Secondary Logo

Journal Logo

ORIGINAL ARTICLES

Serum Potassium Handling at Pre- and Posthemodialysis in Patients with End-Stage Renal Disease

Muto, Shigeaki*; Sebata, Kiyoko; Ohashi, Michie; Yamada, Takako; Matsumoto, Hisako; Mukouyama, Tomoko; Namiki, Sachiko; Kusano, Eiji*; Asano, Yasushi*

Author Information
doi: 10.1097/01.MAT.0000094485.54491.68
  • Free

Abstract

Because of diminished renal K+ excretion as well as impaired extrarenal K+ tolerance, 1–3 patients with end-stage renal disease (ESRD) are at a high risk for life threatening hyperkalemia. Extrarenal K+ disposition is indispensable for the maintenance of normal serum K+ levels. In healthy humans, only approximately 50% of an intravenous or oral K+ load is excreted into the urine during the first 4 to 6 hours. If all the retained K+ stayed in the extracellular compartment, life threatening hyperkalemia would develop. This is prevented by an acute translocation of K+ into cells. The mechanisms underlying the impaired extrarenal K+ tolerance are not entirely clear in patients with ESRD. Patients treated with maintenance hemodialysis (HD) regularly accumulate K+ during the period between dialyses, and K+ removal is one of the major functions of chronic HD. Because intracellular–extracellular K+ flux is known to be affected by numerous metabolic variables, 4,5 factors other than plasma-dialysate K+ gradient, including insulin, aldosterone, blood pH, epinephrine, and plasma bicarbonate, may also play a role in determining the extrarenal K+ balance. In addition to these factors, suppression of Na+/K+-ATPase activity associated with uremia can be involved in the impaired extrarenal K+ elimination. 4 However, it is not known whether these factors actually influence serum K+ handling during dialysis. Although a postdialysis rebound of the serum or plasma concentrations of several substances has been reported, 6,7 the rapidity and the degree of the rise of serum K+ were described in some HD patients despite the use of a large surface area dialyzer. 8 However, factors influencing the behavior of the postdialysis serum K+ have not fully been investigated. Furthermore, the investigation of these issues has been performed only at 6 hours postdialysis. 8 Because the serum K+ concentration in ESRD patients with anuria is exclusively dependent on the internal distribution of K+ between the intracellular and extracellular compartments, 4,5 measurements of both extracellular and intracellular K+ concentrations are necessary. However, intracellular K+ concentrations at postdialysis have not yet been measured. Therefore, the current study simultaneously measured serum and red blood cell (RBC) K+ as well as blood urea nitrogen (BUN), plasma pH, bicarbonate, insulin, and aldosterone at pre- and end-dialysis and at 5, 11, and 19 hours postdialysis to evaluate the factors influencing serum K+ at end- and postdialysis. We also compared the changes in serum K+ levels with those in BUN levels at the periods described above. Our patients were studied under their current dietary conditions; fasting was avoided because this has been shown to have an unfavorable effect on hyperkalemia. 9

Subjects and Method

Patients

Table 1 shows patients’ profiles. Twenty-five patients with ESRD receiving maintenance HD were studied. They were selected because they repeatedly, although not regularly, had demonstrated predialysis serum K+ levels of more than 5.5 mEq/L, and 10 of these patients received potassium binding exchange resin (sodium polystyrene sulfate) for treatment of hyperkalemia. Patients included 18 women and 7 men aged 44 to 78 years (mean age 62.4 ± 1.7 (standard error of the mean [SEM]) years). The patients had been on HD for 2.1 to 20.7 years. The renal diagnosis of the patients was chronic glomerulonephritis (n = 14), non-insulin dependent diabetic nephropathy (n = 8), lupus nephritis (n = 1), polycystic kidney disease (n = 1), and neurogenic bladder (n = 1). All the patients were oliguric (<100 ml/24 hr) or anephric. All patients were fully informed of the purposes and procedures of this study and gave consent.

