specific subgroups of HD patients [16,19,22,30–32]. Matching the patient's haemodynamic and biochemical profile to an individualized dialysis prescription has been indicated as the most effective strategy to reduce the risk of cardiac arrest during HD [3]. In this perspective it seems crucial to deepen the knowledge on the specific effects of HD-induced Ca2+ and K+ changes on cardiac electrical activity.

The aim of the present study was to explore the impact of HD-induced changes in plasma Ca2+ and K+ concentrations on the duration of the ventricular repolarization by using a computational model of the human ventricular cardiomyocyte [33]. To this purpose, we tested two different levels of Ca2+ and K+ concentration in the dialysis bath and compared measured QTc interval variations to simulated data on action potential duration (APD). Trough simulation the impact of each individual ion concentration change on the human ventricular AP was assessed and the effect of concurrent K+ and Ca2+ variations was predicted.



   Subjects and methods
Top
 Abstract
 Introduction
 Subjects and methods
Results
 Discussion
 Conclusions
 References
 
Inclusion criteria
End-stage renal disease (ESRD) patients were considered eligible for the study if on HD three times a week for at least 6 months and aged between 50 and 90 years. None of them had been receiving anti-arrhythmic or anti-hypertensive treatment with beta-blockers. None had been receiving variable doses of digitalis. Further exclusion criteria were mental disease, presence pacemaker or cardiac stimulator and diabetic neuropathy. During the treatments any oral or intravenous intake of K+ and Ca2+ had to be avoided. The study was performed at the Nephrology and Dialysis Units of the Malpighi Hospital in Bologna and the Infermi Hospital in Rimini. Informed patient consent was obtained.

Alternative dialysate Ca2+ HDs
To assess the effect of serum Ca2+ levels on QTc 23 HD patients (5 women and 18 men, 71.1 ± 9.9 years old) were enrolled in a crossover design. The causes of ESRD were chronic glomerulonephritis (n = 8), diabetic nephropathy (n = 2), nephritis caused by abuse of analgesics (n = 2), vascular nephropathy (n = 4), polycystic kidney disease (n = 2) and nephrosclerosis (n = 5).

Each patient underwent two acetate-free biofiltration (AFB) dialysis sessions. AFB is a mixed diffusive–convective dialysis therapy characterized by the total absence of any buffer in the dialysis bath. Patient's acid–base balance is restored by infusing in postdilution mode a sterile solution of sodium bicarbonate at a concentration ranging from 145 to 167 mEq/L. The infusion flow rate was set between 2 and 2.4 L/h depending on individual prescription. The two study sessions were identical (time of dialysis, blood, dialysate and infusion flow rate, bath conductivity and size of the polyacrilonitrile haemofilter) except for Ca2+ concentration in the dialysate. Namely, the dialysate (in mM: Na+ 139, K+ 3, Mg2+ 0.37, Cl– 145 and glucose 5.55) Ca2+ concentration was set at 1.25 or 2 mM. The ultrafiltration rate was set depending on patient needs; it was 0.80 ± 0.22 L/h in 1.25 mM Ca2+ HD and 0.79 ± 0.20 L/h in 2 mM Ca2+ HD (NS).

Alternative dialysate K+ HDs
Twelve patients (6 women and 6 men with a mean age of 71.3 ± 9.0 years) were studied to investigate the impact of different K+ levels. The causes of ESRD were chronic glomerulonephritis (n = 2), diabetic nephropathy (n = 1), nephritis caused by abuse of analgesics (n = 1), vascular nephropathy (n = 2), polycystic kidney disease (n = 4) and nephrosclerosis (n = 2).

Each patient underwent two identical AFB dialysis sessions except for K+ concentration in the dialysate. Namely, the dialysis bath (in mM: Na+ 139, Ca2+ 2, Mg2+ 0.37, Cl– 146 and glucose 5.55) contained 2 or 3 mM K+. The ultrafiltration rate was set depending on patient needs; it was 0.76 ± 0.24 L/h in 2 mM K+ HD and 0.83 ± 0.21 L/h in 3 mM K+ HD (NS).

