Background

Hyperkalemia is observed in patients with chronic kidney disease (CKD) because of a gradual decrease in renal potassium excretion with CKD stages progression [1, 2]. Therefore, they require hemodialysis (HD) therapy to manage electrolyte disorders, especially potassium imbalance. The serum potassium concentration affects the excitation of the myocardial stimulatory conduction system, and changes in the concentrations may lead to severe complications such as lethal arrhythmia [3]. Hyperkalemia is a known cause of sudden death among patients undergoing HD [2, 4]. Conversely, the pre-dialysis serum potassium concentration is often low in elderly adults with malnutrition undergoing HD. However, hypokalemia is also associated with an increased risk of mortality [5] and is a trigger for lethal arrhythmia [6]. Furthermore, a low potassium dialysate concentration and hypokalemia may increase the risk of sudden cardiac death [7]. Therefore, in patients undergoing HD, both hyperkalemia and hypokalemia may be associated with an increased risk of mortality [5, 8, 9]. In Japan, a dialysate potassium concentration of 2.0 mEq/L is mainly used; however, it is difficult to change this concentration because limited dialysates with different potassium concentrations are commercially available, and they are supplied through a central dialysate delivery system (CDDS) [10]. Here, we have reported a case wherein lethal arrhythmia was prevented by maintaining the serum potassium concentration by increasing the dialysate potassium concentration and administering anti-arrhythmic agents.

Case presentation

A 56-year-old man with Emery-Dreifuss muscular dystrophy and chronic heart failure was admitted to our hospital due to overhydration and diuretic resistance. His medical history included cardiac dysfunction with a low ejection fraction of 33%, ventricular tachycardia (VT), and advanced atrioventricular block complicated by muscular dystrophy. Therefore, a cardiac resynchronization therapy defibrillator (CRT-D) was implanted at the age of 50 years, and treatment with 1.25 mg of enalapril, 80 mg of furosemide, 1 mg of warfarin, 10 mg of rabeprazole, and 25 μg of levothyroxine was initiated. His height was 157 cm and body weight (BW) was 98 kg on admission; however, his BW was 61 kg a year and a half ago. The patient’s vital signs on admission were as follows: blood pressure, 87/68 mmHg; pulse rate, 67 beats/min; and oxygen saturation of the peripheral artery (room air), 92%. Laboratory findings demonstrated the following: white blood cell count, 5460/μL; hemoglobin level, 10.5 g/dL; platelet count, 11.6 × 103/μL; total protein level, 6.5 g/dL; serum albumin level, 3.5 g/dL; blood urea nitrogen level, 36 mg/dL; serum creatinine level, 1.44 mg/dL; serum sodium level, 136 mEq/L; serum potassium level, 5.0 mEq/L; serum calcium level, 8.9 mg/dL; serum phosphate level, 3.8 mg/dL; brain natriuretic peptide level, 183 pg/mL; KL-6 level, 239 U/mL; thyroid stimulating hormone level, 31.1 μIU/mL; free T4 level, 1.01 ng/mL; and free T3 level, 1.47 pg/mL. On physical examination, his heart sounds were regular due to pacing rhythm without murmur; respiratory sounds were coarse crackles; and face, trunk, and bilateral lower extremities were edematous. His chest radiograph, electrocardiogram, and echocardiogram on admission are shown in Fig. 1. Echocardiography showed a decrease in the ejection fraction (35%) with moderate mitral regurgitation. After admission, although the furosemide dosage was increased and tolvaptan was initiated to reduce excess body fluid, his urine volume did not increase. Thereafter, intra-aortic balloon pumping and continuous renal replacement therapy were initiated in the intensive care unit because of acute deterioration in cardiac function and persistent excess body fluid. The excess body fluid was removed, reducing the BW from 97 to 73 kg, and his hemodynamic status gradually improved. However, after cessation of intra-aortic balloon pumping, his urine volume significantly decreased; therefore, HD was needed to manage his body fluid status and mineral and electrolytes imbalance. After initiation of HD therapy, CRT-D was activated twice (42nd and 44th days of hospitalization) to defibrillate the pulseless VT during HD. Despite an infusion of potassium chloride solution and initiation of amiodarone, the patient frequently presented with episodes of non-sustained VT with a decrease in the serum potassium concentration. Therefore, the decrease in the serum potassium concentration after HD and changes in the serum potassium concentration during HD were considered to influence the occurrence of VT. The dialysate potassium concentration was changed from 2.0 to 3.5 mEq/L to minimize changes in the serum potassium concentration during HD. Thereafter, his serum potassium concentration did not fluctuate, and the changes in the serum potassium concentration before and after HD disappeared (Fig. 2). The dialysate potassium concentration (3.5 mEq/L) was adjusted until the 133rd day of hospitalization. An increase in his food intake led to a consequent increase in potassium intake; therefore, the dialysate potassium concentration was decreased to 3.0 mEq/L, along with taking carvedilol, and 1 week later, to 2.0 mEq/L. During this period, defibrillation by the CRT-D did not occur probably because of the improvement of potassium intake and the effect of the anti-arrhythmic agents (amiodarone and carvedilol). The patient was discharged on the 162nd day of hospitalization. He continued HD therapy in another maintenance dialysis unit without CRT-D operation.