Table 1
Table 1:
Patients’ Profile

The patients received dialyses three times a week for 4 hours using dialyzers with a surface area of 1.0 to 1.8 m2. Dialysis with capillary dialyzers was performed three times a week using bicarbonate as a buffer in the dialysate. All patients continued to receive their usual medication, including phosphate binders (calcium or magnesium salts) and allopurinol, and, in some cases, active vitamin D3 (alfacalcidol or calcitriol), Ca2+ channel blocker (benidipine or amlodipine), cadraladine, and doxazosin; nine patients received angiotensin converting enzyme inhibitor (enalapril), only 1 patient received metoprolol, but no patient was receiving digitalis glycosides. Four patients with diabetes received neutral insulin, biphasic isophane insulin, or isophane insulin before breakfast and dinner.

Study Design

It was our purpose to study the patients under their normal, everyday conditions. For this reason, food was allowed. The measurements were performed after the first weekly HD (i.e. on a Tuesday or Wednesday). The patients took their normal light breakfast at home and then went to the clinic. All patients were kept supine, and blood was drawn for measurements of serum K+, insulin, aldosterone, partial pressure of carbon dioxide (pCO2), plasma pH, and RBC K+ concentrations at predialysis (0900 hours), end-dialysis (1300 hours), and 5 (1800 hours), 11 (2400 hours), and 19 hours (0800 hours on the day after HD) postdialysis. At 0900 hours, standard HD fistula needles were placed in the patient. The needle close to the arteriovenous anastomosis was used for blood sampling for measurements of the parameters described above and for dialysis therapy. A standard 4 hour HD was performed using several dialyzers (Table 1), and a dialysate with 2 mEq/L of K+, 140 mEq/L of Na+, 25 mEq/L of bicarbonate, and 5.5 mM of glucose was used. At 1100 hours, patients had a lunch containing approximately 12 mEq of K+. Immediately after blood samples were taken at 5 hours postdialysis (1800 hours), the patients had dinner containing approximately 16 mEq of K+. Furthermore, after blood samples were taken at 19 hours postdialysis, the patients ate breakfast.

Analytical Procedures

Heparinized blood was used for determinations of pH and pCO2 with a blood gas analyzer (Gastat-2, Kanagawa, Japan), serum was separated immediately for subsequent determinations of K+ and BUN with an autoanalyzer (Hitachi 7600–210, Tokyo, Japan), and insulin and aldosterone were separated for radioimmunoassay. Bicarbonate concentration was calculated with the Henderson-Hasselbalch equation. Serum concentrations of insulin and aldosterone were measured by radioimmunoassay at Kotobiken Medical Laboratories (Tokyo, Japan).

Measurement of Red Blood Cell K+ Concentrations

Hematocrit (Ht) was obtained with a multichannel automated hematology analyzer (Sysmex SE-9000, Kobe, Japan). Plasma and whole blood K+ concentrations were determined with an autoanalyzer (Hitachi 7600–210, Tokyo, Japan) and RBC K+ concentrations by the standard formula 10: RBC K+ (mEq/l RBC) = 100/Ht (Kt+ − Kp+ × [1 −Ht/100]), where Kt+ and Kp+ are the K+ concentrations of whole blood and plasma, respectively, in mEq/l, and Ht is the Ht value expressed as a percent. Kt+, Kp+, and Ht were measured in triplicate.

Data Analysis and Statistics

The data are expressed as means ± SEM. Comparisons were performed by Student’s t-test or by one-way analysis of variance (ANOVA) in combination with Fisher’s protected least significant difference, where appropriate. p < 0.05 was considered significant.