QT interval analysis
Twelve-lead electrocardiogram (ECG) was continuously recorded (H-12 holter, Mortara Instruments Inc., Milwaukee, WI, USA). ECGs were sampled at 180 Hz. At the end of the recording session, the ECGs were transferred to the hard disk of a personal computer. Sections of ECGs containing five consecutive sinus beats were extracted for all the recordings at t = 0 min and at the end of the treatment. For each section, QT and RR intervals of the three central sinus beats were calculated by a semiautomatic procedure.

The computer-automated technique identified the end of the T wave by the intersection between the baseline and the slope of the T wave between 30 and 70% of peak amplitude [34]. The only analyst intervention was the identification of the baseline. T- wave amplitude was measured from T-wave peak to TP baseline. T waves with an amplitude less than 100 µV were excluded.

For each beat, the QT interval was obtained as the mean on the eight independent leads. QT intervals were corrected for heart rate with Bazzet's formula: QTc = QT/ {surd}RR. Repeatability of the QTc measurement from one session to the next independent of any changes in electrolyte composition was assessed on a small subset of patients (16 haemodialysis sessions, 4 patients). No significant difference was found between measurements obtained from the same patient (3.5 ± 19.2 ms, with a coefficient of variation of 4.5%).

Statistical analysis
All the results are expressed as mean ± standard deviation. Repeated measures (Geisser–Greenhouse correction) ANOVA and Tukey–Kramer Multiple-Comparison Test were applied to every measured parameter for which normal distributions were found (Skewness, Kurtosis and Omnibus tests). Wilcoxon Rank-Sum nonparametric test was applied to parameters for which the normal distribution assumption was not met. Multiple regression analysis was used to determine significant correlations with QTc changes. All the tests were performed with the NCSS statistical tool (NCSS, Kaysville, UT, USA).

Action potential model
The ventricular action potential was simulated using the Ten Tusscher et al. [33] model of human ventricular cell. The model describes all the main membrane currents and active transport mechanisms and the regulation of intracellular Ca2+ concentrations.

The model has been validated against a wide set of experimental data [33]. With respect to the original model formulation, intracellular Na+ and K+ concentrations were kept constant during the simulations. [K+]i was set to 138.3 mM. [Na+]i was kept at 11.6 mM in the pre-dialysis simulations and to 12.9 mM in end-dialysis simulations. In addition, the Ca2+-dependent inactivation gate of ICa, fCa, was modified according to the following formula on the basis of recent studies demonstrating that the Ca2+ channel inactivation process depends more strongly on local Ca2+ than on membrane potential [35–37]:


Formula

where


Formula

Model differential equations were implemented in Simulink (Mathworks Inc., Natick, MA, USA).

The computer simulator was used to reproduce the pre- and end-dialysis conditions. The average plasma Ca2+, Na+ and K+ concentrations recorded at the beginning and at the end of HD sessions were set as extracelullar concentrations in the model. Heart rate recorded at the beginning of treatment was used as model input for both (pre- and end-dialysis) simulations in order to exclude the heart rate adaptation of APD and to make comparisons with the heart rate-corrected QT.

Simulations were performed considering an in silico M-cell and lasted until a steady state AP was reached (typically 100 s). The action potential duration was measured as the interval between the action potential upstroke and the 90% repolarization level of the action potential (APD90).



   Results
Top
 Abstract
 Introduction
 Subjects and methods
 Results
Discussion
 Conclusions
 References
 
Electrolyte changes during HD
As expected, HD with 2 mM Ca2+ dialysate led to a significant increase in the serum Ca2+ level at the end of the treatment, whereas 1.25 mM Ca2+ HD significantly decreased the plasma Ca2+ (Table 1) because of the significant convective removal during AFB [38]. The final Ca2+ levels were different between the two HD protocols (1.38 ± 0.11 mM versus 1.03 ± 0.10 mM, P < 0.001). Plasma pH was higher in 1.25 mM Ca2+ HD. No differences were observed in K+, Na+ and HCO3– between the two HDs.


