Composition of Hemodialysis Fluid

Sodium

Sodium is the major determinant of volume and tonicity of extracellular fluids. As sodium can cross the dialyzer membrane readily, its concentration in dialysis fluid (Na _D ) plays a role in cardiovascular stability during extracorporeal therapy. Acute changes in plasma sodium concentrations are known risks for brain cell damage. Long-term changes in sodium balance can affect patient morbidity via dialysis prescription noncompliance worsening of edema, and blood pressure control.

Because plasma is an aqueous solution of crystalloids and proteins, and plasma proteins (on average 70 g/L) occupy a certain volume, then the volume of plasma water is somewhat less than that of whole plasma. Therefore the concentration of sodium in plasma water, Na _PW , is always greater than that measured in total plasma and can be estimated using certain formulas.

However, the concentration of sodium in the ultrafiltrate, Na _UF , is lower than that in plasma water because of the Donnan effect: some cations cannot cross the dialyzer membrane because they are retained in the blood by negatively charged proteins. When calculating the concentration of sodium available for diffusion, clinicians must correct Na _PW (calculated or measured) for the Donnan effect.

Dialysate with both “high” and “low” Na _D values has been tried, with different clinical effects. Hyponatric dialysate (130 mmol/L) is reported to result in less thirst and interdialytic weight gain. However, not all of the patients treated with hyponatric dialysate in one study showed an improvement in hypertension control, presumably because of high dietary sodium intake or possibly stimulated renin secretion. In another study, dialytic dehydration in hyponatremic patients was obtained predominantly from the extracellular volume, with a high incidence of dialysis dysequilibrium, cramps, and hypotension.

Raising Na _D from 130 to 136 mmol/L resulted in a decrease in reported muscle cramps, and in another study, raising the Na _D from 132.5 through 135 mmol/L to 142 through 145 mmol/L led to lower rates of headache, nausea, and vomiting. Moreover, some patients achieved better immediate hypertension control with higher dialysate sodium concentration, probably as a result of better achievement of “dry weight.” On the other hand, when excretion of sodium is limited in anuric patients undergoing long-term dialysis, positive sodium balance at the end of dialysis sessions can contribute to thirst, greater interdialytic weight gain, and “volume-dependent” arterial hypertension over the long term.

There are different approaches to “normalizing” Na _D concentration. One is to adjust it to Na _PW , to prevent a drop in plasma osmolarity secondary to diffusive losses. Then, the only sodium removed is by convection. Another approach is to aim for normal sodium balance at the end of treatment. Because daily sodium and water intakes are about 100 mmol and 1 L, respectively, adequate Na _D should permit sodium and water removal in this proportion, resulting in an Na _D of approximately 145 mmol/L.

With modern dialysis machines it is possible to change Na _D during the treatment continually, performing so-called sodium profiling. Usually, Na _D is hypertonic at the beginning of treatment, counteracting urea flux from cells to extracellular space while urea removal is at its peak. Then Na _D is reduced progressively, approaching normal at the end of dialysis. Increased Na _D at the time of peak ultrafiltration rate can increase refilling of extracellular compartment with improved venous refill. The limitations of using sodium profiling to limit intradialytic symptoms are the risk of positive sodium balance at the end of dialysis, difficulties in modeling complex interactions among Na _PW , serum protein concentration, total body water, and plasma refilling rate, and variations in the temporal relationship between decreased circulating blood volume and hypotension. The results of these studies depend upon the final dialysate sodium delivered being that which is desired, and manufacturers are allowed an error in the A and B dialysate concentrates. In the future, improved sodium kinetic modeling may help create a software biofeedback loop based on online signals from the patient-machine complex.

Potassium

In patients undergoing long-term dialysis and consuming a liberal diet, daily potassium intake varies between 60 and 80 mmol. With consideration of that fact, thrice-weekly dialysis with a potassium bath of 1.5 to 2.0 mmol/L seems to produce an acceptable potassium balance in most patients. However, relative hypoinsulinemia and metabolic acidosis can shift potassium from the intracellular space to the extracellular space. Clinical scenarios of cytolysis (ischemia, hemolysis, trauma, internal bleeding), renal tubular acidosis type 4 or fasting in patients with diabetes, administration of various medicines (angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, nonsteroidal antiinflammatory drugs, trimethoprim, and nonselective beta blockers) also can contribute to hyperkalemia. Rapid correction of hyperkalemia in metabolic acidosis theoretically can hyperpolarize cells, with persistence of intracellular acidosis. All of these factors, together with dietary variations, dictate personalization of dialysate potassium content. For life-threatening hyperkalemia, a zero-potassium bath is feasible. The contribution of dialysis with a low-potassium bath to the risks of dangerous ventricular ectopy and cardiac arrest is unclear. Supraphysiologic dialysate bicarbonate concentrations in conjunction with higher acetate concentrations may exacerbate the propensity for cardiac arrhythmias by altering the QTc interval resulting from shifts of potassium between the extra- and intracellular space. For patients with poor potassium intake or increased losses through diarrhea, dialysis fluid containing 4 mmol/L of potassium can be advised.