Results

Serum and Red Blood Cell K+ Concentrations at Pre-, End-, and Postdialysis

Serum and RBC K+ levels as well as BUN levels at pre-, end- and 5, 11, and 19 hours postdialysis are shown in Figure 1. The pre- and end-dialysis serum K+ concentrations were 5.21 ± 0.13 and 3.28 ± 0.05 mEq/L (n = 25), respectively, and serum K+ concentrations at end-dialysis were significantly (p < 0.001) decreased. BUN levels at end-dialysis (24.1 ± 1.2 mg/dl, p < 0.001, n = 15) were significantly smaller than those at predialysis (74.4 ± 3.0 mg/dl). On the other hand, RBC K+ concentrations at pre- and end-dialysis were 104.6 ± 1.6 and 104.5 ± 2.1 mEq/L RBC (n = 25), respectively, and they were not significantly different between the two periods. At 5 hours postdialysis, serum K+ concentrations rose rapidly from 3.28 ± 0.05 to 4.13 ± 0.10 mEq/L (n = 25, p < 0.001), but RBC K+ levels were reduced from 104.5 ± 2.1 to 98.4 ± 2.3 mEq/L RBC (n = 25, p < 0.001). BUN levels at 5 hours postdialysis (33.3 ± 1.6 mg/dl) were also greater than those at end-dialysis. At 11 hours postdialysis, serum and RBC K+ concentrations were 4.25 ± 0.08 mEq/L (n = 25) and 101.8 ± 2.3 mEq/L RBC (n = 25), respectively, and these values were significantly greater than those at 5 hours postdialysis (Figure 1). BUN levels at 11 hours postdialysis (40.1 ± 1.6 mg/dl) were significantly greater than those at end-dialysis and 5 hours postdialysis. At 19 hours postdialysis, serum K+ concentrations significantly (p < 0.001) increased to 4.71 ± 0.08 mEq/L (n = 25), values that were significantly (p < 0.001) lower than those at predialysis but were significantly greater than those at end-dialysis and at 5 and 11 hours postdialysis. At this time point, RBC K+ concentrations were 102.8 ± 2.6 mEq/L RBC (n = 25), values that were not significantly different from those at pre-, end-dialysis, or at 11 hours postdialysis but were significantly (p < 0.001) greater than those at 5 hours postdialysis. BUN levels at 19 hours postdialysis (45.3 ± 1.7 mg/dl) were greater than those at end-dialysis and at 5 and 11 hours postdialysis but were smaller than those at predialysis.

Figure 1
Figure 1:
Serum and red blood cell (RBC) K+ levels and blood urea nitrogen (BUN) levels at pre-, end-, and 5, 11, and 19 hours postdialysis. Results of serum and RBC K+ concentrations are given as mean ± standard error (SE) of 25 patients. Results of BUN levels are given as mean ± SE of 15 patients. *p < 0.001 vs. predialysis; †p < 0.001 vs. end-dialysis.

The percent increase in serum K+ concentrations at 5 hours postdialysis (25.9 ± 2.2%, n = 25) was significantly greater than that from 5 to 11 hours postdialysis (3.2 ± 1.4%, n = 25) and from 11 to 19 hours postdialysis (11.1 ± 1.3%, n = 25). Also, the increase in serum K+ concentrations per hour at 5 hours postdialysis (0.17 ± 0.016 mEq/L/hr, n = 25) was significantly greater than that from 5 to 11 hours postdialysis (0.02 ± 0.01 mEq/L/hr, n = 25) and that from 11 to 19 hours postdialysis (0.11 ± 0.06 mEq/L/hr, n = 25). Therefore, the postdialysis serum K+ rebound was indeed prominent at an earlier period after dialysis. The percent increase in BUN levels at 5 hours postdialysis (39.8 ± 3.1%, n = 15) was also significantly greater than that from 5 to 11 hours postdialysis (18.5 ± 2.6%, n = 15) and from 11 to 19 hours postdialysis (12.4 ± 2.6%, n = 15). Furthermore, the increase in BUN levels per hour at 5 hours postdialysis (1.85 ± 0.14 mg/dl/hr, n = 15) was significantly greater than that from 5 to 11 hours postdialysis (1.12 ± 0.08 mg/dl/hr, n = 15) and that from 11 to 19 hours postdialysis (0.65 ± 0.03 mg/dl/hr, n = 15). From 5 to 11 hours postdialysis, the percent increase in serum K+ levels was significantly smaller than that from 11 to 19 hours postdialysis, whereas the percent increase in BUN levels was not significantly different from that from 11 to 19 hours postdialysis. Therefore, the postdialysis rebound of BUN and serum K+ was indeed observed, but its pattern was different between BUN and serum K+.