View this table:
[in this window]
[in a new window]

  		Table 1 Pre- and end-dialysis laboratory data measured on the haemodialysis procedures with alternative Ca2+ dialysate

 
Two different levels of K+ in the dialysis bath led to different end dialysis plasma K+ concentrations. Three millimolar K+ HD caused a significantly less pronounced decrease in the serum K+ level with respect to 2 mM K+ HD (3.43 ± 0.51 mM versus 2.86 ± 0.42 mM, P < 0.01). No other differences were observed between the two protocols (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 2 Pre- and end-dialysis laboratory data measured on the haemodialysis procedures with alternative K+ dialysate

 
Effect of HD-induced Ca2+ and K+ changes on the QTc interval
The QTc interval changes during the HD sessions were not univocal and their trend (prolongation or shortening) was dependent on the level of Ca2+ and K+ in the dialysate. Indeed, none of the four tested protocols was able to induce a significant post- versus pre-dialysis alteration of the QTc. Nevertheless, despite the same initial values, QTc at the end of the sessions was found significantly different both between 1.25 and 2 mM Ca2+ HDs and between 2 and 3 mM K+ HDs (Tables 3 and 4). The end-dialysis QTc interval was longer with lower Ca2+ concentration in the dialysate (424 ± 33 ms versus 400 ± 28 ms, P < 0.05) and with lower K+ dialysate (420 ± 35 ms versus 399 ± 36 ms, P < 0.05).


View this table:
[in this window]
[in a new window]

  		Table 3 Effect of haemodialysis with alternative Ca2+ dialysate on HR, QT and QTc

 

View this table:
[in this window]
[in a new window]

  		Table 4 Effect of haemodialysis with alternative K+ dialysate on HR, QT and QTc

 
In Figure 1, QTc values at the beginning and at the end of sessions with alternative dialysate Ca2+ or K+ concentrations are reported for each patient. QTc was prolonged in 18 out of 23 patients by dialysis with 1.25 mM Ca2+, whereas it was shortened in 14 patients by dialysis with 2 mM Ca2+. QTc was prolonged in 8 out of 12 patients by dialysis with 2 mM K+, whereas it was shortened in 10 patients by dialysis with 3 mM K+.


Figure 1
View larger version (32K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 Individual values of QTc before and at the end of HD sessions with different dialysate Ca2+ (upper panels) or K+ (lower panels) concentrations.

 
The change in the QTc interval induced by the study HD sessions correlated inversely with change in serum Ca2+ (P < 0.01, Figure 2). In multiple regression analysis, where the changes in Na+, Ca2+, K+, pH and HCO3– were assumed as independent variables, this was the only independent correlation with QTc change.


Figure 2
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 Scatter plot and regression line showing the significant inverse correlation between QTc interval duration and serum Ca2+ concentration changes during HD sessions (35 patients, 70 sessions).

 
Effect of HD-induced Ca2+ and K+ changes on APD
Simulated changes in ventricular APs induced by HD with alternative dialysate compositions were in good agreement with our in vivo results on the QTc interval. Representative traces of APs corresponding to 1.25 and 2 mM Ca2+ HDs are depicted in Figure 3. As expected, K+ removal during HD induced membrane hyperpolarization independently on the simulated external Ca2+ concentration (Figure 3, end-dialysis APs). The lower Ca2+ content in the dialysate led to prolongation of AP under the end-dialysis condition (+5.3%), whereas the higher Ca2+ content in the dialysate shortened AP (–2.7%, Figure 4, upper panel). Similarly, the lower K+ concentration increased APD90 (+2.1%), whereas the higher K+ decreased it (–2.3%, Figure 4, lower panel).


Figure 3
View larger version (15K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 Simulated ventricular action potentials (APs) under the conditions of pre- and end-dialysis with different dialysate Ca2+. Simulations were performed by using as input to the ventricular cell model the average measured K+, Ca2+ and Na+ concentrations as reported in Table 1. The lower Ca2+ content in the dialysate leads to prolonged AP under the end-dialysis condition, whereas the higher Ca2+ content in the dialysate shortens AP.

 

Figure 4
View larger version (19K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4 Comparison of percent variations in measured QTc (mean ± SD) and simulated APD in HD with alternative dialysate Ca2+ (upper panels) or K+ (lower panels). According to experimental results, APD increases at lower dialysate K+ or Ca2+, whereas it decreases at higher K+ or Ca2. Both QTc and APD were normalized with respect to their average pre-dialysis value.