Fig. 1
figure 1

a Chest radiograph on admission. b Electrocardiogram on admission. c Echocardiography on admission

Fig. 2
figure 2

Trends in dialysate potassium concentrations and pre- and post-dialysis serum potassium levels during hospitalization. Abbreviations: KCL potassium chloride, CRT-D cardiac resynchronization therapy defibrillator

Discussion and conclusions

Many maintenance dialysis facilities in Japan use the CDDS [10]. This system is useful for the supply of the dialysate with the same electrolyte concentrations for multiple HD machines. Our patient was treated with a standard dialysate used in Japan composed of Na+, 140 mEq/L; K+, 2.0 mEq/L; Cl, 110 mEq/L; Ca2+, 3.0 mEq/L; Mg2+, 1.0 mEq/L; HCO3, 30 mEq/L; and glucose, 100 mg/dL. A dialysate potassium concentration of 2.0 mEq/L is suitable for patients with hyperkalemia undergoing HD; however, a dialysate potassium concentration < 2.0 mEq/L is associated with an increased risk of sudden cardiac arrest due to hypokalemia [3]. In this case, decreases in serum potassium concentration during HD were suspected to be the cause of lethal arrhythmia. The increase in the dialysate potassium concentration to 3.5 mEq/L using a personal dialysate delivery system contributed to the disappearance of lethal arrhythmia. According to the Dialysis Outcomes and Practice Patterns Study, a dialysate potassium concentration of 2.0 mEq/L is used for > 99% patients in the clinical setting of HD therapy in Japan [2]. However, other countries use various dialysate potassium concentrations that can be as low as 1.0 mEq/L and higher than 3.0 mEq/L [2]. Furthermore, the dialysate is usually supplied using a personal dialysate delivery system in many countries except in Japan; therefore, dialysate potassium concentrations can easily be changed using a personal dialysate delivery system compared with those using the CDDS. According to a survey report by the Japanese Society for Dialysis Therapy on the dialysate prescription, 46% of the surveyed facilities reported modifications in their dialysate [11]. Since the CDDS cannot supply individualized dialysate composition [12], the composition can be modified using a personal dialysate delivery system.

The acid–base status influences the intracellular and extracellular potassium homeostasis. In particular, the bicarbonate can affect the potassium redistribution between the intracellular and extracellular fluid, and hypokalemia is induced by the increase in serum bicarbonate concentration, which leads to the increased cell sodium uptake, stimulation of Na+, K+-ATPase activity, and net cellular potassium uptake [13]. In this patient, the bicarbonate concentration before HD was similar during his clinical course (20.5 mEq/L, 21.4 mEq/L, and 20.0 mEq/L at 47th day, 82nd day, and 117th day, respectively). Therefore, there was no association between the serum potassium concentration and bicarbonate concentration before HD in this case. However, we cannot comment on the association between them after HD because the serum bicarbonate after HD was not examined. Calcium and magnesium contribute to the arrythmia development during HD since these cations play an important role in the development of the ventricular action potential and propagation of the electrical impulse [14]. Furthermore, low calcium dialysate (< 2.5 mEq/L), higher corrected serum calcium, and increasing serum dialysate calcium gradient have been associated with an increased risk of sudden cardiac death [15]. According to the laboratory findings obtained on the 40th day of hospitalization, which was closer to the day when CRT-D was activated to defibrillate the pulseless VT during HD, serum corrected calcium concentration before and after HD was 9.0 mg/dL and 9.3 mg/dL, respectively, and serum magnesium concentration before HD was 3.0 mg/dL, with the dialysate calcium concentration of 3.0 mEq/L. Thereafter, no hypocalcemia and hypomagnesemia were confirmed in his clinical course; therefore, in this case, serum calcium and magnesium concentration were considered to have no association with ventricular arrhythmias. In addition, high systolic blood pressure [16], intradialytic hypotension [17], and fluid overload [18] have been associated with the development of cardiac arrhythmias. In both the HD sessions with CRT-D activation in response to the pulseless VT, systolic blood pressure was maintained at around 100 mmHg and no intradialytic hypotension occurred during HD. However, when this HD session was initiated, the body fluid status was noted to be excessive because of the rapid decrease in urine volume; therefore, excessive body fluid might be associated with the occurrence of ventricular arrhythmia.