Figure 2 shows the relationship between the predialysis serum K+ concentrations, the end-dialysis serum K+ concentrations, and the magnitude of the decrease in serum K+ at end-dialysis. The end-dialysis serum K+ concentrations positively correlated with the predialysis serum K+ concentrations (r = 0.611, p < 0.001). There was a significant negative correlation (r = −0.934, p < 0.001) between the predialysis serum K+ concentrations and the magnitude of the decrease in serum K+ at end-dialysis. We also observed that serum K+ concentrations at predialysis positively correlated with those at 5 hours (r = 0.674, p < 0.001), 11 hours (r = 0.672, p < 0.001), and 19 hours postdialysis (r = 0.823, p < 0.001). There were also weak but significant correlations between the predialysis serum K+ concentrations and the magnitude of the increase in serum K+ concentrations at 5 hours (r = 0.498, p < 0.05) and 11 hours (r = 0.410, p < 0.05) postdialysis. The correlation between the predialysis serum K+ concentrations and the magnitude of the increase in serum K+ concentrations at postdialysis was stronger at 19 hours postdialysis (r = 0.645, p < 0.001). The magnitude of the decrease in serum K+ at end-dialysis negatively correlated with serum K+ concentrations at 5 hours (r = −0.525, p < 0.01), 11 hours (r = −0.537, p < 0.005), and 19 hours (r = −0.684, p < 0.001) postdialysis. There were also inverse correlations between the magnitude of the decrease in serum K+ at end-dialysis and the magnitude of the increase in serum K+ at 5 hours (r = −0.499, p < 0.05), 11 hours (r = −0.472, p < 0.05), and 19 hours (r = −0.708, p < 0.001) postdialysis.

Figure 2
Figure 2:
Correlation between the predialysis serum K+ concentrations and the end-dialysis serum K+ concentrations as well as the decrease in serum K+ concentrations at end-dialysis (ΔK+[HD]). The predialysis serum K+ concentrations positively correlated with the end-dialysis serum K+ concentrations and negatively correlated with the decrease in serum K+ concentrations at end-dialysis. HD, hemodialysis.

Blood Gas Parameters, Serum Insulin, and Aldosterone at Pre-, End-, and Postdialysis

Plasma pH, plasma bicarbonate, serum insulin, and serum aldosterone at pre- and end-dialysis and at 5, 11, and 19 hours postdialysis are shown in Figure 3. Plasma pH and bicarbonate at end-dialysis were significantly (p < 0.001) greater than those at predialysis. Plasma pH and bicarbonate at 5 hours postdialysis were not significantly different from those at end-dialysis. From 5 to 11 hours postdialysis, plasma pH significantly (p < 0.001) declined, whereas plasma bicarbonate concentrations were not significantly changed. From 11 to 19 hours postdialysis, both plasma pH and bicarbonate concentrations significantly decreased. Serum insulin levels at end-dialysis were significantly (p < 0.001) greater than those at predialysis, indicating that the increased serum insulin levels at this time point are caused by a direct stimulation of insulin secretion in pancreatic β cells by food intake at 1100 hours. At 5 and 11 hours postdialysis, serum insulin levels significantly deceased to similar levels of those at predialysis. At 19 hours postdialysis, serum insulin levels were significantly smaller than those at predialysis and at 5 and 11 hours postdialysis. Serum aldosterone concentrations at end-dialysis were significantly (p < 0.05) smaller than those at predialysis and thereafter remained stable. Similarly, Blumberg et al.11 reported that during the first 1 hour of HD, plasma aldosterone concentrations significantly decreased.

Figure 3
Figure 3:
Plasma pH, bicarbonate, insulin, and aldosterone at pre-, end-, and 5, 11, and 19 hours postdialysis. Results of plasma pH, bicarbonate, and serum insulin are given as mean ± SE of 25 patients. Results of serum aldosterone are given as mean ± SE of 10 patients. *p < 0.001 vs. predialysis; **p < 0.005 vs. predialysis; †p < 0.001 vs. end-dialysis.