 
We compared HD-induced ventricular AP variations with experimental data from previous clinical studies. When using the experimental electrolyte values by Nappi et al. [16], who also performed HDs with different dialysate Ca2+ concentrations, the simulated ventricular APDs qualitatively reproduced the in vivo results on the QTc interval, as shown in Figure 5. In particular, also in this case both QTc and APD90 were inversely correlated to the dialysate Ca2+ level. Covic et al. [18] compared patients with increasing QTc and patients with decreasing QTc during HD. The different experimental QTc trends were reproduced by simulations at the cellular level (QTc: +5.8% versus –3.4%, APD90: +3.4% versus –1.8%). Similarly, the QTc prolongation found by Morris et al. [11] when patients experienced a HD-induced Ca2+ decrease was reproduced (QTc +3.5%, APD90 +3.9%) as well as the QTc shortening found by Floccari et al. [20] who observed a HD-induced Ca2+ increase accompanied by a limited K+ decrease (QTc: –1.4%, APD90: –1.9%).


Figure 5
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5 Comparison of percent variations in QTc measured by Nappi et al. and simulated APD in HD with alternative dialysate Ca2+. Simulations were performed by using as input to the ventricular cell model the average measured K+, Ca2+, Na+ and heart rate reported in [16]. According to experimental results, APD is inversely correlated to the dialysate Ca2+ level.

 
Combined effect of Ca2+ and K+ on APD
The impact of concurrent K+ and Ca2+ variations on APD90 was further analysed (Figure 6). A typically large HD-induced K+ decrease (from 6 mM to 3 mM) was tested with three different concurrent Ca2+ pre- versus end-dialysis variations (from 1.1 mM to 0.9, 1.1 and 1.3 mM, respectively). When a low end-dialysis K+ plasma concentration was accompanied by a decreased Ca2+ concentration the AP was largely prolonged (APD90: +16%). By observing the time course of the total membrane current, the AP prolongation can be mainly ascribed to a less rapidly activated and less pronounced positive repolarizing current under the end-dialysis condition with respect to the pre-dialysis (Figure 6, lower panel). The AP prolongation was markedly reduced when the HD-induced K+ reduction was accompanied by a constant (APD90: +11%) or increased (APD90: +7%) Ca2+ concentration. Under these conditions, a slightly lower negative inward current can be observed in the first stage of the plateau phase, due to a more pronounced Ca2+-dependent inactivation of the ICa current sustaining the AP plateau, whereas the positive repolarizing current seemed to be anticipated in time with no significant differences in morphology.


Figure 6
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6 Simulated ventricular APs (upper panel) and total membrane current (lower panel, positive values correspond to outward currents, the y-scale was zoomed to appreciate differences in the repolarization phase) under a typical pre-dialysis condition (K+ = 6 mM and Ca2+ = 1.1 mM, thin line) and under three end-dialysis conditions with different plasma Ca2+ concentrations. When a low end-dialysis K+ plasma concentration (3 mM) is accompanied by a decreased Ca2+ concentration (thick line), the AP is prolonged by 16%. Such prolongation is markedly reduced when the dialysis-induced K+ reduction is accompanied by a constant (dashed line) or increased (dotted line) Ca2+ concentration at the end of dialysis with respect to the pre-dialysis.

 


   Discussion
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
Conclusions
 References
 
Our results pointed out that the impact of HD sessions on the QT interval is not univocal and it strictly depends on Ca2+ and K+ levels in the dialysate. The QTc interval was prolonged in dialysis with low versus high Ca2+ and in dialysis with low versus high K+ (Figure 4). For the first time, these trends were confirmed by in silico analysis at the cellular level showing prolongation of ventricular AP at low Ca2+ and low K+ and predicting a critically long QT when simulating low Ca2+ and K+ simultaneously (Figure 6).