Furthermore, a decrease in the serum potassium concentration is likely to be associated with a decrease in potassium intake, as seen in patients with malnutrition undergoing HD, in addition to the loss of potassium from the blood to the dialysate. In particular, severe hypokalemia after HD is occasionally observed in hospitalized patients. Therefore, we investigated and compared the serum potassium concentrations between hospitalized patients and outpatients undergoing HD from January 2018 to December 2019. As shown in Fig. 3a, the serum potassium concentration before HD was distributed to the lower side in hospitalized patients compared to that in outpatients, and the proportion of the patients with a serum potassium concentration < 3.5 mEq/L was significantly higher among hospitalized patients (385 of 2593 patients) than among outpatients (43 of 3286 patients; p < 0.0001). Furthermore, after HD, the frequency of serum potassium concentration < 3.5 mEq/L was 47% (Fig. 3b). The incidence of hypokalemia after HD in Japanese patients was previously reported to be 44.8% [9], which is similar to our results.

Fig. 3
figure 3

a Distribution of pre-dialysis serum potassium levels in patients undergoing hemodialysis at our hospital. Black square represents hospitalized patients and white square represents outpatients. b Distribution of post-dialysis serum potassium levels in patients undergoing hemodialysis at our hospital. Black square represents hospitalized patients, and white square represents outpatients. Abbreviations: HD hemodialysis

Table 1 presents a summary of the literature review regarding the relationship between the dialysate potassium concentration and arrhythmia for this particular patient population. Jadoul et al. [19] reported that patients undergoing HD with pre-dialysis serum potassium concentrations of < 5.0 mEq/L could be at risk of sudden death due to the effects of post-dialysis hypokalemia, and prescribing a dialysate potassium concentration of > 3.0 mEq/L might reduce this risk. Ohnishi et al. [20] also reported an increase in the mortality rate associated with a post-dialysis serum potassium concentration of < 3.0 mEq/L. In this case, VT occurred even when the post-HD serum potassium concentration exceeded 3.0 mEq/L because of the high susceptibility of VT to recur as a primary disease. Therefore, along with administering amiodarone, the dialysate potassium concentration was adjusted from 2.0 to 3.5 mEq/L to prevent further decrease in serum potassium concentration after HD and minimize changes in potassium concentration after HD. Thereafter, the lethal arrhythmia disappeared, and this would be associated with the adjustment of dialysate potassium concentration and amiodarone usage. A reduction in the serum-to-dialysate gradient of potassium concentration decreases arrhythmogenic complications or the number of emergency department visits [21, 22]. Moreover, the risk of death was reportedly higher in patients with large differences in serum potassium concentrations before and after HD [20]. Therefore, also in this case, the adjustment of the dialysate potassium concentration might have contributed to the disappearance of lethal arrythmia. However, there have been no specific recommendations for the adjustment of dialysate potassium concentrations in Japan. At our dialysis center, adjustment of dialysate potassium concentration was performed, as shown in Table 2. An adjusted 2.36 mEq/mL sterilized potassium-chloride solution was prepared by our pharmaceutical department. The adjusted solution was added to the undiluted dialysate solution. To ensure the safety and reliability of the adjusted dialysate compositions, we routinely checked the dialysate potassium concentration supplied to the HD machine before HD.

Table 1 Literature review of the relationship between the dialysate potassium concentration and arrhythmia in patients undergoing HD
Table 2 Adjustment of the dialysate potassium concentration in our dialysis unit

In conclusion, we encountered a case in which lethal arrhythmia was prevented by maintaining serum potassium concentration by increasing the dialysate potassium concentration, in addition to the use of anti-arrhythmic agents. In the acute phase of patients with frequent lethal arrhythmia undergoing HD, an increase in the dialysate potassium concentration may be an effective method for preventing arrhythmogenic complications.