At end-dialysis, there were no significant correlations between the magnitude of the decrease in serum K+ levels and changes in plasma pH, bicarbonate, insulin, or aldosterone. On the other hand, from end-dialysis to 5 hours postdialysis, the magnitude of the increase in serum K+ concentrations negatively correlated with the magnitude of the decrease in serum insulin (r = −0.435, p < 0.05), although they did not correlate with changes in plasma pH, bicarbonate, or aldosterone. Serum K+ concentrations at 5 hours postdialysis were not significantly correlated with changes in serum insulin, plasma pH, bicarbonate, or aldosterone. At 11 or 19 hours postdialysis, there were no significant correlations between the magnitude of the increase in serum K+ concentrations, serum K+ concentrations, and changes in plasma pH, bicarbonate, or insulin.

Discussion

Blumberg et al.8 examined plasma K+ handling during and after HD in 14 patients with ESRD and concluded that a rather high dialytic removal of K+ did not necessarily prevent a rapid postdialysis rebound of plasma K+. Feig et al.12 reported that the fractional decrement of plasma K+ concentrations at end-dialysis is a function of predialysis K+ concentrations. However, both groups have not yet measured the intracellular K+ levels during dialysis or postdialysis. The current study extended these studies to measure RBC K+ levels and to determine factors influencing the behavior of serum K+ levels at end- and postdialysis.

Although all patients took food containing approximately 12 mEq of K+ 2 hours after the start of dialysis as a lunch, serum K+ levels at end-dialysis significantly decreased (Figure 1). The magnitude of the decrease in serum K+ at end-dialysis negatively correlated with the predialysis serum K+ concentrations (Figure 2). These findings indicate that dialysis is most efficient in lowering serum K+ concentrations when the predialysis serum K+ levels were greater and the concentration gradient for K+ between the blood and the dialysate was greater. The current findings are similar to those reported by Feig et al.12 and Blumberg et al.8 In dialysis, a large quality of K+ is abruptly removed from the extracellular space. Most of this K+ originates in the cells, traversing the cell membrane, the extracellular space, and the dialysis membrane before reaching the dialysate. However, 4 hours after HD, RBC K+ levels never decreased (Figure 1). These findings are in good agreement with those of Paskalev et al.13 and Rombola et al.,14 except for those described by Paraskevopoulos et al.15 Paraskevopoulos et al.15 reported that RBC K+ levels were significantly decreased from 95.9 to 93.8 mEq/L after HD. The reasons for this discrepancy are not known at present. The plausible explanation is that the discrepancy might be caused by the differences in the dietary conditions and method for measurement of RBC K+ levels. Even if there is no agreement on the value of RBC K+ as a reliable index of K+ content in other tissues, 16 it was shown to be a good index of the direction and magnitude of changes in total body K+ in dialysis patients during HD treatment. 17 Therefore, the current data indicate that, during dialysis, K+ simultaneously shifts from cells into the extracellular space and from the extracellular space into cells with a similar magnitude. Here, the question arises as to how K+ shifts from the extracellular space into cells. Because increased levels of plasma bicarbonate, 5,18 pH, 4,5,19 and insulin 4,5,20 are known to enhance cellular K+ uptake, one can speculate that the increased plasma bicarbonate, pH, or insulin at end-dialysis may promote cellular K+ uptake, although, at end-dialysis, there were no significant correlations between the decrease in serum K+ concentrations and the increases in plasma bicarbonate, pH, or serum insulin. Because the decrease in serum K+ by dialysis is most prominent, the above-described factors lowering serum K+ levels must have been masked.

During postdialysis, serum K+ levels increased, but the magnitude of the increase in serum K+ was not parallel with time. From the end-dialysis to 5 hours postdialysis, serum K+ levels rapidly increased with a greater magnitude, whereas from 5 to 11 hours postdialysis, they increased more slowly, and from 11 to 19 hours postdialysis, they once again more rapidly increased as compared with those observed from 5 to 11 hours postdialysis (Figure 1). Blumberg et al.8 also reported that in spite of the large amount of more than 100 mEq of K+ removed by dialysis, an impressive rebound of plasma K+ concentration occurred even at 1 hour postdialysis.