Ventricular repolarization and Ca2+
Lengthening of the QTc interval due to prolongation of phase 2 of the AP is a known electrocardiographic manifestation of hypocalcaemia and it might trigger after-depolarizations leading to dysrhythmias and torsades de pointes. The known effect of hypercalcaemia on the electrocardiogram is the opposite of hypocalcaemia with the hallmark of abnormal shortening of the QTc interval, but clinically significant rhythm disturbances associated with hypercalcaemia are rare. The significant inverse correlation between QTc prolongation and variations in plasma Ca2+ that we found is in agreement with previous reports on patients undergoing haemodialysis [14,16,18,20]. Nappi et al. [16] also performed HDs with different dialysate Ca2+ concentrations and measured the QT intervals by means of a tangent-based method. They found QTc changes similar to those we presented in this study and in our simulations APs of the digital cell well reproduced the calcium dependence of the alterations they observed (Figure 5). Genovesi et al. [14] showed the same QTc sensitivity to plasma Ca2+ variations we observed, but greater QTc prolongations. In fact, when analysing the linear regression between QTc and Ca2+ variations, they found a similar slope (63 versus 56 ms/mM) but different intercept (34 versus 10 ms) with respect to our results. Such systematic difference is likely due to differences in methodology and population characteristics.

Ventricular repolarization and K+
K+ plays an important role in maintaining the electrical potential across the cellular membrane, as well as in depolarization and repolarization of the cardiac myocytes. Alterations in serum K+ levels such as the HD-induced K+ removal may lead to electrocardiographic changes and severe arrhythmias. Previous studies failed to find a significant correlation between QTc prolongation and variations in plasma K+ [14,16,18,20]. To the best of our knowledge, the present study compared for the first time the effect on QTc of two dialysate formulations designed to have different end-dialysis K+ plasma levels (e.g. identical bath composition except for K+ concentration). We found that the duration of ventricular repolarization at the end of dialysis was significantly related to K+, being longer in sessions with lower K+ (Table 4). Buemi et al. [22] recently tested two different protocols of K+ removal leading to differences in plasma K+ during dialysis but to equal end-dialysis values. Coherently with our results they found prolonged QTc in the first half of the sessions with faster K+ removal, when lower plasma K+ concentrations were induced.

Combined effect of Ca2+ and K+ changes
Results from previous studies on QTc in HD indicated that when both Ca2+ and K+ change during the treatment, the overall result on QTc is not easily predictable. We pointed out that numerical simulation of the ventricular AP allows quantitative predictions on the complex dependence of APD on simultaneous changes in Ca2+ and K+. In silico analysis showed that K+ removal can lead to APD90 increase even if performed under conditions of constant plasma Ca2+ (dashed versus thin line in Figure 6) or increasing plasma Ca2+ (dotted versus thin line in Figure 6). On the other hand, when the HD-induced plasma K+ decrease is not large, its prolonging effect can be overwhelmed by a concomitant shortening effect of increasing plasma Ca2+, as it was the case in both the 2 mM Ca2+ and 3 mM K+ HDs. Under those conditions, the experimental observation of QTc shortening was well reproduced by simulated APs (Figure 4). Notably, this sort of analysis was successfully applied to data from other clinical studies reporting either HD-induced shortening or prolongation of QTc (see Figure 6 and the Results section).

Importantly, our in silico analysis predicted a critically long QT when simulating a low end-dialysis K+ plasma concentration accompanied by a decreased Ca2+ concentration (Figure 6). It is worth to note that the Ca2+ and K+ concentrations we used for that simulation are within the range of the measured values. By considering the average pre-dialysis QTc (410 ms, Tables 3 and 4) and a percent HD-induced increase equal to those obtained for APD (+16%, Figure 6) an end-dialysis critically prolonged (>440 ms [7,39]) QTc value of 476 ms was predicted. The HD-induced QT prolongation related to Ca2+ and K+ removal could be further exacerbated in uraemic patients receiving any drug that is capable of inhibiting K+ channels, so leading to the clinical manifestation of the acquired long QT syndrome, as recently discussed by Gussak and Gussak [40].

In silico analysis
In this study, in silico analysis has been performed by means of a mathematical model of the human ventricular AP. This sort of computational approach has proven valuable to link information obtained at the molecular level to the whole cell and to the clinical phenotype in a number of inherited [41,42] and acquired [43,44] cardiac diseases. For the first time in this study, it has been applied to analyse HD-induced modifications in ventricular electrical activity. The application of computational cell biology to the analysis of ECG changes associated to HD can contribute to deepen the knowledge on the cardiac impact of HD therapy. On the other hand, HD represents an unique model to study the effects of ionic changes on the ventricular repolarization process under relatively stable and controlled conditions, and it may contribute to clarify the mechanisms involved in the genesis of electrocardiographic repolarization abnormalities related to acquired prolonged QT.

Limitations
The two experimental protocols were designed to separately analyse the effects of Ca2+ and K+; a more complete analysis could be performed by applying a factorial design on a larger patient population. Actually, we also found a difference in end-dialysis pH between 2 mM Ca2+ and 1.25 mM Ca2+ HDs. Nevertheless, regression analysis excluded a direct influence of pH on QTc, as already reported in previous studies [16].

The model we used has been developed to describe a physiological condition [33], it does not take into account the known electrical remodelling effects of uraemic cardiomyopathy and it does not incorporate a reliable description of intracellular Na+ dynamics under large non-physiological variations of extracellular electrolytes. Thus, we set intracellular Na+ and K+ concentrations to constant values. A detailed model of human ‘uraemic cardiomyocyte’ should be identified for a more quantitative description of the cardiac electrical activity in uraemic patients.



   Conclusions
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Conclusions
References
 
The present study showed that the impact of HD on ventricular repolarization duration strictly depends on the level of Ca2+ and K+ in the dialysate. In silico analysis proved to be a valuable tool to gain insights into the impact of Ca2+ and K+ changes at the cellular level and predicted a critical QTc prolongation when a low end-dialysis K+ plasma concentration is accompanied by a decreased Ca2+ concentration. Our results suggest that Ca2+ content in the dialysate should be designed so as not to lower serum Ca2+, especially in sessions at risk for end-dialysis hypokalaemia. This aspect has to be carefully considered when designing new profiled and/or personalized dialysates.



   Acknowledgments
 
We wish to thank Luca Corazza, Filippo Zanetti and Giorgia Alessandrini for their help in data collection and analysis. The work was partially supported by Hospal SpA (Bologna, Italy). Mortara Rangoni Europe SRL (Casalecchio di Reno, Italy) equipped the study with the H-12 Holter recorder.

Conflict of interest statement. Fabio Grandi is a full-time employee of Hospal SpA. Dr Severi, Grandi E, Pes, Badiali and Santoro do not hold any shares and are not consultants for companies related to nephrology. Dr Pes has been supported by Hospal SpA.



   Notes
 
* Stefano Severi and Eleonora Grandi equally contributed to this work. Back



   References
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Conclusions
 References
 