At 5 hours postdialysis, serum K+ levels increased, but RBC K+ levels decreased, indicating that, at this time point, the movement of K+ from cells into the extracellular fluid is greater than the movement of K+ from the extracellular fluid into cells. From end-dialysis to 5 hours postdialysis, blood gas parameters were virtually unchanged (Figure 3). Also, the increase in serum K+ at 5 hours postdialysis did not correlate with changes in plasma pH, bicarbonate, or aldosterone. Therefore, the changes in plasma pH, bicarbonate, or aldosterone are probably not important for K+ shifts out of cells into the extracellular fluid. On the other hand, at this time point, serum insulin levels significantly decreased, and the decrease in serum insulin negatively correlated with the increase in serum K+. Therefore, we can speculate that at 5 hours postdialysis, the decrease in serum insulin levels may induce an inhibition of cellular K+ uptake, which in turn results in an increase in serum K+ concentrations. This is supported by the reports of Gifford et al., 9 Goecke et al., 21 and Allon et al.22 Gifford et al.9 observed that hyperkalemia occurred during fasting in patients with ESRD. Under these conditions, plasma insulin levels were suppressed, but the rise in serum K+ was not related to either differences in serum bicarbonate levels or K+ intake. From these results, they concluded that hyperkalemia may develop during fasting as a result of insulinopenia, causing transfer of K+ out of cells into the extracellular space. Goeke et al.21 reported that somatostatin infusion resulted in a greater increase of plasma K+ in rats with renal failure as compared with that in control rats. Allon et al.22 have shown that 18 hours of fasting in patients receiving HD induced hyperkalemia and insulinopenia and that the hyperkalemia and insulinopenia were prevented by insulin plus glucose infusion at physiologic doses. Future studies will be required to clarify whether and how insulinopenia actually inhibits cellular K+ uptake in patients with ESRD.

From 5 to 11 hours postdialysis, both serum and RBC K+ levels significantly increased. From 5 to 11 hours postdialysis, plasma pH significantly decreased, but concentrations of plasma bicarbonate or aldosterone were not significantly changed. Serum insulin levels were not significantly changed either. However, the increase in serum insulin levels after dinner could decrease serum K+ levels because the patients took dinner immediately after the blood samples were taken at 5 hours postdialysis. This is supported by the fact that, from 5 to 11 hours postdialysis, the magnitude of the increase in serum K+ was trivial, but the magnitude of the increase in RBC K+ was relatively greater (Figure 1). From 11 to 19 hours postdialysis, serum K+ levels significantly increased, but the RBC K+ levels remained constant, indicating that the movement of K+ from cells into the extracellular space is relatively greater than the movement of K+ from the extracellular space into cells. From 11 to 19 hours postdialysis, plasma pH, concentrations of plasma bicarbonate, and serum insulin were significantly reduced, but serum aldosterone concentrations were unchanged. At 11 or 19 hours postdialysis, there were no significant correlations between the magnitude of the increases in serum K+ and changes in plasma pH, bicarbonate, or insulin. Therefore, from 11 to 19 hours postdialysis, changes in plasma pH, bicarbonate, insulin, or aldosterone are probably not important in the K+ shifts from out of cells into the extracellular space.