  1. Abe S, Yoshizawa M, Nakanishi N, et al. Electrocardiographic abnormalities in patients receiving hemodialysis. Am Heart J (1996) 131:1137–1144.[CrossRef][Web of Science][Medline]
  2. Gruppo Emodialisi e Patologie Cardiovasculari. Multicentre, cross-sectional study of ventricular arrhythmias in chronically haemodialysed patients. Lancet (1988) 2:305–309.[Web of Science][Medline]
  3. Karnik JA, Young BS, Lew NL, et al. Cardiac arrest and sudden death in dialysis units. Kidney Int (2001) 60:350–357.[CrossRef][Web of Science][Medline]
  4. Algra A, Tijssen JG, Roelandt JR, et al. QTc prolongation measured by standard 12-lead electrocardiography is an independent risk factor for sudden death due to cardiac arrest. Circulation (1991) 83:1888–1894.[Abstract/Free Full Text]
  5. Schwartz PJ, Stramba-Badiale M, Segantini A, et al. Prolongation of the QT interval and the sudden infant death syndrome. N Engl J Med (1998) 338:1709–1714.[Abstract/Free Full Text]
  6. Priori SG, Bloise R, Crotti L. The long QT syndrome. Europace (2001) 3:16–27.[Free Full Text]
  7. Schouten EG, Dekker JM, Meppelink P, et al. QT interval prolongation predicts cardiovascular mortality in an apparently healthy population. Circulation (1991) 84:1516–1523.[Abstract/Free Full Text]
  8. Ahmed R, Kiya F, Kitano K, et al. Effects of combined changes in serum calcium and potassium on QT interval. A study by Holter electrocardiographic monitoring during hemodialysis. Jpn Heart J (1987) 28:813–828.[Medline]
  9. Erem C, Kulan K, Goldeli O, et al. Impact of hemodialysis on QT interval. Acta Cardiol (1995) 50:177–185.[Web of Science][Medline]
  10. Lorincz I, Matyus J, Zilahi Z, et al. QT dispersion in patients with end-stage renal failure and during hemodialysis. J Am Soc Nephrol (1999) 10:1297–1302.[Abstract/Free Full Text]
  11. Morris ST, Galiatsou E, Stewart GA, et al. QT dispersion before and after hemodialysis. J Am Soc Nephrol (1999) 10:160–163.[Abstract/Free Full Text]
  12. Yetkin E, Ileri M, Tandofan I, et al. Increased QT interval dispersion after hemodialysis: role of peridialytic electrolyte gradients. Angiology (2000) 51:499–504.[Web of Science][Medline]
  13. Meier P, Vogt P, Blanc E. Ventricular arrhythmias and sudden cardiac death in end-stage renal disease patients on chronic hemodialysis. Nephron (2001) 87:199–214.[CrossRef][Web of Science][Medline]
  14. Genovesi S, Rivera R, Fabbrini P, et al. Dynamic QT interval analysis in uraemic patients receiving chronic haemodialysis. J Hypertens (2003) 21:1921.[CrossRef][Web of Science][Medline]
  15. Cupisti A, Galetta F, Caprioli R, et al. Potassium removal increases the QTc interval dispersion during hemodialysis. Nephron (1999) 82:122–126.[CrossRef][Web of Science][Medline]
  16. Nappi SE, Virtanen VK, Saha HHT, et al. QTc dispersion increases during hemodialysis with low-calcium dialysate. Kidney Int (2000) 57:2117–2122.[CrossRef][Web of Science][Medline]
  17. Howse M, Sastry S, Bell GM. Changes in the corrected QT interval and corrected QT dispersion during haemodialysis. Postgrad Med J (2002) 78:273–275.[Abstract/Free Full Text]
  18. Covic A, Diaconita M, Gusbeth-Tatomir P, et al. Haemodialysis increases QTc interval but not QTc dispersion in ESRD patients without manifest cardiac disease. Nephrol Dial Transplant (2002) 17:2170–2177.[Abstract/Free Full Text]
  19. Severi S, Vecchietti S, Cavalcanti S, et al. Electrocardiographic changes during hemodiafiltration with different potassium removal rates. Blood Purif (2003) 2:381–388.
  20. Floccari F, Aloisi E, Nostro L, et al. QTc interval and QTc dispersion during haemodiafiltration. Nephrology (2004) 9:335–340.[CrossRef][Medline]
  21. Maule S, Veglio M, Mecca F, et al. Autonomic neuropathy and QT interval in hemodialysed patients. Clin Auton Res (2004) V14:233–239.[Web of Science][Medline]
  22. Buemi M, Aloisi E, Coppolino G, et al. The effect of two different protocols of potassium haemodiafiltration on QT dispersion. Nephrol Dial Transplant (2005) 20:1148–1154.[Abstract/Free Full Text]
  23. Severi S, Ciandrini A, Grandi E, et al. Cardiac response to hemodialysis with different cardiovascular tolerance: heart rate variability and QT interval analysis. Hemodial Int (2006) 10:287–293.[CrossRef][Medline]
  24. Drighil A, Madias JE, Yazidi A, et al. P-wave and QRS complex measurements in patients undergoing hemodialysis. J Electrocardiol (2006) 10 5. [Epub ahead of print].
  25. Kantarci G, Ozener C, Tokay S, et al. QT dispersion in hemodialysis and CAPD patients. Nephron (2002) 91:739–741.[CrossRef][Web of Science][Medline]