Blumberg et al.8 reported that the decrease in plasma K+ at 6 hours postdialysis was not correlated with the predialysis plasma K+. In sharp contrast with their report, we observed that both the magnitude of the increase in serum K+ and serum K+ levels at 5, 11, and 19 hours postdialysis correlated positively with serum K+ levels at predialysis and correlated negatively with the magnitude of the decrease in serum K+ levels at end-dialysis. These results indicate that both the magnitude of the postdialysis serum K+ increase and the postdialysis serum K+ levels can be determined by serum K+ levels at predialysis and by the magnitude of the decrease in serum K+ at end-dialysis. The correlations between both the magnitude of the increase in serum K+ and the serum K+ levels at 19 hours postdialysis and the serum K+ levels at predialysis were stronger than the correlations between both the increase in serum K+ and serum K+ levels at 5 and 11 hours postdialysis and the predialysis serum K+ levels. The correlations between both the increase in serum K+ and serum K+ levels at 19 hours postdialysis and the decrease in serum K+ at end-dialysis were also stronger than the correlations between both the magnitude of the increase in serum K+ and serum K+ levels at 5 and 11 hours postdialysis and the decrease in serum K+ at end-dialysis. BUN levels were minimal at end-dialysis, increased at later postdialysis periods, and were relatively greater at 19 hours postdialysis (Figure 1). Izumo et al.23 demonstrated that the rate of 86Rb pump influx was greater in normal RBCs incubated with postdialysis plasma compared with predialysis RBCs. The defect of the Na+/K+ pump has been shown to improve during dialysis with uremic toxin removal. 24,25 Kelly et al.26 have shown that increased levels of digoxin-like substances have been reported in uremic patients. Therefore, our current findings and these previous reports suggest the possibility that the defect of the Na+/K+ pump activity induced by the substances stored in uremic blood and removed by dialysis may be involved in the postdialysis serum K+ rebound. This possibility will have to await future investigation.

Hypertonic NaCl infusion during hemodiafiltration therapy has been reported to contribute to the interdialytic increase in plasma K+ concentrations. 27 However, in our study, there were no episodes of hypotension during HD, and hypertonic NaCl was not administered at all. The use of angiotensin converting enzyme inhibitors is also associated with an increased risk of developing hyperkalemia in chronic HD patients. 28 In the current study, we compared serum K+ levels at pre-, end-, and postdialysis between the 9 patients who received enalapril and the 16 patients who did not. However, there were no significant differences in serum K+ levels at each time point between the two groups of patients. Therefore, in our study, the use of enalapril does not influence the serum K+ levels at pre-, end-, or postdialysis.

We concluded that (1) the decrease in serum K+ levels during dialysis exclusively depends on removal of K+ by dialysis and is not associated with changes of plasma pH, bicarbonate, insulin, or aldosterone; (2) the magnitude of the increase in serum K+ levels at postdialysis is independent of changes in plasma pH, bicarbonate, and aldosterone and is influenced by both the predialysis serum K+ levels and the decrease in serum K+ levels at end-dialysis; and (3) the serum K+ increase at 5 hours postdialysis can also be related to the decrease in serum insulin. Throughout the study, we did not measure concentrations of epinephrine or norepinephrine in blood, and our data do not exclude the possibility that differences in catecholamine concentrations may have contributed to the increase in serum K+ at postdialysis.