  26. Malik M, Batchvarov VN. Measurement, interpretation and clinical potential of QT dispersion. J Am Coll Cardiol (2000) 36:1749–1766.[Abstract/Free Full Text]
  27. Vecchietti S, Langley P, Severi S, et al. Effect of premature ventricular beats on manual and automatic repolarization measurements. J Electrocardiol (2004) 37:181–189.[CrossRef][Web of Science][Medline]
  28. Braunwald E, Sonnenblick E, Ross J. The ECG and electrolyte abnormalities. Braunwald E, ed. (1992) Philadelphia: Saunders. 149–151.
  29. Morrison G, Michelson EL, Brown S, et al. Mechanism and prevention of cardiac arrhythmias in chronic hemodialysis patients. Kidney Int (1980) 17:811–819.[Web of Science][Medline]
  30. Redaelli B, Locatelli F, Limido D, et al. Effect of a new model of hemodialysis potassium removal on the control of ventricular arrhythmias. Kidney Int (1996) 5:609–617.
  31. Toussaint N, Cooney P, Kerr PG. Review of dialysate calcium concentration in hemodialysis. Hemodial Int (2006) 10:326–337.[CrossRef][Medline]
  32. Locatelli F, Covic A, Chazot C, et al. Optimal composition of the dialysate, with emphasis on its influence on blood pressure. Nephrol Dial Transplant (2004) 19:785–796.[Abstract/Free Full Text]
  33. Ten Tusscher KH, Noble D, Noble PJ, et al. A model for human ventricular tissue. Am J Physiol Heart Circ Physiol (2004) 286:H1573–H1589.[Abstract/Free Full Text]
  34. Murray A, McLaughlin NB, Campbell RW. Measuring QT dispersion: man versus machine. In: Heart (1997) 77:539–542.[Abstract/Free Full Text]
  35. Linz KW, Meyer R. Control of L-type calcium current during the action potential of guinea-pig ventricular myocytes. J Physiol (1998) 513:425–442.[Abstract/Free Full Text]
  36. Peterson BZ, DeMaria CD, Adelman JP, et al. Calmodulin is the Ca2+ sensor for Ca2+-dependent inactivation of L-type calcium channels. Neuron (1999) 22:549–558.[CrossRef][Web of Science][Medline]
  37. Peterson BZ, Lee JS, Mulle JG, et al. Critical determinants of Ca2+-dependent inactivation within an EF-hand motif of L-type Ca2+ channels. Biophys J (2000) 78:1906–1920.[Web of Science][Medline]
  38. De Vincenzi A, Bellazzi RG, Santagostino M, et al. High- and low-flux acetate-free biofiltration: experimental assessment of calcium mass balance and intact parathyroid hormone behaviour. Nephrol Dial Transplant (1994) 9:1305–1309.[Abstract/Free Full Text]
  39. Schwartz PJ, Wolf S. QT interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation (1978) 57:1074–1077.[Abstract/Free Full Text]
  40. Gussak I, Gussak HM. Sudden cardiac death in nephrology: focus on acquired long QT syndrome. Nephrol Dial Transplant (2007) 22:12–14.[Free Full Text]
  41. Clancy CE, Rudy Y. Na+ channel mutation that causes both Brugada and long-QT syndrome phenotypes: a simulation study of mechanism. Circulation (2002) 105:1208–1213.[Abstract/Free Full Text]
  42. Vecchietti S, Grandi E, Severi S, et al. In silico assessment of Y1795C and Y1795H SCN5A mutations: implication for inherited arrhythmogenic syndromes. Am J Physiol Heart Circ Physiol (2007) 292:H56–H65.[Abstract/Free Full Text]
  43. Gima K, Rudy Y. Ionic current basis of electrocardiographic waveforms: a model study. Circ Res (2002) 90:889–896.[Abstract/Free Full Text]
  44. Rodriguez B, Ferrero JM Jr, Trenor B. Mechanistic investigation of extracellular K+ accumulation during acute myocardial ischemia: a simulation study. Am J Physiol Heart Circ Physiol (2002) 283:H490–H500.[Abstract/Free Full Text]
Received for publication: 28. 7.07
Accepted in revised form: 28. 9.07


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nephrol Dial TransplantHome page
T. B. Drueke and M. Touam
Calcium balance in haemodialysis--do not lower the dialysate calcium concentration too much (con part)
Nephrol. Dial. Transplant., October 1, 2009; 24(10): 2990 - 2993.
[Abstract] [Full Text] [PDF]

Home page
 Phil Trans R Soc AHome page
S. Severi, C. Corsi, and E. Cerbai
From in vivo plasma composition to in vitro cardiac electrophysiology and in silico virtual heart: the extracellular calcium enigma
Phil Trans R Soc A, June 13, 2009; 367(1896): 2203 - 2223.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/4/1378    most recent
gfm765v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Severi, S.
Right arrow Articles by Santoro, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Severi, S.
Right arrow Articles by Santoro, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?