Acknowledgment

This study was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References

1. Bia MJ, DeFronzo RA: Extrarenal potassium homeostasis. Am J Physiol 240: F257–F268, 1981.
2. Fernandez J, Oster JR, Perez GO. Impaired extrarenal disposal of an acute oral potassium load in patients with end-stage renal disease on chronic hemodialysis. Miner Electrolyte Metab 12: 125–129, 1986.
3. Perez GO, Pelleya R, Oster JR, Kem DC, Vaamonde CA: Blunted kaliuresis after an acute load in patients with chronic renal failure. Kidney Int 24: 656–662, 1983.
4. Perez GO: Extrarenal control of serum potassium, in Massry SG, Glassock RJ (eds), Massry and Glassock’s Textbook of Nephrology, 4th ed. Philadelphia: Lippincott Williams & Wilkins, 2001, pp. 289–295.
5. Tannen RL: Dyskalemia, in Massry SG, Glassock RJ (eds), Massry and Glassock’s Textbook of Nephrology, 4th ed. Philadelphia: Lippincott Williams & Wilkins, 2001, pp. 295–307.
6. Pedrini LA, Zereik S, Rasmy S: Causes, kinetics and clinical implications of post-hemodialysis urea rebound. Kidney Int 34: 817–824, 1988.
7. DeSol CA, Umans JG: Phosphate kinetics during high flux hemodialysis. J Am Soc Nephrol 4: 1214–1218, 1993.
8. Blumberg A, Roser HW, Zehnder C, Muller-Brand J: Plasma potassium in patients with terminal failure during and after haemodialysis: relationship with dialytic potassium removal and total body potassium. Nephrol Dial Transplant 12: 1629–1634, 1997.
9. Gifford JD, Rutsky EA, Kirk KA, McDaniel HG: Control of serum potassium during fasting in patients with end-stage renal disease. Kidney Int 35: 90–94, 1989.
10. Lans HS, Stein IF, Meyer KA: The relation of serum potassium to erythrocyte potassium in normal subjects and patients with K deficiency. Am J Med Sci 223: 65–74, 1952.
11. Blumberg A, Weidmann P, Shaw S, Gnadinger M: Effect of various therapeutic approaches on plasma potassium and major regulating factors in terminal renal failure. Am J Med 85: 507–512, 1988.
12. Feig PU, Shook A, Stern RH: Effect of potassium removal during hemodialysis on the plasma potassium concentration. Nephron 27: 25–30, 1981.
13. Paskalev D, Tchankova P, Jankova T, Steiner M, Nenov D: Alteration in red blood cell sodium and potassium concentrations and increased lipid peroxidation in patients with chronic renal failure on maintenance hemodialysis: a hypothesis. Am J Nephrol 14: 246–248, 1994.
14. Rombola G, Colussi G, De Ferari ME, Frontini A, Minetti L: Cardiac arrhythmias and electrolyte changes during haemodialysis. Nephrol Dial Transplant 7: 318–322, 1992.
15. Paraskevopoulos A, Agroyannis B, Kopelias L, et al: Changes in erythrocyte calcium and potassium in patients during HD and CAPD. Int J Artif Organs 23: 750–753, 2000.
16. Patric J: Assessment of body potassium stores. Kidney Int 11: 476–490, 1977.
17. Johny KV, Lawrence JR, O’Halloran MW, Wellby ML, Worthley BW: Studies on total body, serum, and erythrocyte potassium in patients on maintenance haemodialysis. The value of erythrocyte potassium as a measure of body potassium. Nephron 7: 230–240, 1970.
18. Iqbal Z, Friedman EA: Preferred treatment of hyperkalemia in renal insufficiency: survey of nephrology training-program directors. N Engl J Med 320: 60–61, 1989.
19. Adrogue HJ, Madias NE: Changes in plasma potassium concentration during acute acid-base disturbances. Am J Med 71: 456–467, 1981.
20. Alvestrand A, Wahren J, Smith D, DeFronzo RA: Insulin-mediated potassium uptake is normal in uremic and healthy subjects. Am J Physiol 246: E174–E180, 1984.
21. Goecke IA, Bonilla S, Marusic ET, Alvo M: Enhanced insulin sensitivity in extrarenal potassium handling in uremic rats. Kidney Int 39: 39–43, 1991.
22. Allon M, Takeshian A, Shanklin N: Effect of insulin-plus glucose infusion with or without epinephrine on fasting hyperkalemia. Kidney Int 43: 212–217, 1993.
23. Izumo H, Izumo S, DeLuise M, Flier JS: Erythrocyte Na, K pump in uremia. Acute correction of a transporter defect by hemodialysis. J Clin Invest 74: 581–588, 1984.
24. Kaji D, Kahn T: Na+-K+ pump in chronic renal failure. Am J Physiol 252: F785–F793, 1987.
25. Hamlyn JM, Blaustein MP, Bova S, et al: Identification and characterization of a ouabain-like compound from human plasma. Proc Natl Acad Sci U S A 88: 6259–6263, 1991.
26. Kelly RA, O’Hara DS, Mitch WE, et al: Endogenous digitalis-like factors in hypertension and chronic renal insufficiency. Kidney Int 30: 723–729, 1986.
27. De Nicola L, Bellizzi V, Minutolo R, et al: Effect of dialysate sodium concentration on interdialytic increase of potassium. J Am Soc Nephrol 11: 2337–2343, 2000.
28. Knoll G, Sahgal A, Nair RC, et al: Renin-angiotensin blockade and the risk of hyperkalemia in chronic hemodialysis patients. Am J Med 112: 110–114, 2002.
Copyright © 2003 by the American Society for Artificial Internal Organs